You are on page 1of 126




Euro-Mediterranean Wildland Fire Laboratory,
a “wall-less” Laboratory
for Wildland Fire Sciences and Technologies
in the Euro-Mediterranean Region

Deliverable D-08-06

Towards a Euro-Mediterranean
Wildland Fire Danger Rating System

P014: Giovanni BOVIO, Raffaella MARZANO

P023: Inmaculada AGUADO, Emilio CHUVIECO, Jesús MARTÍNEZ
Héctor NIETO, Javier SALAS
P027: Wanda BEROLO, Pierre CARREGA, Dennis FOX, Nuno GERONIMO,
Jean-Guillaume ROBIN
P030: Israel GÓMEZ, Pilar MARTÍN, Javier MARTÍNEZ-VEGA, Lara VILAR
P036: Ioannis GITAS, Michael KARTERIS, Spyros TSAKALIDIS

December 2006

The views expressed are purely those of the writers and may not, in any circumstances, be regarded as stating an
official position of the European Commission

1 Scope and objectives...............................................................................................................................................1
1.1 First phase.....................................................................................................................................................1
1.2 Second phase................................................................................................................................................2
1.3 Third phase....................................................................................................................................................2
2 Proposed methods for wildland fire danger estimation at European scale .............................................................3
3 Review of fire danger rating System developed in MEGAFIRES project................................................................4
3.1 Introduction....................................................................................................................................................4
3.2 Use of meteorological indices and satellite data ...........................................................................................4
3.3 Figures...........................................................................................................................................................6
3.4 Tables............................................................................................................................................................8
3.5 Analysis of long-term fire risk on a European level .......................................................................................9
3.5.1 Selection of risk variables....................................................................................................................9
3.5.2 Techniques to estimate large fire occurrence ...................................................................................10
3.5.3 Figures...............................................................................................................................................12
3.5.4 Tables................................................................................................................................................14
4 Review of fire risk estimation method developed for the SPREAD project ...........................................................15
4.1 Input data for fire risk mapping....................................................................................................................16
4.2 Human danger ignition ................................................................................................................................16
4.3 Probability of Ignition and Fuel Moisture Content .......................................................................................17
4.3.1 Fuel Moisture Content of live fuels ....................................................................................................17
4.3.2 Fuel Moisture Content of dead fuels .................................................................................................17
4.3.3 Probability of Ignition related to the FMC ..........................................................................................17
4.4 Propagation Danger ....................................................................................................................................18
4.4.1 Average Rate of Spread and Flame Length......................................................................................18
4.4.2 Propagation danger versus RoS and FL...........................................................................................18
4.5 Wildland Fire Danger Assessment ..............................................................................................................19
4.6 Figures.........................................................................................................................................................19
5 Review of the European Forest Fire Information System-Risk Forecast ..............................................................22
5.1 Introduction..................................................................................................................................................22
5.2 Meteorological danger indices.....................................................................................................................23
5.3 FPI Model. ...................................................................................................................................................23
5.3.1 General presentation .........................................................................................................................23
5.3.2 Input data...........................................................................................................................................24
5.3.3 Computation of FPI............................................................................................................................26
5.4 Long-term indices........................................................................................................................................26
5.4.1 Probability of Fire Occurrence...........................................................................................................26
5.4.2 Likely Damage...................................................................................................................................27
5.5 Figures.........................................................................................................................................................28
5.6 Tables..........................................................................................................................................................29
6 Basic structure and characteristics of the proposed Euro-Mediterranean Wildland Fire Risk Index ....................31
6.1 General presentation...................................................................................................................................31
6.2 Figures.........................................................................................................................................................33
7 Components of the fire risk system: Ignition Danger Index...................................................................................34
7.1 Fuel Moisture...............................................................................................................................................34
7.1.1 Introduction........................................................................................................................................34
7.1.2 Live fuels (satellite information).........................................................................................................34
7.1.3 Study case in Central Spain ..............................................................................................................35
7.1.4 Probability of ignition related to live fuel moisture content ................................................................39
7.1.5 Probability of Ignition in Dead Fuels (Meteorological Index).............................................................40
7.1.6 Figures...............................................................................................................................................44
7.2 Probability of Ignition (Human Factors).......................................................................................................49
7.2.1 Introduction........................................................................................................................................49
7.2.2 Estimation of human ignition danger at regional scale: the case of Alpes-Maritimes (France) ........50
7.2.3 The third model: statistical approach B. ............................................................................................53
7.2.4 Model comparison and conclusion. ...................................................................................................53
7.2.5 Study Area and Wildland Fire Database ...........................................................................................54
7.2.6 Methodology: Geographically Weighted Regression (GWR), the linear and logistic case ...............55
7.2.7 Results and discussion......................................................................................................................55


7.2.8 Conclusion.........................................................................................................................................56
7.2.9 Figures...............................................................................................................................................57
7.2.10 Tables................................................................................................................................................61
7.3 Human ignition danger in Southern Europe based on fire occurrence maps .............................................63
7.3.1 Fire Occurrence Hot Spot Areas .......................................................................................................63
7.3.2 Figures...............................................................................................................................................65
7.3.3 Tables................................................................................................................................................67
8 Propagation danger index .....................................................................................................................................68
8.1 Average rate of spread and Flame Length..................................................................................................68
8.2 Propagation danger (PD) ............................................................................................................................69
8.3 Figures.........................................................................................................................................................69
9 Vulnerability index..................................................................................................................................................70
9.1 Population vulnerability ...............................................................................................................................70
9.1.1 Introduction........................................................................................................................................70
9.1.2 Example of population vulnerability mapping in Piemonte Region (Italy) .........................................72
9.1.3 Estimation of population vulnerability at Euro-Mediterranean scale .................................................73
9.1.4 Figures...............................................................................................................................................76
9.1.5 Tables................................................................................................................................................80
9.2 Vulnerability related to environmental value ...............................................................................................82
9.2.1 Presentation ......................................................................................................................................82
9.2.2 Figures...............................................................................................................................................83
9.3 Potential soil erosion ...................................................................................................................................85
9.3.1 Introduction........................................................................................................................................85
9.3.2 Objectives..........................................................................................................................................85
9.3.3 A European scale model to predict potential erosion risk .................................................................86
9.3.4 An operational model for the Mediterranean context ........................................................................89
9.3.5 Figures...............................................................................................................................................92
9.3.6 Tables................................................................................................................................................99
10 Euro-Mediterranean Wildland Fire Risk Index.....................................................................................................100
10.1 General presentation.................................................................................................................................100
10.2 Figures.......................................................................................................................................................101
11 References ..........................................................................................................................................................102
12 Annex: Mapping post-fire soil erosion risk...........................................................................................................114
12.1 Introduction................................................................................................................................................114
12.2 The impact of forest fires on soil erodibility ...............................................................................................114
12.3 Site description ..........................................................................................................................................115
12.4 Methods.....................................................................................................................................................116
12.4.1 Mapping soil erosion risk.................................................................................................................116
12.4.2 Soil erosion factors ..........................................................................................................................116
12.4.3 Partial validation of the soil erosion risk map ..................................................................................117
12.5 Results.......................................................................................................................................................117
12.5.1 Distribution of the soil erosion factors .............................................................................................117
12.5.2 The soil erosion risk map ................................................................................................................117
12.5.3 Partial model validation ...................................................................................................................118
12.6 Discussion .................................................................................................................................................118
12.7 Conclusions ...............................................................................................................................................119
12.8 Figures.......................................................................................................................................................120
12.9 Tables........................................................................................................................................................122



The basis and structure have been defined taking

This deliverable is strongly related and in a sense
into account the proposals and suggestion of the
the succession of the deliverables D-08-02 (Wildland
projects analysed in the previous chapter.
fire danger and hazards: a state of the art), D-08-05
In this chapter, it has been determinant the scale of
(Common methods for mapping the wildland fire
work (global) and the availability of homogenous
danger), and D-08-03 (Towards a Euro-Mediterranean
information for entire area of study.
Wildland Fire Danger Rating System: basis, structure).
The structure of the system follows a hierarchic
Deliverable D-08-03 addresses the design of the
scheme with three sub-indices concerning the fire
basis and structure of a prototype of a Euro-
danger associated to ignition, propagation and
Mediterranean Wildland Fire Danger Rating System.
Deliverable D-08-06 addresses the final design of
the system. In chapters 7 to 9, the variables that would have to
integrate each one of the mentioned sub-indices are
The present work has been divided in three main
The methods for their obtaining and calculation are
In chapters 2 to 5, a review of methods already
developed and applied operationally or semi-
The input variables include the fuel moisture content
operationally for estimating and mapping wildland fire
(live and dead), the risk of ignition associated to man,
risks at Euro-Mediterranean scale is presented.
models of simulation of fire behaviour, vulnerability
The results of two previous European projects
related to population, environmental value of territory,
(Megafires and Spread) and the European Forest Fire
and potential soil erosion.
Information System-Risk Forecast –EFFIS- developed
Considering the state of the art and the available
by the JRC are analysed.
information, it has not been possible to analysis all of
In chapter 6 the basic structure of the wildland fire them with the same detail and precision.
danger rating system is proposed.
Final conclusions about the integration of the
This system is composed by the three most
proposed variables and the distribution of the index are
important components in forest fire risk: ignition,
included in chapter 10.
propagation and vulnerability.
Chapter 11 is dedicated to the numerous references
and the annex (chapter 12) presents a specific work
dealing with mapping the post-fire risk of soil erosion




The main objective of the Wildland Fire Risk and
Hazard Unit of the Eufirelab Project (Unit 8) is the First includes a review of methods already
design of the basic structure and characteristics of a developed and applied operationally or semi-
prototype of a Euro-Mediterranean Wildland Fire operationally for estimating and mapping wildland fire
Danger Rating System. risks at Euro-Mediterranean scale (Megafires and
With that intention, the work of this unit has been Spread projects, European Forest Fire Information
structured in three main sections: System-Risk Forecast - EFFIS -).
- Revision of the variables used in existing fire danger In this section it has been realized a detailed
indices (deliverables D-08-02 and D-08-07), analysis of the proposals, at global scale and for the
- Analysis of methods for mapping the wildland fire Euro-Mediterranean basin, about which we have been
danger (deliverable D-08-05), and able to find information.
- Design and development of a prototype of Euro-
Mediterranean Wildland Fire Danger Rating System
(deliverables D-08-03 and D-08-06), based on the
information analyzed in the previous deliverables.
On the basis of the state of the art (existing
systems, necessary inputs, algorithms...) and the
information availability on the variables of interest, the
consortium is intended to develop a prototype of the
Euro-Mediterranean Wildland Fire Danger Rating
System, designing the basic structure and determining
its characteristics.
This final objective will be developed in two
In the first (D-08-03), a proposal of the basic
structure of the index and a broad selection of the
variables to use is carried out.
In the second one (D-08-06), that we present next,
will be define more precisely the structure and the
variables to include in an index that can be used at
global scale for the Euro-Mediterranean Basin.
According to these objectives, the deliverable D-08-
06 "Towards to Euro-Mediterranean Wildland Fire
Danger Rating System" is carried out in three phases.

D-08-06.doc 1


In the second phase is proposed the basic structure In the third phase, after the proposal of this basic
of this risk index to integrate the most important structure, the variables that would have to integrate
components in the ignition, propagation and each one of the mentioned sub-indices are analyzed. In
vulnerability of forest fires. some cases the methods for their obtaining and
This was carried out taking into account the calculation are proposed.
proposals and suggestion made in these previous As it has been previously commented, the selected
projects and the analysis performed in preceding variables will depend significantly on the scale of work
deliverables of this work unit (8), concerning variables and the availability of information.
of risk and methods of integration. For this reason, in this section of the deliverable,
In this section it has been determinant the scale of study cases at local scale are also presented, with the
work (global) and the availability of information (more or intention of transferring that proposal from the local
less homogenous) for entire area of study. scale to a global scale.
These two premises have supposed the proposal of The input variables in the ignition sub-index include
a relatively simple structure, which could become more the fuel moisture content (live and dead) and the risk
complex and rich in local scales of work as well as in associated to man.
the case of having available more homogenous In the sub-index of propagation, models of
information about variables of interest for whole area. simulation of fire behavior are included. Finally,
Even so, as we see in the last section of this variables associated to population, environmental
deliverable, it is not at all simple to propose an value, and soil erosion, are incorporated in the
operative index for a region so extensive and diverse vulnerability sub-index.
(natural and culturally) as Southern Europe. Considering the state of the art and the available
information, it has not been possible to take into
The structure of this index follows a hierarchic
account all of them with the same detail and precision.
scheme in which the final index is obtained from three
sub-indices concerning the fire danger associated to Finally, some recommendations about the
ignition, propagation and vulnerability. integration of the mentioned variables are proposed
These sub-indices offer independent information and some ideas about the distribution of the index are
that it can be used individually. included.
For example, the ignition danger index can be very
useful for the design of the vigilance and monitoring
system, whereas the propagation risk index can be
very interesting for the location of extinction resources.
The vulnerability shows us those zones that could
suppose greater losses in human, ecological and
economic terms.
The combination of these three sub-indices will offer
an overall view about the forest fire danger, considering
all the key elements that need to be valued in an index
of these characteristics.

D-08-06.doc 2


As it was mentioned in the description of work, the Wildland fire danger may be considered at different
review of the state of the art on fire danger and hazard spatial and temporal resolutions: global and local scale
assessment and mapping is one of the main objectives on one hand, and short-term and long-term on the
of Unit 08. other.
The review was divided in two major parts: Both scales of assessments are very important for
- the analysis of the individual environmental and fire management.
anthropogenic variables related with forest fire Considering only spatial scale, global approaches
danger and hazard (accomplished in D-08-02), and assume to cover from millions to dozen million square
- the description of the models and methods currently kilometres, while local studies are focused on hundreds
used to combine those variables to estimate fire to few thousands square kilometres.
danger and hazard and produce maps of their Global analysis can contribute to establish general
spatial distribution at different scales (completed in guidelines for fire management at international level,
D-08-05). such as at European Union level, while local scales are
adapted to specific fire prevention resources of small
A long tradition of research, both European and
regions or provinces.
worldwide, has been devoted to those issues, specially
at local and regional scales, but few attempts has been The type of information to be used in fire risk
made in order to create an integrate fire risk assessment systems is strongly dependent on the
assessment system at coarser scales (i.e. Pan- spatial scale, since global approaches could most
European). probably not include some critical variables that are
In the above mentioned reviews missing links in the easily available at local or regional scale.
integration and spatial assessment of risk information
As it was previously mentioned, in this review
were observed as most of the works were produced
section we will include only those methods for wildland
and tested in specific study areas and they didn´t offer
fire danger estimation that had been proposed and
much information on how to extrapolate the methods to
effectively implemented at European scale, that is,
other areas and, even most important, to other spatial
methods that have produced maps of fire risk/danger
for the whole Europe or, at least, an area covering
The objective of this section is to review existing several European countries.
methods, systems or products that have been Systems that were theoretically developed for being
proposed for fire danger/risk estimation at European applied at global scales but were never implemented or
scale. only in specific test areas are not included in this
This review will allow us to analyse and compare the review.
structure and components of those systems and also to
We have found three systems that fulfill the above
identify their goodness and, eventually, potential
mentioned requirements, two of them are the result of
limitations for being operationally applied to evaluate
european projects “Megafires” and “Spread” and the
fire risk in the framework of national or supranational
last one is a fire risk forecasting system developed and
decision support systems.
implemented by the Joint Research Centre (JRC) in the
This information is required to cope with the second
framework of the INFOREST Action
objective of Unit 08, which is to propose the structure of
a Euro-Mediterranean Widland Fire Danger Rating

D-08-06.doc 3



One of the objectives addressed in the MEGAFiReS
project ( was to In Europe there is no uniform approach to fire
develop methodologies for fire risk rating and mapping, danger rating and different methods have been applied
at both global and local scales, based on satellite and developed in different Mediterranean countries.
remote sensing and Geographic Information System Some studies were made in Minerve I and II
(GIS) techniques. European research projects to compare the fire danger
For the global scale, it was intended to present a rating capabilities of different meteorological danger
comprehensive view of fire risk assessment at the indices, in the attempt to find a common model to be
European Mediterranean countries. applied in the European Mediterranean area, but the
MEGAFiReS project undertook, for the first time in final results left some uncertainty about the best model
Europe, a cross-country evaluation of trends in fuel to use (BOVIO et al., 1994; VIEGAS et al., 1996).
moisture content of Mediterranean species using
One of the problems faced by the danger indices is
coarse resolution remote sensing data.
that they are not specifically tailored to assess extreme
Three different aspects of fire risk assessment were fire danger conditions.
investigated in MEGAFIRES: In MEGAFIReS a first global attempt was made to
- Short-tem fire risk: foliage moisture estimation from develop a system for rating, from ground weather
satellite data measurements, extreme fire conditions in the European
- Fire danger rating using meteorological indices and Mediterranean basin.
satellite data The system is empirical and is made up of a
- Long-term fire risk mapping combined set of selected existing fire danger indices.
Fuel moisture estimation was investigated in On the other hand, the possibility of using satellite
MEGAFIRES using high and low resolution sensors. data to estimate the variables needed to compute the
This was one of the first attempts to correlate daily- danger indices specifically tailored for large fires was
acquired low resolution satellite data with moisture investigated.
content of living foliage on a significant European scale. The objective was to have a more accurate spatial
Validation of the satellite-derived information was distribution of the fire danger estimation that could be
based on the measurements of foliage moisture less dependent on the location and density of the
content of different species carried out during 2 years weather stations.
on 4 networks of test-land plots located in three A global historical fire and meteorological databases
different Euro-Mediterranean countries. were addressed for the whole European Mediterranean
No Pan-European maps were produced and, basin.
therefore, we will not describe here the proposed Meteorological variables required to compute fire
methodology. danger indices were obtained from the database of the
However, the study clearly showed the potential of MARS project implemented at the Joint Research
low-resolution satellite data such as NOAA-AVHRR to Center.
monitor and map foliage moisture content at global In this database, in the time period considered, daily
scales. ground weather data recorded at 12:00 a.m. from about
On the contrary, the other two aspects of fire risk 360 weather stations were interpolated on 1389 grid
that were included in the project, produced methods cells of 50x50 Km (Figure 1).
and results at Euro-Mediterranean scale and, therefore, Historical meteorological data included maximum
they will be described below. and minimum air temperature, vapour pressure, wind
speed, rainfall, potential evaporation and solar
With these weather data, 28 fire danger indices
were computed for each grid cell (daily values for five
The indices computed had different features,
addressing specific components of fire danger (see
table 1).
Due to the high heterogeneity of the fire
environment in the territory, the area of investigation
was restricted and stratified using land cover and
climatic data.

D-08-06.doc 4

Only those grid cells with at least 25% of their area The negative signs found in coefficients were for
covered by wildlands (i.e. the sum of grassland, indices that directly indicated moisture content values,
shrubland and forest land according to the CORINE while positive signs were for indices that increased with
Land Cover map legend) were retained for further danger level.
investigation, so those grid cells mostly covered by
In the typical Mediterranean climatic types (Csa and
urban and agricultural lands were masked out.
Csb classes) the models showed better performances,
In addition, only the Köppen climatic types BSh, while in the semi-arid zone (BS class) predictions were
BSk, Csa and Csb were considered truly representative more uncertain.
of the investigated fire environment. Nevertheless, in all zones and, therefore, at global
The other climatic types, except for some transitory level, the logistic models developed showed better
situations, were in different conditions, often with a performances of each of the 28 individual fire danger
winter-spring fire season and a generally lower level of indices considered in the study.
fire danger.
The indices finally selected should not be
In Figure 2 the area selected for the investigation,
considered as being generally better than the others.
stratified in the climatic types is showed.
Exploratory data analysis showed that many indices
The restriction of this defined area of interest that were not included in the equations had good
resulted in a reduction of the grid cells from 1389 to discriminating power.
499, which multiplied by the number of days considered Those selected were a “set” of danger indices that
(5 fire seasons of 122 days each) give a total of together improved the rating capabilities of
304 390 statistical units. meteorological severe fire conditions.
The empirical model was build using logistic
To assess the model fitting, observed and predicted
regression in order to identify extreme/non extreme fire
EDD were compared.
danger conditions.
This was done grouping the observations according
The objective was to identify the set of indices that
to different criteria (probability classes, climatic zones
together could be used to best estimate the
and month) and also applying the performance score
meteorological conditions for a Danger Day (DD).
for fire danger indices introduced by (MANDALLAZ and
It must be pointed out that large fires often last for YE 1997)
more than one day, that is to say that meteorological
The spatial fitting of the model to the data was
danger conditions associated with them remain for a
qualitatively assessed by means of monthly maps of
certain period.
DDs plotted over the monthly values of the model
Actually, it often happens that a large fire that burns
output in each grid cell. One example of these monthly
for several days spread most of the burned area in few
maps is presented in figure 3.
highly severe days.
The agreement between the spatial distribution of
In order to consider a reasonable statistical unit an the estimated fire severity and large fires ocurrence,
approximation was introduced defining as Extreme changed from year to year, but an overall correct
Danger Day (EDD) a day with at least 1 large fire behaviour of the logistic model was recognised.
burning in the grid cell, excluding multi-day burning The logistic regression analysis provided a first
fires and the day when the fire was extinguished, that is approximation to underline the meteorological indices
the last day, when is was reasonable to expect a lower better adapted to predict large fire occurrence.
level of fire danger. The authors proposed the logistic model as a first
prototype that must undergo further analysis.
The dependent variable in the logistic regression
Besides, the authors consider the usefulness of
was a binomial variable that identified an EDD (value 1
incorporating other refinements such as the integration
as danger day or 0 as non danger day), while the
of satellite images for improving the spatialization of the
explanatory variables were a selected set of danger
indices that provided the best fit.
Exploratory data analysis of indices led to identify Within MEGAFIRES project, the potential use of
and exclude some indices with undesirable distribution satellite data for the spatial extrapolation of specific
properties. components of meteorological fire danger was
After test and build models, better results were
obtained when separating the 3 climatic zones i.e. The indices considered were the Keetch-Byram
building different models in each zone and including Drought Index (KBDI; KEETCH and BYRAM, 1968) and
indices of different types in each model (short, mid and the Canadian Drought Code (DC; VAN WAGNER 1987).
long-term moisture content estimators, spread potential
The analyses were performed at three different
indices and composite indices), showing that when
levels, in a region of Spain, in the Iberian Peninsula
different components of fire danger are taken into
and in the whole European Mediterranean basin.
account the estimates are improved.
Both temporal and spatial dimensions were
All models were globally significant at the 1% level addressed when analysing correlation of satellite
and so were for individual variable coefficients, with variables with meteorological danger indices.
some exception.

D-08-06.doc 5

For Andalusia and the Iberian Peninsula the results The strongest correlation period, from 19 to 23 July,
confirmed the correlation between NDVI and KBDI was with R=-0,74 and 92 records of available data, was
quite poor, but the one between NDVI and DC was then used to develop a regression equation to derive
interesting. DC values as a function of NDVI (figure 4).
For the global approach both meteorological danger The common background of some meteorological
indices (DC and KBDI) were computed with records of fire danger indices and remote sensing-derived indices
the JRC-MARS database for a selection of 56 weather was confirmed by observation at the regional level and
stations. partially at the European Mediterranean basin level,
The satellite data were the same mosaic of NOAA- and this is particularly important for the long-term
AVHRR images used from the correlation analysis with components of fire danger.
the logistic model, with forested area masked out in
Furthermore, the DC, better than the KDBI, seemed
circles of 10 km radius around selected weather
to be more related to living vegetation vigour and its
seasonal trend in Mediterranean environment, as it is
Regarding the temporal analysis, the expected signs
monitored by remote sensing techniques.
of correlation between NDVI and DC (inverse relation)
were obtained for 44 out of 56 weather stations. These consistent similarities, regardless of the local
Due to the few dates available, the statistical climatic differences, strongly support the use of satellite
significance at the 5% level could be established for data to estimate temporal trends in some components
only 9 of them. of meteorological fire danger, and encourage further
investigation on its potential integration into currently
Regarding the spatial dimension, the correlation
operational fire danger indices.
between NDVI and DC calculated for all available dates
showed the expected trend for 31 dates out of 35.

Figure 1: Layout of grid cells in the European Mediterranean Basin and location of large summer fires 1991-1995

Figure 2: Area of investigation (study area) stratified in the climatic types

D-08-06.doc 6

Figure 3: July 1994 DDs and model values in the study area

Figure 4: Logistic model output estimated from ST in the European Mediterranean basin (July 19-23, 1996).
The mapped area is the one for which the logistic model was developed, i.e. the climatic zones Csa, Csb and
BS, with urban and agricultural 50x50 km2 grid cells masked out

D-08-06.doc 7


Table 1: Indices used in the study

Category N Index Description Variable

1 BEHAVE fine fuel moisture model (ROTHERMEL et al. 1986) BEHAVE

Short-term 2 Canadian Fine Fuel Moisture Code (FFMC) (VAN WAGNER 1987) FFMC
3 Moisture content estimate in Mark5 McArthur’s Fire Danger Meter (NOBLE et al) Mark5
estimators 4 McArthur ‘67 fuel moisture model (MCARTHUR 1967) McArth67
5 USA NFDRS 1-Hour Timelag Fuel Moisture (DEEMING et al. 1978) h1
6 NFDRS 10-Hour Timelag Fuel Moisture (BRADSHAW et al. 1983) h10
7 NFDRS 100-Hour Timelag Fuel Moisture (BRADSHAW et al. 1983) h100
Mid-term and
8 Fosberg’s et al. 1000-Hour Timelag Fuel Moisture (FOSBERG et al. 1981) h1000
moisture 9 Canadian Build Up Index (BUI) (VAN WAGNER 1987) BUI
10 Canadian Drought Code (DC) (VAN WAGNER 1987) DC
11 Canadian Duff Moisture Code (DMC) (VAN WAGNER 1987) DMC
12 Keetch-Byram Drought Index (KEETCH and BYRAM 1968)- revision, 1988) KB
13 Canadian Initial Spread Index (ISI) (VAN WAGNER 1987) ISI
Spread 14 Mark3 McArthur’s Fire Danger Meter (NOBLE et al. 1980) Mark3
estimators 15 Mark5 McArthur’s Fire Danger Meter (NOBLE et al. 1980) Mark5F
16 Italian Fire Danger Index (PALMIERI et al. 1992) Impi
17 Canadian Fire Weather Index (FWI) (VAN WAGNER 1987) FWI
18 Portuguese Index (GONÇALVES and LOURENÇO 1990) PortoCif
19 Spanish ICONA Method – probability of ignition (ICONA 1993) Prob
20 Orieux Index (ORIEUX 1979) Orieux
21 Carrega ’87 Index (CARREGA 1990) Carrega
22 Sol Numerical Risk (SOL 1990) SolRisNum
23 IREPI index (BOVIO et al. 1994) IREPI
24 Anderson et al. adsorption (ANDERSON et al. 1978) EWmax
Equilibrium 25 Anderson et al. desorption (ANDERSON et al. 1978) EDmax
26 Van Wagner adsorption (VAN WAGNER 1982) EWmaxW
(EMC) models 27 Van Wagner desorption (VAN WAGNER 1982) EDmaxW
28 Simard’s Equilibrium Moisture Content equations (SIMARD 1968) SEMCmax

D-08-06.doc 8

3.5 ANALYSIS OF LONG-TERM FIRE RISK ON A A whole set of 52 variables taken from demographic
EUROPEAN LEVEL and agricultural census were extracted for
164 provinces of the study area.
The second MEGAFIRES task regarding risk
assessment at Euro-Mediterranen scale was related Regarding the geographical variables, a brief
with the definition of an integrated long-term fire risk description of the spatial analysis undertaken follows:
index. Elevation
The main goal of this analysis (CHUVIECO et al.
1999b) was to identify structural factors of fire risk at Elevation was obtained from two sources: National
European scale. maps at a 1:1,000,000 scale for Spain and Italy, and
As in the case of meteorological indices, fire the GLOBE project (compiled by the U.S. Geological
occurrence was focused on large events (above 500 Survey at 1 Km2 resolution). From the elevation data
hectares). (Figure 5), mean slope and roughness were computed
using algorithms provided by Idrisi G.I.S. (EASTMAN
The basis for the analysis was the compilation of a 1993)
database for the whole study area (Greece, Italy, South
of France, Spain and Portugal). Land cover
The variables were obtained from National or Global
databases: CORINE, Digital Chart of World, NGDC Land cover, which is one of the layers generated for
Globe project, DMSP data. the European Environmental Agency (European
Environtmental Agency 1996) was extracted from the
Census data were the most difficult to compile, due
to the problem of ensuring consistency among the CORINE program.
different countries. Since some regions were missing from the
published CD-ROM, the coverage was completed with
Census data defined the geographical unit of
reference. data directly provided by the EEA.
Since most of the human variables were only The European legend of the CORINE land-cover
program was simplified to six general fuel type
available at a provincial level (NUT-3), this division was
used for all the variables. categories: Grasslands, Shrublands, Perennial,
Broadleaf, Agriculture, and Non-Vegetated (Figure 6).
In the case of geographical layers which cover the Roads and railways density
whole territory (such as land cover, elevation and
DMSP data), the average value for each province was The density of roads and railways was computed
computed to assure consistency among variables. from national maps and the Digital chart of the World.
Within the study area, there are 164 NUT-3
provinces: 48 in Spain, 20 in Italy, 52 in Greece, 18 in Urban areas
Portugal and 26 in France.
Provincial average sizes vary from one country to Urban areas were assumed to be related to fire risk,
another, Italy having the largest and Greece the since recreational uses of forests are among the main
smallest. causes of fire ignition, which can be either accidental or
due to carelessness.
3.5.1 Selection of risk variables Since urban areas are very dynamic, satellite
The selection of variables relating to long-term information was chosen as a source to map urban land
trends of fire occurrence was carried out using previous cover.
technical literature reviews, (CHUVIECO et al. 1997) and Data from the Defence Meteorological Satellite
considering the limitations of data among the different Program (DMSP) provide a global view of city lights,
countries. since this satellite includes a very sensitive radiometer
Three groups of variables were identified: first the operating at night in the visible spectrum (Figure 7).
geographical ones, that deal with terrain features
(climate, land cover, roads, rails, etc.); secondly the The National Geophysical Data Center (NGDC) in
demographic variables, which were related to Boulder, Colorado, has processed these data to
population characteristics, and finally agricultural generate a world database of city lights. (ELVIDGE et al.
variables, dealing with agricultural structure. 1997).
Some variables that could be, at least theoretically, City lights data from Eurasia were extracted from
related to human risk, such as property size, the Internet server at NDGC.
unemployment, and hunting practices were not Climatic regions
considered, because of the difficulty in either obtaining
data or homogenising them among the different Climatic regions were generated from the Joint
countries. Research Center climatic database archived at the
Since rural economies have strongly changed in MARS unit.
most Mediterranean countries during the last 30 years, Average values for the last 30 years were used to
dynamic variables were also included in the analysis, classify each cell, according with the Koeppen method.
by comparing present values with those measured in The original Koeppen system was reduced to a
1960. smaller number of classes and those not presented in
the MEGAFiReS study area were discarded.

D-08-06.doc 9

The final map includes the following climates: The variables finally selected were: LDENSI90
BSh+Csa, BSk, Cfa, Cfb+Cfc, Csb+Csc, Dfb+Dfc. (population density), POBACT90 (Active population
The proportion of each provincial unit per climate 90), LALTIMED (mean elevation), LDECIDU (% of
type was extracted after a cross-tabulation between the broadleaf forest), and CLIM_BCS (Proportion of area
Koeppen system and the provincial limit. with Koeppen climates B and CS).
Table 2 provides the LR coefficients for the final
3.5.2 Techniques to estimate large fire occurrence
variables and their level of significance.
The number of large fires and burned area in each The signs obtained were logical since the fires are
province of the EU Mediterranean countries was expected to be higher at greater population densities,
compiled from national statistics (Figure 8). higher elevations, more arid climatic conditions, less
The study period covered was from 1991 to 1995. active population and less area covered by broadleaf
Presence of large fires in each provincial unit was
selected as the target variable used to estimate fire The variables considered in the equation offer
occurrence from the auxiliary variables previously trends in the correlation with the dependent variable
mentioned. according to what was expected (see column in
The final objective was to better understand how table 2).
these variables explain patterns of large fire occurrence The significance level varies among them, being
in the EU Mediterranean basin. more important for active population, climate and
population density, which should be considered as the
Two techniques were selected to perform this most related to large fire occurrence
analysis: Logistic regression and Artificial Neural
Networks. Table 3 offers an assessment on the performance of
Both techniques had been previously used in the the model.
estimation of fire occurrence providing coherent results.
A global accuracy of 78.26% was obtained in the
Additional analysis was performed with the number
estimation of large fires between 1991 and 1995.
of large fires per provincial unit.
This prediction is quite acceptable, considering the
In this case, linear regression analysis was used
great diversity of the study area, not only in a
instead of logistic regression, since this latter technique
geographical sense but also taking into account the
can only be applied to dicotonomous variables (fire/not
different national particularities regarding fire
Neural networks were applied both to the presence
This performance is even better if only omission
or absence of fires and the number of fires per
errors are considered, since only 11% of the provinces
provincial unit.
where fires occurred were not classified as such. Estimation using logistic regression
Commission errors were higher (almost 40% of the
Different models were tested including different set provinces predicted as having a large fire were not
of variables. affected), but these errors are less critical than the
omission’s from a fire management point of view.
In order to avoid the effect of multiple correlation The LR function derived from the final model makes
and the noise produced by the large number of it possible to compare the geographical distribution of
variables, a previous analysis was carried out. expected versus observed occurrence of large fires
Multiple correlations were computed after grouping (figure 9).
the variables into three categories (geographical, Most provinces were predicted correctly.
demographic and agricultural variables).
These variables were input into three different LR, To perform a final test on the LR model, a new
one for each of the above groups, in order to find out equation was generated from a random sample of 60%
the most significant variables to explain fire occurrence. of all the provinces.
In all cases, the occurrence of large fires (0 not The other 40% of provinces were used as test
affected, 1 affected) was used as dependent variable. cases.
LR was performed with the Stepwise Backward The LR model was created following a similar
Selection algorithm included in the SPSS statistical approach to the previous one.
package. A threshold below 0.5 was also selected to
discriminate between occurrence / not occurrence.
In order to build a global LR model, which includes In this case, as could be anticipated, the fitting is
the geographical, demographic and agricultural poorer that in the previous equation (table 4).
variables, a new correlation matrix was computed
among them, discarding some variables. A global accuracy of 60% was achieved, which
The final model was computed from 161 cases. could be considered a good estimation, especially
Three were discarded because they offered a high taking into account the rate of omission errors (37,5%).
bias. The same variables as the previous model were
identified as significant to explain fire occurrence in
these test provinces.
Geographical distribution of the estimations is
included in Figure 10.

D-08-06.doc 10
EUFIRELAB Estimation using linear regression After a comparative analysis, the omission errors
(the most important ones) were higher for the logistic
Linear regression analysis was applied to estimate regression, having also lower global accuracy than
the number of large fires in each provincial unit. ANN.
Following similar criteria to the logistic regression, a
previous selection of significant variables in each The network work better for typical Mediterranean
thematic group (geographical, demographic and fires (those occurring at summer time), while it failed in
agricultural) was undertaken. those provinces affected by winter fires (provinces
Afterwards, exploratory analysis to the resultant located in the North of Spain, Greece and Italy, and
variables was carried out to assure normal distribution some in South of France).
of the independent variables. Errors from logistic regression did not present a
Some were log-transformed. clear spatial pattern, affecting provinces of different fire
Criteria for selecting the independent variables were characteristics.
based on iterative stepwise forward method with In spite of the flexibility and fitting power of the ANN
thresholds of significance at 0.05. one of the main drawbacks of this technique refers to
The final model was: complexity to find out what are the most significant
NF= 13.467 - 0.273 POBACT90 + 0.161 DIFPOBACT variables that affect fire occurrence.
+ 0.051 BSHCSA - 0.113 DENCAP90 + 0.048 In this aspect, the ANN is like a “black box” with few
DENOVI90 - 0.060 AGRICULT -0.073 DIFREN, analytical possibilities to measure the influence of the
independent variables in the estimation.
where An indirect method to find out the most critical
- NF is the number of large fires; variables of the model was undertaken in the Megafires
- POBACT90 is the percentage of active population in project.
each province for 1990; This method is based on replacing the original
- DIFPOBACT is the difference in active population values of each input variable by random values, after
between 1960 and 1990; the network is trained.
- BSHCSA is the proportion of provincial area It was assumed that the increase in the RMS error
covered by KÖPPEN climates Bsh and Csa; produced by such a change should reflect the relative
- DENCAP90 is the density of goats for 1990; importance of that variable in the whole fitting.
- DENOVI90 is the density of sheeps for 1990; If the variable was significant, randomising it should
- AGRICULT is the proportion of agricultural cover in give us a greater RMS with respect to a marginal one.
each province and Repeating this step with each variable showed the
- DIFREN is the difference in agricultural renters relative importance of all of them.
between 1960 and 1990. Randomising Csb-Csc climates, average distance to
The RMS error of this estimation was 5.4 fires. the roads, active population in 1990, agricultural area, Estimation using artificial neural networks differences in population working in the services
between 1960 and 1990 and population density
The predictive power of artificial neural networks provided the highest increment in RMS.
(ANN) was compared with logistic regression to Therefore, these should be the most critical
estimate large fires ocurrence (CHUVIECO et al. 1998 variables in the estimation of fire-not fire.
and 1999, CARVACHO 1998). Conclusions
The independent variables were the same (human
factors, climatic units, vegetation classes and According to the Logistic Regression model the
topography) used in the regression models. variables most clearly related to large fire occurrences
Also, the same provinces randomly selected to fit was: the proportion of BS-Cs climates, population
the logistic regression were used to train the ANN. density, elevation, unemployment and lack of broadleaf
Then, the trained model was applied to the other cover.
40% of the provinces, studying the predictive capacity The ANN offered a robust estimation of fire
of this technique in comparision with logistic regression. occurrence, but didn't provide as many insights on the
Figure 10 includes the spatial assessment of both most critical variables.
predictions., where is displayed the estimated versus Estimated fire occurrence maps proved to be a
observed provinces with fires/no fires. useful tool for managing global trends of fire risk, by
Table 5 includes the error matrix of ANN analyses. pointing out those regions, which offer a more
consistent and stable risk of being affected by large fire
In the final map (figure 10), red and yellow colors events.
indicate correct predictions (either estimated or Evident difficulties found were related witht the
observed or non estimated and non observed), while availability and homogeneity of socio-economic data at
orange and magenta the errors. European Scale.
The most critical are omission errors (displayed in
magenta) where the models did not predict fires in
provinces that were actually affected

D-08-06.doc 11

3.5.3 Figures

Figure 5: Elevation of theMegafires study area

Figure 6: Fuel type map generated from the Corine Land Cover database

Figure 7: DMSP city lights map of the study area (Source NGDC

D-08-06.doc 12

Figure 8: Occurrence of large fires (above 500 hectares) in the study area 1991-1995).

Figure 9: Actual versus predicted occurrence of large fires by LR techniques 1991-1995)

Figure 10: Observed versus estimated occurrences of large fires in Southern Europe using logistic regression
analysis (top) and neural network analysis (below).
Est-Obs means fires that were both estimated and observed; Nest-Nobs, means both non-estimated and non-
observed; Est-Nobs and Nest-Obs imply commission and omission errors, respectively

D-08-06.doc 13

3.5.4 Tables

Table 2: Coefficients of the final logistic regression model.

Dependent variable occurrence/Not occurrence of large forest fires 1991-1995)
Variable B S.E. Wald df Significance R Exp(B)
LDENSI91 .8349 .2564 10.6024 1 .0011 .1963 2.3045
POBACT90 -.2168 .0500 18.7953 1 .0000 -.2743 .8051
LALTIMED .6631 .2245 8.7219 1 .0031 .1735 1.9408
LDECIDU -.3524 .1572 5.0240 1 .0250 -.1164 .7030
CLIM_BS .0229 .0063 13.1270 1 .0003 .2233 1.0232

Table 3: Observed versus predicted cases

Predicted (cases) % Correct

Observed (cases) 0 1
0 38 25 60.32 %
1 10 88 89.80 %
Overall 78.26 %

Table 4: Error matrix for the estimation of large fire occurrence using logistic regression
(observed versus predicted cases. Test provinces)
Predicted No Fire Fire Totals Comission error
No Fire 14 15 29 51,72%
Fire 11 25 36 30,56%
Totals 25 40 65
Comission error 44 % 37,5% Global accuracy 60,23%

Table 5: Error matrix for the estimation of large fire occurrence using ANN Observed. Test provinces
Predicted No Fire Fire Totals Comission error
No Fire 17 12 29 41,39%
Fire 8 28 36 22,22%
Totals 25 40 65
Comission error 32% 30 % Global accuracy 69,23%

D-08-06.doc 14


The Spread project ( is a The danger component was considered in a broad
European funded project which tried to assess fire risk perspective, covering the probability of a fuel ignites
conditions at several spatial scales, including the Pan- (ignition danger) and the potential hazard that this fire
European. propagates in space and time (propagation danger).
Following Spread Deliverable D162 entitled “Fire The consideration of both components forms the
risk mapping (II): Validated methods and digital wildland fire danger assessment (WFDA) component of
products covering the whole EU Mediterranean Basin the WFR index defined by the Spread project, as
and selected study sites (local – regional – global illustrated in Figure 11.
scale)” we present here a summary of the fire risk
Therefore, the WFDA is based on the estimation of
integration scheme proposed for the European
two properties: ignition danger and propagation danger.
Mediterranean area (EUMed).
The former is related to the causal agents of fire as
Acording to previous literature reviews on fire risk well as the conditions and properties of the fuel, while
terminology, the Spread project studied the fire risk in the latter is associated to estimating the behaviour of
terms of physical probability of fire occurrence the fire.
(danger), on one hand, and potential effects caused by Fire behaviour is mainly described with reference to
the fire on the other (vulnerability) (CHUVIECO et al. the propagation rate and the intensity of the flame front.
The fire danger assessment system includes an
The former refers to the potential that a fire occurs
estimation of fuel moisture content derived from
in a particular area and time on one hand, and to its
satellite data (live fuels) and meteorological variables
propagation capability on the other.
(dead fuels), as well as an estimation of the historical
The other component of fire risk is named fire
patterns of human-caused fires, and the fire
vulnerability, and concerns the potential effects of a
propagation potential, generated from fire-behaviour
fire, either on human values and lives and
simulation programs.
environmental resources.
Other factors, such as risk associated to lighting or
The final index should be computed as the product
flammability, as well as the vulnerability component
of the two components
could not be derived for the whole EUMed, and was not
Trying to adopt an approach that may be both addressed in this project.
scientifically sound and operationally applicable, the
All the input variables were geographically
components that include those two concepts of danger
referenced and included into a dedicated Geographic
and vulnerability were adapted within Spread Project to
Information System (GIS).
different spatial and temporal scales, but it was more
The results were derived for the whole EUMed area
focused on the EUMed scale.
at 1x1 km grid size (with the exception of
meteorological data that were only available at 50x50
km2 grid sizes, and NOAA-AVHRR images with 4.4 x
4.4 km2 pixel size).
The resulting product showed promising potential for
helping fire managers to simulate different danger
scenarios, as well as to obtain a single evaluation of
fire danger conditions for the whole EUMed area size.
A demo was run as a Spread web service during
summer 2004, in which only the fire danger branch
(WFDA) of the framework was implemented and
dynamically updated in quasi-real time during 1 month.

D-08-06.doc 15

Once provided a general explanation of the scheme The human danger ignition HDI layer was generated
adopted in the Spread project, a more detailed from historical data of fire occurrence.
definition of each variable and the data inputs for A probability surface was generated from the
generating the index at EUMed scale follows in this location of ignition points using the kernel density
section. approach.
Kernel density estimation is based on the estimation
The WFDA is a combination of the following
of the density at each intersection of a grid
intermediate maps:
superimposed on the data, after placing a probability
density (kernel) over each point event (GATRELL et al.
- HDI: Human Danger Ignition (static)
1996, LEVINE 2002).
- PD: Propagation Danger (static)
KOUTSIAS et al. (2002, 2004) introduced it in fire
- PI: Probability of Ignition related to fuel moisture
occurrence for assessing fire occurrence patterns at
content (dynamic))
landscape level by addressing some of the inherent
Static means that the intermediate map does not positional inaccuracies of the fire ignition locations.
change throughout the season, dynamic means that
The number of fires observed at community level for
the intermediate map will have to be updated every
the period 1992-2000 was calculated from national fire
day. So the final Wildland Fire Danger Assessment
statistics and expressed using the community
map (WFDA) has to be updated daily.
The adaptive kernel density estimation mode was
chosen due to the non-homogeneous spatial
distribution of community centroids.
The adaptive approach allows for the adjustment of
the bandwidth size in relation to the concentration of
the interpolated points (WORTON 1989).
Locally varying bandwidth size of 10 community
centroids proved to perform best showing a reasonable
variability in the resulting density surfaces avoiding an
under- or over-smoothing.
To avoid over- or under-estimation within and
among countries because of heterogeneities of
different sources the kernel density values within each
country were processed before merging them.
Actually, kernel densities were reclassified to 10
classes based on the equal area criterion within each
country, presupposing equivalence for fire hot spot
areas among the countries
The kernel density interpolation produced
continuous, fire occurrence density surfaces, which in
the Mediterranean context is mainly related to human
fire danger factors, and therefore can be considered as
a static representation of fire danger associated to
human factors thorough a whole fire season.
Figure 12 shows the results of this analysis, which
provides a global view of fire ocurrence distribution,
which is mainly caused by human factors, with higher
occurrences in the NW of Spain and Portugal, SW of
Italy and Greece, the Southern part of the Maritime
Alps, and most territory of Corsica and Sardinia

D-08-06.doc 16


The meteorological data to compute the index were
extracted from the MARS database maintained at the
4.3.1 Fuel Moisture Content of live fuels JRC.
The spatial resolution of this database is 50x50 km.
The estimation of Fuel moisture content (FMC) of
The index was computed daily and transferred to
live fuels was based on an empirical formula derived
the Internet mapserver designed for the semi-
from satellite data.
operational testing of the summer of 2004.
The formula was generated for grasslands and
shrublands based on the analysis of multitemporal Figure 14 shows the spatial distribution of FMC
series of NOAA-AVHRR images. (CHUVIECO et al. estimated from meteorological variables for the same
2004b) period as for the figure 13 (in this case, just a single
The index was derived and tested in Central Spain. day, instead of an 8-day period).
For the calibration study, AVHRR images were
As it can be observed, the spatial patterns are
acquired daily at the University of Alcala’s HRPT
similar in the two figures, in spite of being generated
receiving station, but 8-day composites were created to
from two very different sources
avoid cloud and atmospheric contamination.
The compositing criterion was maximizing surface 4.3.3 Probability of Ignition related to the FMC
temperature of the daily series.
The empirical index was based in two different This PI index was derived from a combination of the
two previous products.
equations, one for grasslands and the other one for
shrublands (Cistus ladanifer was the species selected Both dead and live FMC were converted first to
for the calibration phase). probability of ignition, based on average values of
moisture of extinction (ME). (CHUVIECO et al. 2004a).
For extending this experience to the whole area The values of ME are dependent on the fuel
covered by EU Mediterranean countries, it was used complex, and therefore a fuel type map with a ME
the 10-day NDVI composites of MARS NOAA-AVHRR parameter defined for each fuel type is required to
images produced by the Joint Research Centre (JRC). derive this index.
The composites were derived from LAC data Since no fuel map of Europe with such a
resampled from the original 1.1x1.1 nadir resolution to specification is yet available, an estimation of fuel type
4.4x4.4 km2 pixel size. distribution was extracted from the CORINE land cover
To maintain the original time sequence used for and a ME value assigned to each fuel type.
developing the empirical formula for FMC estimation, in Once the FMC of dead and live fuels and their
the application to the EUMed area the 10-day respective PI were obtained, a global PI for each pixel
composite was updated every 8 days. was computed following:
In addition auxiliary information on the extents of
PIf = PIlive * Live proportion + PIdeath * Death proportion
grasslands and shrublands in each pixel needed to be
derived. Live and Dead proportion expresses the fraction of
This information was extracted from the Corine land live and dead fuel particles in each pixel, which is also
cover map of Europe. dependent on the fuel type.
Figure 13 shows an example of the FMC generated
As a first approach, this map was again generated
from AVHRR images for the whole EUMed area.
from the CORINE land cover map.
4.3.2 Fuel Moisture Content of dead fuels
Figure 15 shows an example of the PIf map
This map is based on the 10h moisture code of the computed for the same days as shown in previous
USA National Fire Danger Rating System, which figures.
requires a basic set of input meteorological variables It should be considered as an integration of the fuel
(air temperature and relative humidity). moisture status of dead and live fuels.
It was shown that this index could be used to The spatial resolution of the AVHRR data improves
estimate the average FMC of dead fuels in Central the resolution provided by the MARS meteorological
Spain (AGUADO et al., in preparation), the area used as database, and therefore, geographical patterns are
calibration site. more evident in the final product.
Additional testing should be done in other
Mediterranean regions for operational use.
The index was computed every day with weather
observations taken at noon.
Moisture content was expressed as percent of dry

D-08-06.doc 17


To provide the propagation danger map, a set of
4.4.1 Average Rate of Spread and Flame Length prototype plots was created, considering all the fuel
models from ROTHERMEL classification and a certain
In order to obtain a global view of risk associated to
slope percentage (from 0 to 100%, with a step of 5%).
fuel loads, terrain characteristics and wind flows, a
The total number of plots was 273.
global simulation analysis was performed.
Each plot consists of a grid of cells with 11 columns
This analysis tried to obtain average values of rate
x 11 rows (each cell measured 328.083 x 328.083
of spread and flame length, considering different wind
and topographic conditions for the estimated fuel maps
The ignition point was located in the middle of the
of the whole EUMed area.
This attempt should be considered as a general
For each plot, many input parameter combinations
overview of average expected fire behaviour at global
were used to simulate the wildland fire behaviour and
scale, in order to rank different danger levels according
the average rate of spread and flame lenght were also
to the combination of fuel and terrain spatial patterns.
The estimation of the average RoS was based on
The parameters considered for variation were: 1-hr
several simulations performed by the Autonomous
dead fuel moisture, 10-hr dead fuel moisture, 100-hr
University of Barcelona for different fuel types, slope
dead fuel moisture, live herbaceous moisture and wind
ranges and wind flows.
speed and direction.
As a simulation kernel the wildland simulator Bevins
Average values of rate of spread (RoS, in m/min)
proposed by COLLINS, which is based on the fireLib
were computed for the different fuel types.
library (COLLINS, 1996) was used.
Finally, the RoS values were scaled into a 0-1
FireLib is a library that encapsulates the BEHAVE
range, by normalizing the values between the
fire behaviour algorithm (MORGAN et al., 2001).
maximum and minimum values (figure 16).
In particular, this simulator uses a cell automata
Fuel types were derived from the Corine land cover.
approach to evaluate fire spread.
The resulting map is considered static, since no
The terrain is divided into square cells and a
specific conditions are simulated (wind or FMC), but
neighbourhood relationship is used to evaluate whether
only general patterns of propagation rates.
a cell will be burnt and at what time the fire will reach
the burnt cells. Same process as above was performed to compute
As inputs, this simulator accepts maps of the terrain, the average flame length (FL), measured in metres and
vegetation characteristics, wind and the initial ignition normalized into a 0-1 scale (figure 17).
4.4.2 Propagation danger versus RoS and FL
The output generated by the simulator consists of
two maps of the terrain. It derives from the combination of the two
In the first one, each cell is labelled with its ignition intermediate products previously described: RoS and
time; in the second one, each cell is labelled with its FL.
flame lenght. The results of the simulations were mapped at
This information must be used to calculate the rate EUmed scale using CORINE land cover (reclassified
of spread and an average from among all flame lenght. into fuel models) and slope maps
To calculate the rate of spread, the distance The maps of FL and RoS were then normalized
between the ignition point and each particular cell in the using linear fitting and multiplied to produce PD:
terrain is divided by the ignition time of that particular
cell. PD = [(RoSi – RoSmin)/(RoSmax - RoSmin)+0.001] *
This calculation is repeated for each cell in the [(FLi – FLmin)/(FLmax - FLmin)+0.001]
terrain to determine the maximum value of the rate of A small constant (0.001) was added to avoid zero
spread. multiplication in case of minimum values. RoS and FL
This maximum value is used as the rate of spread were considered in this formula of equal importance,
for that particular situation. although this could be tuned up in future improvements,
according to further experience or suggestions.
This map is taken as static, i.e. it will not change
throughout the fire season.

D-08-06.doc 18

4.5 WILDLAND FIRE DANGER ASSESSMENT - HDI and PD are static, they do not change
throughout the season,
The final index of Wildland Fire Danger Assessment
- c is a constant to avoid multiplying by zeros (0.001
WFDA proposed in SPREAD integrates the three
is used),
components previously described, the human danger
- PIf changes every 8 days with respect to live fuels
index (HDI), the probability of ignition associated to fuel
and every day with respect to dead fuels.
Status (PIf) and the Propagation Danger (PD), as
follows: Therefore the resulting map of WFDA has to be
updated daily. An example is shown in Figure 18,
WFDA = (HDI+c) * (PIf+c) * (PD+c)
derived using the maps depicted in Figures 12, 15, 16
and 17 for August 10th 2004
All the EUMed maps mentioned so far were made
available as a demo product in quasi-real time (with 2-3
days delay) within the demo of the Spread Project web
based services.

Ignition sources Unintentional

Natural: Lightning
Ignition Danger
Moisture content
Fuel condition Live
Wildland Fire Flammability

Fuel properties
Danger Wind flows

Figure 11: Structure and components of the Wildland Fire Danger assessment (WFDA).
Red boxes mean components that have not been finally implemented in the project

Figure 12: Fire ignition danger in Southern Europe: an estimation of human fire danger factors spatial distribution

D-08-06.doc 19

Figure 13: FMC of live fuels for the first week of August, 2004

Figure 14: Estimation of FMC for dead fuels. Map from 10th August 2004

Figure 15: Probability of ignition from FMC of 10th August 2004

D-08-06.doc 20

Figure 16: Estimated average Rate of Spread (normalized values from 0 to 1)

Figure 17: Estimated average Flame Length (Normalized values from 0 to 1)

Figure 18: Wildland Fire Danger Assessment for August 10th 2004.

D-08-06.doc 21


The fire risk module of EFFIS runs on operational
(EFFIS-Risk Forecast) developed BY the JRC (Joint basis 6 meteorological danger indices (Portuguese,
Research Centre) of the European Commision Spanish, Sol Numerical Risk, Italian, Canadian FWI
and Behave fine fuel moisture).
The European Commission DG Joint Research
In addition a Fire Potential Index (FPI) is computed,
Centre set up since 1999 a research group to work
by integrating:
specifically on the development and implementation of
- forecast meteorological data (to estimate fuel
advanced methods for the evaluation of forest fire risk
moisture content),
and mapping of burnt areas at the European scale.
- satellite data (to estimate the relative fraction of live
These activities led to the development of the
fuels from Relative Greenness) and
European Forest Fire Information System (EFFIS).
- a fuel map (to estimate the fuel properties).
Since the year 2003 EFFIS is part of the Regulation
In this case maps with a spatial resolution of 4.4 km
(EC) No 2152/2003 (Forest Focus) of the European
are generated.
Council and Parliament on monitoring of forests and
Daily EU maps are produced from May 1st to
environmental interactions.
October 31st (from 2006 it will be from February 1st)
All the EFFIS activities are coordinated with DG
processing 0-24 hours and 48-72 hours weather
Environment to reach the final users, Civil Protection
forecast data.
and Forest Services, in the Member States
Together they make the core of the currently named
EFFIS is aimed to provide relevant information for
Expected forest fire risk level is mapped in 5 classes
the protection of forests against fire in Europe
(very low, low, medium, high, and very high) providing
addressing both pre-fire and post-fire conditions.
1 to 3 days risk forecasts over Europe with an average
On the pre-fire phase, EFFIS is focused both on the spatial resolution of about 40 km.
development of systems to provide forest fire risk
Once the indices are computed, they are distributed
forecast based on existing fire risk indices, and on the
to the civil protection and forest fire services via
development of new integrated forest fire risk indicators
(EFFIS - Risk Forecast).
For all indices, maps of past days or averages for a
These indices permit the harmonized assessment of
given historical period can also be generated,
forest fire risk at the European scale.
specifying either a date or a time interval from 1st of
They may be used as tools for the assessment of
May 1st to October 31st
risk situations in cases in which international
cooperation in the field of civil protection is needed.
Currently, the dynamic forest fire risk forecast The long term as well as the advanced integrated
indices are available on the EFFIS web site and sent to approach that will complet the risk assessment system
the Member States Services daily from the 1st of May are currently under further development and are, at the
until the 31st of October. moment, for internal use.
They will be described here as far as the methods,
even at an exploratory research phase, have been
proposed and developed at European scale in several

D-08-06.doc 22


Dynamic indices are short-term indices that assess 5.3.1 General presentation
the probability of fire ignition and spread.
It is a risk index based on the Fire Potential Index
These factors may be derived directly from
created with BURGAN et al 1998), and integrated
meteorological variables, or indirectly by the effect that
through the combination of diverse types of static and
these variables have on vegetation.
dynamic variables from different data sources like
Indices that are computed from meteorological
meteorological stations, satellite images..
variables are referred to as meteorological fire risk
This index has a double structural and dynamic
On the other hand, the indices that evaluate the
The structural component constitutes the
status of the vegetation are the so-called vegetation
information about fuel properties, which is usually used
stress fire risk indices.
by indices at local and regional scale, but not at global
As already mentioned, there is not a consensus on
scale, due to the difficulty to elaborate cartography of
which meteorological fire risk performs best for the
fuel types at this scale.
whole Mediterranean region, although several studies
The dynamic perspective is offered by the
show that the Canadian Fire Weather Index is well
meteorological variables and the remote sensing data,
suited for the estimation of fire risk for the region (SOL,
that serve to characterize the hydrological state of the
dead fuel and the alive fuel presence.
Accordingly, several indices operationally used in
southern Europe are computed in the EFFRFS system. The main assumption behind the FPI model is that
The algorithms in the EFFRFS are those of a “fire potential” can be assessed if the moisture content
system referred to as EUDIC, (BOVIO and CAMIA 2000) of live and dead vegetation is reasonably represented
although the actual computation of the fire risk indices (BURGAN et al 1998).
is now implemented in a distributed geo-database that Therefore, two are the key factors in this model as
enables the on-line integration of all the model inputs. can be seen in Figure 19:
1) live and dead fuel loadings,
The indices were initially computed from data
2) dead fuel moisture content..
collected from a network of meteorological stations
covering Europe in which data are collected at the In the original FPI model, fuel loadings are obtained
station level, and further interpolated to a 50 km by 50 from a fuel model map, while dead fuel moisture
km grid. content is estimated using meteorological data.
At present, these indices are computed from
forecast data, and fire risk indices are produced as Satellite data are used to discriminate between
forecasts for one, two and three days. living and “cured” fuel at the pixel level.
Thus, the model requires the following data inputs:
Six indices are being computed: - A fuel map to define the dead and live fuel loads,
- Behave (Rothermel et al. 1986, VAN WAGNER 1987) and the extinction moisture values for each fuel
based on estimation of the moisture content of fine type.
dead fuel. - Maximum Value Composites (MVC) of NDVI
- Canadian Fire Weather Index (FWI) (VAN WAGNER (Normalized Difference Vegetation Index) to
1987)made up of six normalized indices that calculate the Relative Greenness (RG); the purpose
indicate the daily variation of fuel water content, of the RG in the model is to provide seasonal
initial rate of propagation, and quantity of fuel and adjustment in the proportion of the fuel that is live. In
expected intensity of the flame front. other words, in the FPI model, the proportion of live
- Portuguese index (GONÇALVES and LOURENÇO load varies by pixel, as a function of the assigned
1990) assesses atmospheric conditions in the fuel model, and of the relative greenness. (BURGAN
proximity of the fuel layer. and HARTFORD 1993)
- Spanish ICONA method measures probability of - Meteorological data to estimate the moisture content
ignition (ICONA 1993)based on litter and fine dead of the small dead and cured fuels.
fuels moisture content.
- Sol’s Numerical Risk (SOL 1990, DROUET and SOL Applying the original FPI methodology was
1993) related to ignition and propagation. conceptually possible, but neither the input formats nor
the spatial resolution of the data were comparable to
- Italian Fire Danger Index (PALMIERI et al. 1992)
derived from MC ARTHUR's model. those used in the US.
Consequently, several modifications of the model
European statistics on forest fire data are used for were tested.
the calibration and validation of these indices. The main differences between the original FPI and
All the indices are computed for southern Europe; its European counterpart concern the input data and
however individual civil protection agencies will be able the way these data were processed (e.g. formats,
to download the indices for its own territory. interpolation method, and spatial resolutions).

Based on Sebastián et al (2002).

D-08-06.doc 23

Another important difference is connected to the use Both the RG and the live-ratio were kept at the 4.4
of the fuel map, which in the original model is meant to km spatial resolution of the NDVI images from which
provide a live-ratio and an extinction moisture value for they were derived.
every pixel. Meteorological data
To run the model, a fuel map had to be created,
since such a map did not exist at a European scale. The moisture content of small dead fuels is
5.3.2 Input data considered the major factor in determining fire
ignition.In this model, small dead fuels are represented
The temporal resolution of the FPI index is one day. by the Ten-Hour Time Lag Fuels.
All the input data have this temporal resolution, with The moisture content of small dead fuels (Fm10hr)
the exception of the RG, which is computed with 10 is determined as a proportion of the equilibrium
days Maximum Value Composites (MVC) of NDVI daily moisture content (Emc) of the surrounding atmosphere,
images. as follows: (FOSBERG and DEEMING 1971)
To avoid the uncertainty of pixel location, a 4.4 km Fm10hr = 1.28 x Emc
pixel window was used when deriving the MVC from
which the NDVI is computed (NOAA AVHRR original The Emc determines the amount of water vapour
spatial resolution is 1.1 km). that a given piece of wood can hold. (SIMARD 1968).
Other inputs, in particular, the CORINE Land Cover In order to compute Fm10hr for a given day it is
database (CLC), in its raster version, have a higher necessary to use values for the air temperature,
spatial resolution (0.1 km). relative humidity, precipitation, and percentage of
All the interpolated meteorological data have a lower cloudiness for that day.
spatial resolution (50x50 km). The two last variables are needed to correct for
solar heating and rainfall respectively.
It was decided that the resolution of the NDVI data
would be used as the base resolution of the input data. Except for the percentage of cloudiness – derived
This implied a compromise between the higher from NDVI images - these meteorological variables
resolution of CLC and the lower resolution of were extracted from the MARS meteorological
interpolated meteorological data. database located at the JRC (TERRES 1999).
The FPI would then be derived with a 4.4 km spatial Once this institution receives data at the station
resolution. To match level, then theses data are spatially interpolated into
the spatial resolutions, pixel values were replicated 50x50 km grids according to the CGMS procedure.
for the case of meteorological data, and a two-step This procedure takes the following characteristics
generalisation was performed in the case of the fuel into account: distance, difference in altitude, difference
map, which was initially derived with the same in distance to coast, and climatic barrier separation
resolution as the CLC, i.e. 100 m pixels. (VAN DER GOOT 1997).
The 50x50 km pixel size was imposed by the low Satellite data density of meteorological stations, particularly in some
areas of southern Europe.
At the Joint Research Centre (JRC) a satellite-
tracking antenna receives daily NOAA AVHRR HRPT More specifically, to compute the meteorological
data. Within JRC, these images are pre-processed up inputs of FPI three interpolated variables were
to the computation of the NDVI. extracted from MARS: maximum temperature of the
Finally, the images are geometrically corrected, and day (°C), mean daily vapour pressure (hPa), and total
an overall mosaic of Europe is created at a spatial daily rainfall (mm).
resolution of 4.4 x 4.4 km. The daily rainfall is the sum of precipitation between
Further information on process can be found in 6 UTC on day D and 6 UTC on day D+1.
Kerdiles 1997). For the European FPI, historical NOAA The interpolated actual vapour pressure and the
–AVHRR mosaics processed at JRC were used in interpolated maximum air temperature were used to
several ways: compute the relative humidity - which was not directly
- First, daily cloud masks were extracted from the available from the database- previous computation of
NOAA images and used to derive a daily cloudiness the saturation vapour pressure.
percentage (%) grid for the study area.
- Second, 10-day Maximum Value Composites (MVC)
of NDVI historical images were used to compute
Relative Greenness (RG) (Burgan and Hartford
- Finally, the historical NDVI maximum value at a
given location was used along with the overall
historical maximum to parameterise the live-ratio
values, as explained below.

D-08-06.doc 24

Overlay of the two data sets was done at the CLC’s Derived data sets
spatial resolution, resulting in a 100 m spatial resolution Fuel types map fuel map, which was subsequently generalised to 4.4
km resolution using a two-step procedure.
In Europe, fuel type maps exist only at the national
or local scales in some countries. Notice that this approach has similarities with:
It was thus necessary to derive a European Fuel - other global land cover/vegetation mapping
types map in order to be able to compute the FPI. experiences in which ground truth data was not
This fact implied the need for a method to link the used, such as the 1-km Global Land Cover
ecological aspects of the forest fuels with the fire risk Characteristics database. (LOVELAND et al. 1991),
determinants. - the first phases of other experiences in which
The chosen approach required obtaining, from a ground data was only used in the last phases of the
land cover classification, a general framework for labelling process.
presenting the land cover regions. This is the case, for instance, of the NFDR Fuel
Using GIS tools, these land cover regions were Model Map. (BURGAN et al. 1998)
stratified into sub-regions according to phyto- Live-ratio map
sociological criteria, which accounted for the floristic
composition and other factors governing the distribution Following the original FPI methodology, the
of the vegetation. maximum Live-ratio was initially obtained from a Fuel
types map.
Two databases were thus cross-referenced: the However, the building of such a map, particularly
European classification Corine Land Cover (CLC) and when fieldwork is not feasible, entails some subjectivity.
the Map of Natural Vegetation of Europe (MNVE) In order to address this problem, it was proposed
(Conseil de l’Europe/Commission des Communatés (BURGAN, 1999) the use of a maximum live-ratio map
Européennes 1987). derived through a re-scaling of NDVI historical
The CLC has a fairly good spatial resolution, with a maximum values in each pixel.
minimum mapping unit of 25 ha. The maximum NDVI values were computed for a
However, thematically the CLC can be considered five-year interval 1994-1998), and the live-ratio was
relatively poor, since only 44 classes are distinguished. then derived as:
On the other hand, the MNVE has a lower spatial
resolution (mapping scale of 1:3000.000) but depicts Live-ratio=0.25 + 0.50 x (NDVImax / NDVIabsolute-max)
more than 100 Vegetation associations (V-A). Where
These associations delimit ecological areas that are - NDVImax represents the maximum NDVI for a given
relatively homogeneous, characterised by the location,
predominance of natural or sub-natural primary - NDVI absolute-max is the overall maximum NDVI on
vegetation. any location in the NDVI mosaic.
First of all the CLC was used to mask out The computed live ratio values are scaled between
agricultural and not vegetated land. 0.25 and 0.75, in accordance with the live-ratio values
Then the remaining land cover classes were defined for the NFDRS.
subdivided into different spatial sub-regions according Extinction moisture map
to the existing vegetation associations.
Next, the relationship between these Land-cover / In the FPI model the extinction moisture (ExtM)
V-A units and the standard National Fire Danger Rating value of a pixel is deduced from the fuel type assigned
System (NFDRS) (DEEMING et al. 1978)fuel types was to that pixel.
investigated. Finally, the NFDRS fuel model key was However, the ExtM content of broad vegetation
then used to assign to each identified sub-region a fuel types is normally found in the literature and so it was
type depending on the characteristics of the reasoned that the direct assignment of ExtM values
predominating understory. was simpler, and could be closer to reality, than their
In this process there are several issues that are indirect assignment through the Fuel types map.
worth to stress: Thus it was decided to derive an Extinction moisture
- The MNVE identification of the vegetation map by assigning an ExtM value to each identified sub-
associations is based on knowledge of the region.
ecological factors that govern the vegetation Table 6 shows the CLC classes considered in the
distribution (mainly climate and soil). This analysis and the fuel types assigned to them.
information is also included in the MNVE data set The fuel type before the brackets indicates the
along with the description of each V-A, and was also general assignation.
taken into account when assigning the fuel types Fuel types between brackets correspond to the
with the help of the fuel model key. different identified sub-regions (therefore they depend
- During this process the CLC is essential, because on the predominant vegetation association).
the MNVE depicts potential vegetation associations The fourth column shows the initially assigned ExtM
that sometimes do not coincide with existing ones. values (those from the NFDRS).
In this case the CLC is given priority and the fuel Last column shows the modified extinction moisture
type is assigned directly to the land cover class. values.

D-08-06.doc 25

To obtain this table, several sources of information 5.4 LONG-TERM INDICES

were consulted.
Long-term fire risk indices are those computed from
These comprised the extinction moisture values in
variables that do not change in a short lapse of time.
the Behave and the NFDRS fuel models.
Consequently, these maps are updated with a
All the available documentation was consulted and
yearly (or longer) frequency.
the conclusions derived discussed with experts until the
Two types of long-term indices could be included in
ExtM value was agreed.
These conclusions were used to revise the initially The first one provides the probability of fire
assigned ExtM values, which were only changed in occurrence, while the second one indicates the
case of contradiction with the findings of the experts. likelihood of damage to a forested area in terms of
Table 6 enables the comparison between the ExtM economic or environmental losses.
values initially used, and new ExtM values assigned
5.4.1 Probability of Fire Occurrence2.
when building the ExtM map.
It can be seen in this table that the use of the ExtM This index intends to evaluate the probability of fire
map simplifies the European FPI model. ignition, rather than the propagation of the fire.
Fire propagation is more related to the dynamic
5.3.3 Computation of FPI
indices that have been presented in previous sections.
Once the necessary inputs were derived, following Aspect was included in the model as representative
the original methodology (Burgan et al. 1998), the Fire of the topographic conditions that may influence fire
Potential Index was computed as: ignition.
It is related to the type and condition of the available
FPI = 100 x (1-Fm10hcorrected) x (1-Lr)
fuels for ignition.
Where Fm10hcorrected is the moisture content of the Since the influence of socio-economic factors
small dead fuels corrected by the solar heating and the (human factor) is difficult to model, fire recurrence was
extinction moisture content, and Lr is the Live-ratio introduced in the model as a surrogate of these factors.
corrected by the Relative Greenness, i.e. the part of the Fire statistics for southern Europe show that the
load that is live. probability of fire occurrence is higher in those areas
that have historically suffer a high rate of fire incidence.
FPI values range 1 to 100, with 1 meaning “no fire
risk” and 100 meaning “highest risk”. The method developed is based on multi-variant
For operational purposes this range has been regression techniques.
simplified to 5 risk categories from very low to very
Two sources of data of the European Commission
have been used: on the one hand, the database of fires
The FPI map calculated for the EU is avaliable at
of the Main directorate of Agriculture; on the other
the web site
hand, diverse data bases of the statistical office,
37 independent variables was extracted Of these
two sources of data (see Table 7), including in three
thematic fields: economic, social and environmental
All of the variables were chosen because of their
likely relationship with the forest fires phenomena in the
Mediterranean area as described in the literature on
Some of them, like the fuel types and the
topographic variables, account directly or indirectly for
the hazard.
The others are deemed to somehow affect the
probabilities of ignition.
The third thematic group would integrate the
information relative to fuel types and the fisiographics
variables; the rest would be socio-economic aspects.
In this model the dependent variable is Annual
average of fires, weighted by province (or departement,
in France).
The independent variables were not homogeneous,
neither spatially nor temporally, for all the countries in
the Mediterranean basin, and therefore it was decided
to derive two models.

Based on (Sebastián et al. 2001 and Sebastián, 2004)

D-08-06.doc 26

The first model (Model 1) was built for the five EU 5.4.2 Likely Damage.
Mediterranean countries with the set of 32 variables
There are natural areas that are of particular
that ensure the homogeneity in the temporal and
spatial resolution for the whole study area, i.e.:
This may be due to many different reasons, from the
Portugal, Spain, France, Italy and Greece.
purely economic value of the timber to the unique
Other variables considered important in explaining
environmental qualities of the area.
the fire regime in the Mediterranean, such as surface of
The aim of the likely damage index is to provide a
permanent grasslands, agrarian subsidies, and some
method to highlight areas that should be strongly
economic statistics, were only available (at the required
protected from forest fires.
resolution) for Spain, France and Italy.
Hence, a second model (Model 2) was built for this In the Mediterranean region of Europe the most
reduced region using a total of 37 explicative variables. important quality of most forested areas is their
environmental value.
Concerning the methodology, random sampling
Timber production is usually a secondary asset of
techniques were used to divide the original sample into
these forests.
two subsamples.
An added condition of these forests is the intrinsic
One was used to fit a linear regression model, and
difficulty for regeneration due to the lack of rain and the
the other to validate it.
fragility of soils.
The original sample for Model 1 consisted of
323 observations (i.e. provinces), from which 194 were The likely damage was estimated by assigning to
used to fit the model, and 129 to validate it. each cell a vulnerability degree.
The sample size for Model 2 included
The vulnerability index considers the likely damage
245 observations.
From this sample, 156 observations were used to fit that a fire can cause in a specific area.
the model and 89 to validate the fit. This evaluation can be critical in areas of special
ecological value, in susceptible zones of erosion or
The sampling process was carried out 10 times for prone to the alteration of the hydric balance, and in
both models leading to 10 different and independent areas close to human settlements.
samples to fit the models and 10 corresponding This index considers three factors:
subsets for their validation. - The potential erosion, obtained from the land cover
type, the slope and the regime of rains.
The variable selection was performed by searching
- The protection level of a specific zone, that
for the best intermediate model (sets of explicative
considers its rareness nature, its fragility and its
variables) for each of the obtained sub-samples.
environmental interest.
The information derived from these regressions was
- The distance to human settlements, that considers
then used to fit the final, definitive models.
the human lives and the properties in danger.
The intermediate models were derived using robust
regression, which provides suitable fits even when To derive the mentioned variables, data from the
outliers are present. Eurostat’s database were extracted and processed in a
GIS environment.
An ordered list of the variables that resulted
All the long-term fire risks were normalized between
selected in the final Model 1 (all southern European
the values 0 and 100.
countries) can be found in Table 8; and final Model 2
This range was further divided into five fire risk
(Spain, France and Italy), in Table 9
classes from very low risk to very high risk.
An important consistency was found between the An example of the likely damage index showing the
two models derived. levels of risk and the area for which long-term indices
This supports the thesis of the selected variables as are computed is presented in Figure 21.
the better explicative variables of the forest fire
The two types of long-term indices (probability of fire
phenomena at regional (Mediterranean) scale. The
and likely damage) could be integrated into a single
validation of the models was done through the analysis
of their predictive ability.
This so-called integrated long-term index would
Analysis of the spatial autocorrelation was identify areas that are jointly subject to suffer forest
performed to dextermine whether the selected models damage and high potential losses.
take into account the spatial structure of the dependent The index would help forest fire services locate
variable. those areas to which the highest level of protection
should be given.
6 variables were coincident between the two
models: two types of land cover (presence of shrub and
mixed forest), two variables relative to the agrarian
production (cattle and cereal production), a socio-
economic variable (unemployment) and another
fisiographical one (altitude).
Figure 20 presents an example of a fire probability
map for southern Europe.

D-08-06.doc 27


Figure 19: Structure of FPI model calculation

Figure 20: Example of a fire probability map for southern Europe (San Miguel-Ayanz et al. 2002)

Figure 21: Example of a likely damage map for southern Europe (San Miguel-Ayanz et al 2002)

D-08-06.doc 28


Table 6: Proposed ExtM values for the considered CLC classes.

CLC Description Fuel type Live ratio Initially assigned Ext. New assigned
assigned Moist. (%) Ext. Moisture
Permanent grassland L 0.67 15 15
Agriculture with natural vegetation T 0.67 15 20
Agroforestry areas L(C) 0.67 15 (20) 15
Broad-leaved forest R ( O,F, T ) 0.40 25 (30,15,15) 30
Coniferous forest T (D,H,C, P,A) 0.67 15 (30, 20, 20, 30, 25, 15) 25
Mixed forest R (O,P,C) 0.40 25 (30, 30, 20) 25
Natural grassland L(A) 0.67 15 (15) 15
Moors and heathland O 0.53 30 30
Sclerophillous vegetation F 0.60 15 15
Transitional woodland-scrub T 0.67 15 15
Sparsely vegetated areas A(S) 0.60 15 (25) 15
Inland marshes N 0.40 25 25
Peat bogs S 0.50 25 25
Salt marhes N 0.40 25 25

Table 7: Independents variables used in the study (Sebastián et al., 2004).

Type Variable
Cereal production
Wheat production
Barley production
Corn production
Agrarian production Leguminous production
Agrarian subsidies
Agricultural total production
Economic Variables Permanent grassland area
Agricultural total area
Goat production
Bovine production
Cattle production
Ovine production
Addition goat and ovine production
Primary sector
Production by economic sectors Secondary sector
Tertiary sector
> 25 years old
Unemployed index
< 25 years old
Population density
Social Variables
Settlement > 20.000 inhabitants
Population variables
Protected areas
Road density
Fisiographic variables Slope
Land cover type 1 (%)
Land cover type 2 (%)
Land cover type 3 (%)
Land cover type 4 (%)
Environmental Areas Land cover type 5 (%)
Land cover type 6 (%)
Fuel types
Land cover type 7 (%)
Land cover type 8 (%)
Land cover type 9 (%)
Land cover type 10 (%)
Land cover type 11 (%)
Land cover type 12 (%)

D-08-06.doc 29

Table 8: Selected variables in Model 1 (Sebastian et al., 2001).

Variables in Final Model 1 Reg. Coeff.
Ln(Percentage of land occupied by Shrubs) 0.1195
Ln(Percent. of land occupied by Mixed Forests) 0.0973
Unemployment 0.0944
Ln(Surface of Permanent grasslands) -0.0936
Altitude -0.037
Ln(Percent. of land occupied by Broad leaves) 0.033
Total Cereal Production -0.0352
Total Cattle Production 0.0255

Table 9: Selected variables in Model 2 (Sebastian et al., 2001).

Variables in Final Model 2 Reg. Coeff.
Ln(Percentage of land occupied by Shrubs) 0.3775
Ln(Percent. of land occupied by Burnt areas) 0.2387
Total Cereal Production -0.2615
Ln(Percent. of land occupied by Mixed Forests) 0.2168
Altitude -0.2115
Total Goats production -0.1525
Ln(Perc. Land occup. by Mixed Forest- Agriculture) 0.1154
Total Cattle production -0.1095
Unemployment 0.1087

D-08-06.doc 30



6.1 GENERAL PRESENTATION The assessment of the importance of the expected

damage is closely related to the concept of
Fire risk assessment is routinely performed in most
vulnerability, which can be defined as the potential
developed countries that are affected by forest fires.
damage caused by the natural/technological disaster
Estimating forest fire risk involves identifying the
on persons and values (ecological, economic, etc.).
potentially contributing variables and integrating them
Consequently, an integrated assessment of fire risk
into a mathematical expression, i.e. an index.
should consider both likely occurrence of fires, as well
This index, therefore, quantifies and indicates the
as their potential damages (Figure 22).
level of risk.
A literature review of forest fire risk methods has The Eufirelab unit 8 intends to propose a coherent
shown how different approaches are used for the and integrated scheme for fire risk assessment, which
evaluation of fire risk. would consider a wide range of risk variables, both
The most relevant are producing updating associated to fire ignition and propagation, as well as to
evaluation of fire danger conditions over the Internet. fire vulnerability.
Such a proposal will be based on previous
Most of these operational systems currently rely
experiences which have shown good potential for risk
mainly on meteorological data (see section 2.3), which
is considered the most critical variable for fire ignition
As previously stated, most operational fire risk
and propagation, while other factors of risk are less
systems are focused on physical factors: weather data
considered or they are introduced in an experimental
and fuel status mainly.
However, especially in Europe, a growing concern
Fuel loads and fuel water state is commonly
about the role of human activities in fire ignition and
considered, but the difficulties of obtaining accurate
propagation is widely recognized (KALABOKIDIS et al.,
information on both variables are recognised.
2002; MARTÍNEZ et al., 2004; VÉLEZ et al., 2002).
Therefore, most commonly fuel loads are roughly
Therefore, any risk assessment system should also
estimated from land cover maps, and fuel state is
include the consideration of socio-economic causes of
associated to weather variation.
fire, as well as the vulnerability associated to human
Additionally, human factors of fire risk are rarely
beings (life, properties and values).
considered, in spite of being considered a critical
The proposed scheme will include these factors.
variable in fire ignition (MARTELL et al., 1987).
In addition to this lack of critical variables in fire Another aspect to be taken into consideration to
danger estimation, another weakness of current fire risk design an integrated risk assessment scheme regards
assessment systems is related to the concept of risk the need of having consistent evaluation of risk at
itself. different temporal and spatial scales.
Traditionally fire managers define fire risk The great diversity of natural and socio-economic
considering the chance of starting or spreading a fire in conditions strongly complicates this task, which is
a particular time and space. critical if risk conditions need to be objectively
According to FAO’s terminology (FAO 1986) forest compared throughout Europe.
fire risk is even more restricted, since is defined as “the Obviously, any risk assessment system based on
chance of a fire starting as determined by the presence physical and human variables need to be adapted to
and activity of any causative agent”. local conditions, but a hierarchical design may help the
However, other natural hazards evaluate risk use of a common scheme at different scales, and for a
considering not only the potential occurrence of an broad range of social and natural conditions.
event, but also –and more important in many cases- On the other hand, Fire risk, as any other natural
the potential damage which that event would cause on hazard, is not static, but needs to be updated regularly.
persons and values. The temporal updating would strongly depend on
From this point of view, the traditional concept of fire how dynamic the different risk variables are.
risk, which is still active in most countries and regions, Some would require a very short updating period
is not considering a critical component of risk, which (such as weather data), while others only need revising
affects the susceptibility to fire of the potentially in few months-years (fuel types), or they can be
affected region. considered static (elevation, slope).
As it is well known nowadays, fire is a natural factor
Based on the indices revised in the preceding
in many ecosystems, and therefore, from the
section and considering the availability of spatial
ecosystem point of view, the impact of a given wildfire
information at the required scale (see previous
could be relatively contained, being within the limits of a
deliverables of the unit), Eufirelab unit 8 will aim to
normal natural disturbance.
define the scheme for fire risk integration.
Fire management should greatly benefit from having
It includes the definition of the risk variables, as well
the possibility to consider in the decision making
as the integration of those input variables in synthetic
process the importance of the expected potential
indices for fire risk mapping at European scale.

D-08-06.doc 31

The system will include the assessment of Danger

As far as vulnerability assessment concerns, the
and Vulnerability (Figure 23).
main potential effects to be analysed have been divided
The former is related to the probability of fire
into ecological and socio-economic factors.
occurrence and propagation, while the latter is
Due to the small number of experiences using this
associated to the potential effects of fire on human
component in forest fire risk index at global scale, a
beings and natural values.
preliminary study of the variables to be included in a
The proposed system includes the estimation of two vulnerability index will be undertaken in this deliverable.
danger properties: ignition danger and propagation The socio-economic component of fire vulnerability
danger. will be focused on the potential damages on persons
The former is related to the causal agents of fire and properties.
(excluding lightning because no operational way to The economic value (in different colour in the
model this factor has been produced) as well as the scheme) will not be addressed in this deliverable
conditions and properties of the fuel (live and dead), because the great difficulty to create this information at
while the latter is associated to estimating the European scale.
behaviour of the fire. The ecological component includes potential soil
In this deliverable will be defined the two erosion after a forest fire and the environmental value
components of ignition danger (causal agents and fuel of the territory.
The system will be designed in order to be
The propagation danger is related to fire behaviour
implemented at European scale.
(rate of spread and flame length).
The target resolution will be of 1 km².
The Danger component requires considering The supra-national perspective, which will be
weather data, fuels and socio-economic data that proposed for the EUMed countries, should potentially
should be derived from spatial analysis of point data contribute to the improvement of the European Forest
(both measured and forecasted), as well as remote Fire Information System (EFFIS), currently run at JRC.
sensing images.
More details can be found in following sections.

D-08-06.doc 32


Figure 22: Integrated Fire Danger Rating (after ALLGÖWER et al., 2003)

Live fuels (satellite)

Ignition Danger Dead fuels (meteo)

Probability of ignition Human

(human factors) ignition

EM - WFRI Danger Index

Population exposition

Vulnerability Environmental value

Economic value

Potencial Soil erosion

Figure 23: Scheme of the Euro-Mediterranean Wildland Fire Risk Index (EM-WFRI)

D-08-06.doc 33



7.1.2 Live fuels (satellite information)
7.1.1 Introduction
Remote sensing data have been frequently used to
The moisture content of fuel is a critical parameter in estimate the water status of plants, both in agricultural
fire ignition because flammability is closely dependent and ecological research (CARTER, 1991).
In the forest fire danger literature, the water content
The burning capacity of the leaves and plants is
of plants is commonly expressed as fuel moisture
inversely correlated to the moisture quantity.
content (FMC), defined as the percentage of water
For fire ignition danger, fuel could be divided in two
weight over sample dry weight:
categories: dead and live.
Dead fuels lying on the forest floor (fallen branches, FMC=[(Ww-Wd)/Wd)*100]
litter, and foliage) are the most dangerous because they
are drier than live fuels and more dependent on rapid where Ww is the wet weight and Wd is the dry weight
atmospheric changes. of the same sample.
The moisture content of live fuels has a marginal This variable is mostly obtained through field
role in fire ignition, but it is critical in fire propagation sampling using gravimetric methods (VIEGAS et al.,
modelling because the amount of water is directly 1992).
related to the rate of fire spread (CARLSON & BURGAN, Wet samples are:
2003; SNEEUWJAGT & PEET, 1985; VIEGAS, 1998). - weighed, then
It is not easy to measure the water content of the - oven-dried at 60 or 100 oC, and
- weighed again to determine the dry weight.
leaves and plants in an operational way.
Two main directions are generally used: FMC can be referred to for both live and dead
- water estimation from soil water quantity estimation, species.
using methods of water reservoir -the water quantity
of soil is computed from rainfall values (input) and Within the context of fire danger estimation, good
evapotranspiration (output) correlations between live FMC and multitemporal series
- water estimation from remote sensing of NOAA–AVHRR data have been found for
measurements of vegetation aspect. This first way is herbaceous species using normalized difference
generally more or less integrated in the vegetation index (NDVI) data (CHLADIL & NUNEZ, 1995;
meteorological indices of fire risk. PALTRIDGE & BARBER, 1988), but problems were found
for shrubs and trees (CHUVIECO et al., 1999a; LEBLON,
Commonly, the estimation of dead fuel moisture 2001).
content (FMC) is based on meteorological danger
indices, which attempt to account for the absorption– However, in the remote sensing literature, water
evaporation relationships in inert materials (SIMARD, content is usually expressed as the equivalent water
1968). thickness (EWT: water content/leaf surface), instead of
The application of spatial interpolation techniques is FMC, because EWT is directly related to the absorption
required because meteorological data are frequently not depth of the leaf.
available for fire prone areas. Laboratory spectral measurements have been
Applying those indices to live fuel moisture trends performed to estimate EWT, showing divergent results
could be more complex because live plants are much in the visible and near infrared (NIR) depending on
less dependent on atmospheric conditions than dead whether they were done at leaf or canopy level,
materials, given their mechanisms to extract water from because of the indirect effects of water content changes
the soil reserve and reduce evapotranspiration. on the whole plant (mainly through the modification of
Recent research have obtained strong nonlinear the leaf area index [LAI]).
relationships between live fuels moisture content and However, short wave infrared bands (SWIR: 1.1–2.5
long-term meteorological codes (CASTRO et al., 2003; Am) have proven to be the most sensitive to EWT
VIEGAS et al., 2001), but their results are species variations (BOWMAN, 1989; COHEN, 1991; DATT, 1999),
dependant. In this context, spatial interpolation although additional bands are required to reduce the
techniques could introduce additional noise. uncertainty caused by other variables affecting SWIR
Within the context of fire danger estimation, remote reflectance.
sensing data have shown good correlations with live Simulation studies based on radiate transfer models
fuel moisture content (CHLADIL & NUNEZ, 1995; have recently identified a ratio of the near infrared (NIR)
PALTRIDGE & BARBER, 1988). and SWIR band as the most appropriate for retrieving
EWT at leaf and canopy levels (CECCATO et al., 2001,
Taking into account the previous comments, a 2002), as previous experimental studies had suggested
combination between remote sensing data (related to (HUNT & ROCK, 1989).
live fuel moisture content) and meteorological index
(related to dead fuel moisture content) will be proposed.

D-08-06.doc 34

In spite of this progress, to estimate EWT from The empirical estimation was intended for
reflectance measurements, additional efforts need to be operational retrieval of FMC in fire danger
made to derive FMC from satellite data in fire danger assessments.
studies because the amount of water per area is not as Considering the current limitations of meteorological
critical in fire propagation as the quantity of water per networks and fuel type maps, it was determined that the
dry mass. FMC estimation should not require external data sets
Assuming that the specific leaf weight (SLW= dry other than the information derived from the AVHRR
leaf weight/leaf area) is constant over time for single images and very simple vegetation maps.
species, FMC may be considered a function of EWT The estimation was targeted at grassland and shrub
(CHUVIECO et al., 2003a). species, which are the most dangerous in fire
Still, when this relation changes significantly over propagation of surface fires.
time, FMC may be indirectly estimated as a result of the
The empirical fitting was based on a long time series
effects of plant drying on the decrease in leaf area
of field measurements of FMC for the Cabañeros
index (LAI) values (mainly in shrub species) and
National Park study site (Central Spain), but collecting
chlorophyll content (herbaceous species).
field measurements at other sites with similar species
Therefore, the estimation of FMC from reflectance
validated it.
measurements can be undertaken when the estimation
is restricted to single or (physiologically) similar species. Previous work showed a strong statistical
This explains why strong empirical relations between relationship between AVHRR-derived variables and
FMC and satellite variables have been found by several FMC for the Cabañeros study site using only summer
authors (CECCATO et al., 2003; CHUVIECO et al., 2002; images (CHUVIECO et al., 2003a).
LEBLON, 2001). In this case, 2 years of field data were used for
Recent studies have shown that by better estimating calibrating the model and 2 more years for validation in
other factors affecting canopy reflectance in the NIR the same study site.
and SWIR bands, particularly the leaf area index (LAI), Additionally, strong relations were also found for
it is possible to apply radiative transfer model (RTM) Landsat-TM images (CHUVIECO et al., 2002) and SPOT-
inversion techniques to obtain a reliable estimation of Vegetation images, showing consistent trends among
EWT and FMC (ZARCO-TEJADA et al., 2003). the three sensors (CHUVIECO et al., in press).
This work follows the same trend towards finding
Additionally, plant canopy temperature is affected by
consistent relations between FMC and satellite-derived
FMC changes because water availability is a critical
variables, for operational use of satellite data in fire
parameter in plant evapotranspiration.
danger estimation.
Based on this principle, several authors have tested
In this case, the model is applied to spring and
the use of thermal images to estimate plant water
summer data, uses 4 years for calibration and 2 more
content, mainly on crops (JACKSON et al., 1981; MORAN
for validation in the same study site, as well as five
et al., 1994).
additional validation sites, located far away from the
Forest and shrub canopies are more complex, but
calibration area.
some workers have shown good relationships between
Additionally, it introduces a function of the day of the
the differences in air and surface temperature (ST) and
year to model the effect of the seasonal trends.
fire danger hazard (VIDAL et al., 1994).
Because these differences are closely dependent on Methodology
the density of vegetation cover, the combined use of
surface temperature (ST) and NDVI have shown One of the key elements to obtain a sound empirical
statistically stronger relationships with water content estimation in remote sensing research is the availability
than either of the two variables alone (ALONSO et al., of long time series of field data.
1996; CHUVIECO et al., 1999a; PROSPER-LAGET et al., Field measurements must include valid samples of
1995). large areas, on which the satellite images will be
7.1.3 Study case in Central Spain For this study, the Cabañeros National Park (located
The following study has been published in Remote in Central Spain) was used as a calibration site
Sensing of Environment (CHUVIECO et al., 2004b). because it offered unique opportunities for testing
relations between FMC and satellite-derived variables. Objectives First, the central area of the park is covered by
grassland and shrublands on very gentle slopes.
The objective of the study case is the assessment of
Because it is a protected area, no agricultural
an empirical approach to estimate FMC of
practices are carried out, and therefore, temporal
Mediterranean species based on multitemporal analysis
changes are associated to vegetation seasonal trends
of NOAA–AVHRR images.
rather than crop alterations.
The proposed method is built on statistical fitting of
Fuel types sampled were grasslands (three plots)
field collected FMC and satellite data, using a function
and several shrub species (two plots): Cistus ladanifer,
of the day of the year to take into account the seasonal
Erica australis, Phillyrea angustifolia and Rosmarinus
trends of FMC.

D-08-06.doc 35

Plot sizes were 50 x 50 m, located 3 to 5 km apart, Geometric corrections were based on orbital models
along a range of 20 x 5 km. A complete description of and manual control points and automatic correlation
the fieldwork may be found in CHUVIECO et al. (2003a). improved multitemporal matching.
Daily data were synthesized into 8-day composites
Field measurements were taken from early April to
using maximum NDVI values.
the end of September.
The median value of a 3 x 3 pixel window was
For the model calibration, a time series from 1996 to
extracted from each composite and correlated against
1999 was used.
field measurements.
These samples were collected every 8 days.
When comparing AVHRR images and field
For validation purposes, another field campaign was
measurements, the potential noise caused by the great
carried out in 2001 and 2002.
differences within the area covered may be reduced
In this case, the samples were collected every 16
when using average values of species, instead of single
days because previous analysis did not show shorter
plot averages.
time changes in FMC values.
For instance, average values of grasslands collected
In each plot, three samples per species were
in a length of 10 km (three plots separated linearly 5 km
collected. Average values per plot and period were
each) would be a better representation of what an
AVHRR pixel is actually measuring than single plot
Long-term trends of FMC values were compared for measurements.
the different grassland and shrubland plots located in
Several authors have discussed the pros and cons
Cabañeros in order to test whether they were showing
of empirical and theoretical models in remote sensing
local or regional differences in FMC trends.
research (STRAHLER et al., 1986).
Considering the large distance between plots (3–5
Theoretical models have two main advantages:
km apart), if significant differences in average values
generalizing power and a better understanding of the
between plots of grasslands or shrublands were not
parameters involved.
found, it could be concluded that FMC temporal
However, they are complex to generate because
changes of these two vegetation types are more
they require many input parameters that are often
significant than those changes caused by their spatial
unavailable and are difficult to validate.
Empirical models are commonly based on statistical
Consequently, field measurements of small plots
could be considered as representative of the temporal
They are simpler to formulate and provide a
variation of FMC for large plots.
quantitative validation on their exactness, but they are
In this way, the field measurements could be
difficult to generalize, especially when statistical
soundly related to coarse resolution satellite images.
relations are not based on physical properties.
During 2001 and 2002, a different set of field
In the field of water content estimation, a whole
measurements was taken in other regions of Central
range of theoretical models has been proposed in
Spain for validation purposes (Figure 24): a site located
recent years, most of them based on the radiative
in the province of Segovia was covered with
transfer function (BARET & FOURTY, 1997; CECCATO et
grasslands; Atazar–Alberche with grassland and shrubs
al., 2001; CECCATO et al., 2002b; JACQUEMOUD et al.,
(a mixture of C. ladanifer and R. officinalis); Ibérica and
1996; ZARCO-TEJADA et al., 2003).
Pre-pirineo with shrubs (R. officinalis and other shrub
They are solid approaches but require further
species), and Cádiz with C. ladanifer.
assessment and must demonstrate their operational
These plots are 200 to 500 km apart from the
application with field campaigns.
Cabañeros site and have different elevations but
These models estimate the EWT, which is the
include similar species, parts of the Mediterranean
variable directly associated to leaf water absorption.
ecosystem except of the Pre-pirineo site.
FMC is equal to EWT divided by SLW. EWT can be
The plots were selected so as to include
estimated using a radiative transfer function, but dry
homogeneous plant coverage on as gentle slopes as
matter content cannot be directly retrieved because the
water is masking its effect on reflectance (JACQUEMOUD
However, the shrub species were frequently mixed,
et al., 2000).
even with some trees.
For this reason, ZARCO-TEJADA et al. (2003) use a
Among the different mixtures, only those plots with a
simplified inversion model to obtain dry matter for FMC
significant coverage of C. ladanifer or R. officinalis
estimation, after deriving EWT from a radiative transfer
(more than 60%) were selected for validation purposes.
To assure consistency in the results, the field
protocol of these assessment sites was the same as Several authors (ALONSO et al., 1996; CHLADIL &
that of Cabañeros. NUNEZ, 1995; CHUVIECO et al., 1999a, 2002, 2003a;
AVHRR images were acquired by the University of
proposed empirical fittings for estimating FMC from
Alcala’s HRPT receiving station. Raw digital to
satellite data.
reflectance conversion was based on NOAA
coefficients (including degradation rates), and surface
temperature (ST) was based on methods proposed by
Coll & Caselles 1997).

D-08-06.doc 36

Most commonly, these studies were based on A scatterplot of NDVI against ST for different FMC
AVHRR images, although there are also some values observed in the Caban˜eros site showed the
examples using Landsat-TM images (CHUVIECO et al., trend towards the appearance of low values of FMC
2002). when low values of NDVI and high ST values occur,
The empirical model was based on linear regression both for grasslands and shrublands.
analysis, where FMC was the dependent variable and The trends are more evident for grasslands because
the independent variables were AVHRR variables, they present a wider range of both FMC and NDVI
NDVI and ST, and a function of the day of the year. values.
Two models were generated, one for grasslands and
Following the logic of Verstraete and Pinty 1996),
one for shrubs.
the design of an optimal index for discriminating
The major physiological differences between these
different FMC values should be based on lines
two communities made it advisable to split the fittings.
perpendicular to the main axis of NDVI and ST
Cistus ladanifer was selected as a representative of
variation, which show potential sensitivity for
Mediterranean shrub species because it is widely
discriminating FMC values.
represented in Spain.
Other species of the same family (Cistus sp.) are Additionally, a temporal variable based on the day of
also broadly distributed across the Mediterranean the year (from 1 to 365) was included in the empirical
basin. fitting to take into account seasonal trends in FMC,
following a logic already tested in Mediterranean
The statistical model was built from 88 periods (22
conditions (CASTRO et al., 2003; CHUVIECO et al., in
periods of 8 days during 4 years: 1996 to 1999),
covering spring and summer conditions of the
Considering that these temporal trends are more
Cabañeros National Park.
contrasted for grasslands than shrublands, two different
The time series include a wide range of rain
functions were computed:
patterns, with some dry years 1999 and 1997 with
precipitation close to 200 mm in 6 months), and more FDg = (sin(1.5 x π x (Dy + Dy1/3)/365))4) x 1.3
humid ones 1996 and 1998), with 250 and 230 mm,
respectively. FDc = (sin(1.5 x π x Dy/365))2+ 1) x 0.35

The equation was validated using data from the where FDg and FDc are the functions of the day of
same study area (Cabañeros), as well as the other the year (Dy) for grasslands and C. ladanifer,
study sites previously described during the 2001 and respectively, and the sine angle is computed in radians.
2002 spring and summer seasons. This function was derived by fitting a periodical
function to the temporal average of FMC values of
Satellite variables considered in the linear grassland and C. ladanifer for 6 years of measurements
regression were NDVI and ST. in Cabañeros 1996–2001).
The former would be positively related to FMC The function has a wider variation for grasslands
because the drying of the plant reduces chlorophyll than C. ladanifer, which agrees with the stronger
activity in grasslands, as well as leaf area index in contrast in the water content of herbaceous species.
shrub species. The constant terms were used just to scale the
On the contrary, ST would be expected to be functions in a similar range among them.
negatively related to FMC because the cooling effect of
evapotranspiration is reduced when plants get dry and Results
introduce mechanisms to reduce water loss.
For the 6 years of field data, the t tests applied to the
Before obtaining the estimations of FMC from linear temporal differences of the three grassland plots in the
regression models, an analysis of trends between FMC Cabañeros site, separated between 3 and 5 km, did not
and these two variables (NDVI and ST) was show significant differences among them.
undertaken. Similarly, the average temporal trends of the two
In a similar way to other study areas (ALONSO et al., shrub plots located 3 km apart were not significantly
1996; KALLURI et al., 1998; MORAN et al., 1994; different.
PROSPER-LAGET et al., 1995), NDVI showed a negative Therefore, it could be concluded that the average
correlation with ST in both grasslands and shrublands FMC values of small plots (50 x 50 m) are
for the spring and summer seasons. representative of large areas (several kilometres apart),
This trend must be related to the physiological at least in the Cabañeros site, and consequently,
reaction of plants to higher temperatures and lower temporal data extracted from those small plots can be
moisture contents, which, depending on the plants, may assumed representative of the large plots observable in
change leaf colour, deteriorate leaf structure, modify AVHRR coarse-pixel size images.
leaf angle distribution by leaf curling or reduce LAI by
As mentioned above, the equations to estimate FMC
leaf loss, and/or decrease evapotranspiration.
from AVHRR data were derived from multiple linear
Based on these relationships, some authors have
regression analysis, using NDVI, ST and FD for 4 years
proposed a regression model of NDVI and ST to
of Cabañeros field data 1996 to 1999).
estimate plant evapotranspiration (KALLURI et al., 1998)
and fire hazard levels (PROSPER-LAGET et al., 1995).

D-08-06.doc 37

This severe decrease is reflected in the reduction of

The resultant equations were:
NDVI and increasing ST, but not as steep as the field
FMCg = -57.103 + 284.808 x NDVI – 0.089 x ST + FMC values show.
136.75 x FDg As a result, overestimations in this period reach up
to 60% of FMC.
FMCc = 70.195 + 53.520 x NDVI – 1.435 x ST + Additionally, negative FMC values were estimated in
122.087 x FDc August of the second year, caused by very low NDVI
values (below 0.1 for this period).
- FMCg and FMCc are the estimated FMC values of However, negative estimations are not a major
grasslands and C. ladanifer, respectively; obstacle for operational purposes because a simple
- NDVI is the normalized difference vegetation index
filter could be applied to the empirical model to avoid
(range -1 to + 1); them.
- ST is the surface temperature (in Celsius degrees); Additionally, FMC values of grasslands during most
- FD, the function of the Day of the Year. of July and August are below 30%, which may be
The determination coefficients (r2) obtained were considered the limit for live species.
0.737 (p < 0.001) for grasslands and 0.672 (p < 0.001) Therefore, for practical purposes, grasslands may
for C. ladanifer. be considered as dead fuels for the central part of the
Significance values of independent variables were summer.
lower than 0.01 for NDVI and FDg in the case of
The validation of the C. ladanifer showed an even
grasslands, and ST and FDc in the case of C. ladanifer. better fitting than grasslands, with very close
ST was not significant for grasslands (p>0.5) estimations both in spring and summer in the
because most of its discrimination power is included in
Cabañeros site.
the FDg variable. The highest deviations from the observed FMC
However, it was incorporated to improve spatial values never reached 20%, and for most periods, they
estimations, given that FDg does not change spatially.
are under 10% of FMC.
For the same reason, NDVI was kept in the case of The other two validation sites for grasslands (Ávila–
C. ladanifer, in spite of its low significance (p = 0.187). Segovia and Alberche–Atazar) also showed very good
The contribution of NDVI to FMC estimation is more
fittings, with r2 values of 0.881 and 0.905 (Figure 25).
significant for grasslands because the drying process in There is a slight tendency towards overestimation in
herbaceous species is commonly followed by a loss of Alberche–Atazar and underestimation in Ávila–Segovia,
chlorophyll activity and a LAI decrease.
but the relation in both cases is close to a 1:1.
In both cases, the factor accounting for the seasonal
trends (FD) is very significant (p < 0.001). The scattergram also shows a nonlinear estimation
As expected, NDVI and FMC show a positive trend, especially in Alberche–Atazar, which may be
correlation, whereas for ST it is negative, confirming the related to the saturation of NDVI in the upper part of the
physiological assumptions previously stated. range (BARET & GUYOT, 1991).
In fact, polynomial equations between observed and
The assessment of these equations was carried out estimated FMC values provide r2 values higher than for
on other time periods (2001–2002) in the same study lineal trends, with 0.96 for Ávila–Segovia and 0.95 for
site (Cabañeros), as well as on other study sites, and
good results were obtained in all cases. Nonlinear relationships should also be explored at
Relations are very coherent at all sites, in spite of the calibration stage in the future.
being at a distance of over 200 km for Segovia and
Atazar–Alberche sites, and more than 500 km for the FMC estimation for C. ladanifer shows good fittings
Ibérica site, and with different altitude ranges (up to 500 in all assessment sites: Alberche–Atazar, Cabañeros
m of height increase in the case of Segovia). and Cádiz, although for the latter, only 4 observations in
For the Cádiz site, only four observations were the summer of 2001 were available (Figure 26).
available, but r2 values were also very high (0.96). The empirical model has a slight tendency towards
The Pre-pirineo site had no samples or neither overestimation, especially for lower values of FMC.
grasslands nor C. ladanifer and was not included in this The lower values offer a better fitting between
analysis. estimated and observed FMC with differences lower
than 10% of FMC during mid-summer.
The temporal trends are very well estimated, and
deviations of actual versus predicted FMC values are Considering certain physiological similarities
low and have no consistent bias. between C. ladanifer and another widespread
The worst estimation was observed for grasslands in Mediterranean shrub, R. officinalis, the empirical
late spring (early June) and in the middle of the summer function was also applied to other study sites where this
(August). shrub species had been sampled in the field.
The former is related to the sudden decrease in The results were very positive for all sites
FMC, which in both years, changes from over 170% to (Cabañeros, Atazar–Alberche and Ibérica), with r2 over
just 30–35% in 16 days. 0.85.
The temporal trends also show good fittings, with
nonsignificant biases (Figure 27).

D-08-06.doc 38

An underestimation was observed for the spring The ME of living fuels varies between 12% for some
season, but the fittings improved in the summer, when grass fuels and 200% for the needles of some conifer
fire danger is higher, and therefore, the need for species.
accurate estimations is more demanding. For most live vegetative fuels the ME is in the range
The absolute errors were higher for R. officinalis of 120%-160%, while for dead fuels it is in the range of
than for C. ladanifer, especially in the Ibérica site, which 12%-40%.
may also be caused by the mixture of the field plots,
This method assumes that ME values act as relative
where several shrub species grow in the same area.
thresholds to ignition for each fuel, above which the IP
7.1.4 Probability of ignition related to live fuel dramatically decreases.
moisture content Although, the IPf for FMC values higher than ME
would be zero, a conservative approach is
The empirical model generated in the previous
recommended here, assuming that a marginal IPf exists
section using NDVI, ST and function of Day of the Year
even at high values of FMC.
showed a consistent predictive power to estimate FMC
For this reason, it is proposed to assign a maximum
of grasslands and C. ladanifer, a typical Mediterranean
IPf value of 0.2 to the FMC that equals the ME value of
shrub species.
each fuel.
The model was tested in plots located several
FMC values lower than ME would have IPf values in
hundreds of kilometers apart and with different altitude
the range of 0.2–1, the IPf being linearly inversely
proportional to FMC values.
Therefore, this model may be tested on operational
For FMC values greater than the ME, IPf values
scenarios in Mediterranean conditions, applying both
would range from 0.2 to 0.
spring and summer data.
Null IP (IPf = 0) was assigned to the maximum FMC
The model only requires two basic satellite variables
value recorded in the historical series of FMC field
(NDVI and ST), the day of the year and a regional map
of vegetation types, which distinguishes grasslands
from shrublands. Schematically this method is based on the following
This variable could also be derived from the algorithm (Figure 29):
multitemporal classification of AVHRR data, following
If FMC > ME, then
any of the methods applied to derive global land cover
maps (DEFRIES & TOWNSHEND, 1994). IPf = {1– [(FMC – ME)/(FMCmax – ME)]} × 0.2
Considering the spatial and temporal resolution of else
AVHRR images, the empirical index may be used for
short-term estimations of fuel moisture content (Figure IPf = 0.2 + [(ME – FMC)/(ME – FMCmin)] × 0.8
28). where FMCmax and FMCmin are the maximum and
After estimating the FMC of live fuels through minimum FMC values of each fuel type derived from
satellite information it is essential to transform these field FMC samplings.
values to probability of ignition in order to be able to While these samplings are site specific, they might
assess the probability of ignition in live fuels and to nevertheless be assumed to apply to fairly large regions
combine this probability with probability of ignition with similar environmental conditions.
based on other factors (i.e. human factors etc). ME values are fuel-type specific.
To do so, CHUVIECO et al. (2004a) developed a For dead fuels, ME values were taken from the
simple method to convert FMC values to danger ratings BEHAVE fire behaviour prediction system (Burgan and
Rothermel 1984), ranging from 12% to 40%.
based on computing ignition potential from thresholds of
moisture of extinction adapted to each fuel. Grasslands with FMC values lower than 30% were
Fire danger is restricted to the likelihood of fire treated as dead fuels and assigned ME values of 12%
(model 1) and 15% (model 2) following values proposed
occurrence, given a particular fuel moisture content.
This likelihood will be defined in terms of probability for the BEHAVE model (BURGAN and ROTHERMEL
of ignition associated with fuel moisture content status. 1984).
These values can then be integrated with other For live fuels, threshold values were taken from
variables associated with ignition sources (e.g., specialized literature.
lightning, human), which could also be expressed in For shrubs, an average value of ME was selected
terms of ignition probability. (105%), following the experimental results of
The method proposed from Chuvieco et al. (2004a), DIMITRAKOPOULOS and PAPAIOANNOU (2001) and
to convert FMC values to ignition potential associated BURGAN (1979).
with fuel status (IPf ranging from 0 to 1) is based on the For annual grasslands, the ME threshold was fixed
concept of moisture of extinction (ME). at 40% (ALBINI 1976).
According to ANDERSON (1982), the ME of
This is defined as the threshold moisture content
above which a fire cannot be sustained (ROTHERMEL, grasslands is directly related to fuel depth.
1972). This explains why short grasses have low values
(12%–15%) and tall grasses (more than 70 cm high)
have higher values (25%–30%).

D-08-06.doc 39

The use of linear functions to convert FMC to IPf is The water content of dead fuel is constantly
based on the experience of several authors (such as changing, depending mainly on atmospheric conditions
shown similar trends between FMC and ignition delay Loss or gain of water content will vary depending on
for a wide range of Mediterranean species. the physical and chemical characteristics of the fuel and
The slope coefficient of the linear regressions was the presence of varying atmospheric activity (rain,
related in this case with the flammability of the species. condensation, etc.).
DELICHATSIOS et al. (1991) proposed a similar
The water content of dead fuels is determined by
solution, they termed this regression coefficient
various methods.
“flammability delay increment.”
The most precise one is direct sampling by
The use of a linear function instead of an
gravimetric methods.
exponential one (as suggested by other authors, such
Following this approach, water content is computed
as TRABAUD 1976 and ALBINI and REINHARDT 1995) is
from the difference of wet and dry weights of the
justified by the fact that recent experimental results with
numerous species of the Mediterranean Basin clearly
Most commonly, the water content is expressed as a
show that there is a gradual, linear relation between
percentage of the dry weight (BLACKMAR and FLANNER
flammability (as measured by ignition delay time) and
1968; DESBOIS et al. 1997).
Direct sampling provides exact measurements, but it
is costly and labour intensive, especially when wide
Older studies used a limited number of species and
area estimations are required.
not the ISO 5657-1986(E) methodology for measuring
Additionally, this method does not provide an
ignition time of materials (TRABAUD 1976; VALETTE
instantaneous measurement, as the samples must be
oven dried during a certain number of hours (24 or 48
In a highly exponential (nonlinear) relationship, small
have been commonly suggested).
variations of the FMC in the upper scale of the curve
(i.e., near the threshold of the moisture of extinction) Other methods are based on the use of previously
may result in relatively big differences in the calibrated wooden sticks that are assumed to be good
flammability of the fuels. representatives of certain types of fuel (SIMARD 1968).
However, Mediterranean species seem to respond In the USA, standard 10-hour fuel sticks with an
with gradual fluctuations in their ignition delay time to oven-dry weight of 100 grams are commonly used.
changes in the FMC status (DIMITRAKOPOULOS and The sticks are continuously weighed.
This is especially noticeable in live fuels and highly Finally, meteorological danger indices (MDI) have a
flammable species (i.e., species with relatively low long tradition in fire danger estimation, because they
ignition delay time). comprise different critical variables related to fire
ignition and fire propagation.
Efforts are underway to improve the highly
exponential relationship between the moisture and fuel The meteorological danger indices vary in
flammability in the Canadian Forest Fire Danger Rating complexity and in the number of variables to be
considered, from those that only require temperature
System (MCALPINE, 1995; in BEMMERZOUK, A.M., 1997)
and relative humidity to those which are based on
7.1.5 Probability of Ignition in Dead Fuels complicated numerical models (for a summary of the
(Meteorological Index) models most used see VINEY 1991).
These indices rely on current and past weather Meteorological Danger Indices
conditions, since they also try to estimate the degree of
Dead fuel includes a wide range of materials dryness of different forest fuels.
(senescent grasses, dry leaves, small twigs, and Since most of the countries have a relative dense
organic material in the topsoil). network of weather stations for different purposes
Water content of dead fuels is the most important (agriculture yield prediction, disaster prevention, traffic
factor in determining fire danger potential. regulation, etc.), MDI values can be operationally
On one hand, the water content of the fuel is computed for extended territories.
inversely related to the probability of ignition, due to the Additionally, with the growing availability of
fact that part of the energy necessary to start a fire is automatic weather stations, these indices may be
used up in the process of evaporation right before the computed very frequently and measured in real time.
fire starts to burn (CHANDLER et al. 1983). However, often the location of weather stations is
On the other hand, water content also affects fire not very appropriate for fire danger estimation, since
propagation since the source of the flames is reduced they are located in urban or agricultural areas.
with humid materials, therefore reducing flammability Therefore, spatial interpolation techniques to
(ROTHERMEL 1972). estimate weather variables at forested areas are
For this reason, the estimation of water content in required.
these fuels is a critical variable both in determining fire These interpolation methods always introduce a
ignition in a specific time and place, as well as in certain estimation error, which is added to the actual
predicting the behaviour of the fire itself. estimation of fuel moisture content (FMC).

D-08-06.doc 40

Even so, this method is the most widely used to As noted before, one of the main challenges of using
determine the FMC of dead fuel. these meteorological fire danger indexes is extending
the measurement obtained in the specific weather
There are a lot of fire danger meteorological indices
stations to cover the rest of the territory (FUJIOKA 1987).
which use in different ways air or/and ground
This problem is compounded more and more
temperature, air or/and wood sticks humidity (relative or
because the Forest Services use the GIS technology in
absolute), wind speed, and soil water reserve as an
forest fire prevention when they integrate spatial
expression of the possibility for the plant to get a water
variations in the data.
satisfaction rate.
As a result, once an index with a better predictive
From all these meteorological indexes found in fire
capacity has been selected, it is necessary to get a
prevention literature, the Fine Fuel Moisture Content
spatial distribution of this index.
(FFMC) of the Canadian CFFWI (Canadian Forest Fire
Weather Index) and the 10-h moisture code of the The spatial dimension of meteorological indices can
National Forest Fire Danger Rating System (NFDRS) be obtained using a mapping methodology based on
provide a sound estimation approach for short-term logical relationship between the value of an index (or of
changes in dead fuels FMC. its components) and the environmental parameter
Both the CFFWI and the NFDRS have several values, as it has been explained in a previous
components and their usefulness had been validated deliverable (D-08-05).
for fire prevention in several Mediterranean contexts
This methodology improves the spatial information:
from a map built with some points (the index value of
The FFMC estimates the water content of the top which is indicated) and many empty cells, we get a map
layer on the ground (L-layer) from measurements of with all the cells filled with the index value for each cell.
temperature, relative humidity, wind velocity, and Next section shows an example of this methodology
precipitation registered in the last 24 hours (VAN at local scale.
WAGNER 1987). Spatial interpolation techniques: the case of
This water content index has an empirical character
Alpes-Maritimes (France)
derived from the relationship between these
meteorological variables and the water content of a Mapping the meteorological indices values is in fact
standard fuel (jack pine and lodgepole pine type), at the one of the most important steps of the risk
same time integrating the accumulative effect of determination.
atmospheric conditions in the hours previous to the As explained in previous deliverable D-08-03,
measurements. meteorological indices are always computed from
This index has a timelag of approximately 0.66 days. stations network data.
On the other hand, the index 10-h, which is part of Every data is representative only of the sensor that
the NFDRS, estimates the water content of measured it, and then there are two concepts
combustibles with a width of 1.2 to 2.5 cm (BRADSHAW concerning the use of meteorological data measured in
et al. 1983). a given place, to elaborate a meteorological risk index:
This index uses the concept of Equilibrium Moisture - Either a region is covered for instance by a 20
Content (EMC). stations network, and the index is computed for each
The EMC of a fuel element under given station. That means the risk is known only for 20
environmental conditions is the moisture content that points, and is unknown for all the places, which exist
the element will attain if left for sufficient time in those outside these points. The users, as foresters or
(constant) conditions. firemen have to fill themselves, with their
The EMC is a function of the temperature and the imagination, the numerous places where the value is
relative humidity (SIMARD 1968), as well as the not computed, and in fact they interpolate in their
atmospheric conditions present at the moment of mind the absent values. This risk concept is a
measuring the samples. punctual one and the problem is the needed
A study undertaken in the Department of Geography information is absent on large surfaces.
at University of Alcalá has compared the usefulness of
two meteorological danger indexes in fire prevention
used in Canada (FFMC) and the United States (10-h) in
order to evaluate the FMC of dead fuels in a
Mediterranean territory.
Taking measurements in the field over a period of
6 years compared the codes.
The results did not show a significant difference
between the two; therefore it is recommended that the
index with less meteorological variables (10-h) be used.
This index allows an estimation of FMC of
flammable dead fuels with an RMS around 5% (AGUADO
et al, in preparation).

D-08-06.doc 41

- Or, in the same region covered by 20 stations, it is - This index has been used by French firemen and
considered that there are sub-regions constituted by foresters of the Alpes-Maritimes during more than 10
different areas, which are relatively homogeneous. years 1990-2000) inside of an Expert System
For example, in France, MeteoFrance decided they elaborated by Ecole Nationale Supérieure des Mines
are 7 different « fire regions » in the department of de Paris, and has proved to be very efficient,
Alpes-Maritimes (Nice). For every « fire region » according to the users.
(their surfaces are different from the others), the The width of each pixel: spatial resolution.
meteorological risk index is computed only from the
stations, which are inside the fire region, generally 2 The smaller a pixel is, the more accurate the risk
or 3 stations. In fact, each fire region is «virtual» definition will be, and that may be very interesting,
because each meteorological variable which is used especially in complex topography.
and contribute to compute the index of each fire But such a precision needs computer resources.
region is an average value computed from the
stations. Thus the index is obtained from data Working on a large area as Mediterranean Europe
characterizing a non-existent place which results for instance, at a given moment, does not necessitate
having a very high spatial resolution.
from computed values (means, barycenter from 2 or
3 stations). But, a very high resolution is needed for operational
work, i.e. by foresters or firemen.
In this case we can say that each fire region (from In this case, the spatial scale used has to be very
200 to 1500 km2) index is supposed to represent an accurate.
area which is homogeneous enough, so that in details, In this work we used a DTM of Alpes-Maritimes
the risk is he same in each point of the region. (4400 km2) with 50 m spatial resolution, (degraded to
Obviously, especially in mountainous relief, as very 1 km2 if needed).
often in Mediterranean regions, this homogeneity does The map of figure 30 gives an idea of the
not exist… topography of the region where this methodology was
checked (department of Alpes-Maritimes).
We propose here another possibility, that can be
used everywhere, if a stations network exists: it consists In fact, we decided to work on the south part of this
to obtain with scientific arguments the most probable region because excepted one, all the available
meteorological risk index value for a very small area, a automatic stations are located in this south part,
pixel, the surface of which is for instance from 2500 m2 allowing thus to obtain a higher measurement network
(50 m side) to 1 km2 (instead of some hundred or density for meteorological data.
thousands kilometres…).
As explained in previous deliverables, different
The principle is to interpolate data from a station to environmental parameters can be obtained from the
elaborate the risk value of each pixel. DTM, as:
- Altitude of each pixel;
In the previous deliverables D-08-03 and D-08-05,
- Relative altitude, indicating if a given pixel is at a
we explained the « environmental » regression
lower position (bottom of a valley) or dominating;
methodology: a statistical law is found, establishing a
- Aspect, with a value varying from 1° (North) to 180°
link between the meteorological data of a place (a pixel)
(South) that is the maximum, East and West being
and the environment of this place, especially the
the same;
topographic and relief environment (altitude, relative
- Slope, above the given pixel, in degrees;
altitude, aspect, slope, sea distance).
- Sea distance.
Realizing index risk maps induce to answer to 2
One example of maps is given that allow comparing
sorts of problems and to make choices:
two parameters as aspect, for 50m pixel (fig. 31) and
- To decide the width of each pixel, in a raster GIS;
1km pixel (fig. 32).
- To decide which methodology will be used:
interpolation of data from the network stations and Two different logics: direct interpolation of
computing the index for each pixel, or computing the the index, or interpolation of meteorological
index for each station (from its data) and data in order to compute the index.
interpolation of the index for each pixel.
Direct interpolation of the index, 50 m definition
The meteorological index we have used in this work
is I85/90 CARREGA index for 3 reasons: It is in fact the simpler way to obtain a
meteorological risk index map.
- We know it very well as one of us is the author!
- This index is simple because composed with only The methodology needs a station network and an
3 parameters (wind, air humidity, soil water reserve), environmental regression model.
that allows to understand more easily how maps are We decided to use always the same model, even if
changing from a situation to another, and how according to the date and the corresponding
meteorological situation, the score of the regression
spatial interpolation works;
analysis is not the same.
The environmental parameters used are: altitude
(m), relative altitude (m), aspect (°), slope (°), distance
from the sea (km).

D-08-06.doc 42

Maps of figure 33 and figure 34, at 06h and 12h Thus the interpolation can be made by complex
UTC are risk index interpolation maps using a DTM 50 models or by spatial interpolation as kriging or simpler,
m resolution for each pixel. as IDW method (weight of each point is related to the
The risk index value of each 50 m pixel was distance between other points).
interpolated directly by environmental regressions, from
Direct interpolation of the index, 1 km definition
the risk index value computed for each meteorological
station. The methodology is the same as section i), but the
geographical scale changes: each pixel is 1 km instead
On 28 August at 6:00, the influence of altitude is
of 50m.
obvious at the end of the night: risk is higher on slopes
Comparing both maps of fig. 33-34 and fig. 37-38
facing to South but especially on higher altitudes.
shows the differences concerning the spatial accuracy,
High risk values (red) are not numerous, and the
which is less operational and usable when pixel width is
dominant situation is a low risk value, because we are
just at the beginning of the day, with generally low wind
But the advantage of this 1 km dimension is that less
speed, and high air humidity.
computer capacity and memory storage is used.
The high value of the index in altitude is due to wind
speed. Maps of figure 37 and figure 38, at 06h and 12h
At 12:00, the difference is obvious, due to UTC are risk index interpolation maps using a DTM 1
decreasing air humidity (temperature has increased) km resolution for each pixel.
and increasing wind speed (due to air heating and The risk index value of each 1 km pixel was
turbulence). interpolated directly by environmental regressions, from
the risk index value computed for each meteorological
Interpolation of the meteorological variables which
constitute the risk index value, 50 m definition
Figures 39 to 43 are 1 km definition risk index maps
In this case the interpolated values are not the risk
(index interpolated) for August 31st 2003.
index initially computed for each station.
The sequence of maps allows to evaluate the the
The interpolation concerns the meteorological
risk level which is changing very quickly in time, and
which is different at the same moment from a place to
These ones are interpolated from the station
another one.
network for each pixel of the map, taking into account
During this day a Foehn wind effect occurred at the
its environmental parameter.
end of the morning, with high temperature, very low air
Obtaining its meteorological values allows
humidity and increased (but moderate) wind speed, with
computing for each pixel its risk index value.
a soil water reserve, which was very low.
The environmental parameters used to interpolate A “mixed fire” (forest, crops and houses) started
meteorological data are also: altitude (m), relative about 12:00 UTC very close to the sea, burnt more than
altitude (m), aspect (°), slope (°), distance from the sea 200 ha despite of about 800 firemen and destructed 10
(km). houses.
This fire was studied and explained in D-08-02.
Maps of figure 35 and figure 36, at 06h and 12h
It is interesting to notice that the meteorological risk
UTC are risk index maps resulting from meteorological
index was very high in the place where the fire began
variables interpolation (using a DTM 50 m resolution for
(in dark at 12:00).
each pixel).
The risk index value of each 50 m pixel was Conclusion
computed directly from its meteorological values.
The possibility to obtain a map with a spatial
These values were interpolated by environmental
repartition of risk indices, and especially of a
regressions, from meteorological values of each
meteorological index, seems to be very interesting,
meteorological station.
compared with the simple knowledge of some points
Comparing the two maps (interpolation of the risk only, or to large areas in which the risk value is
index or interpolation of meteorological components of supposed to be the same.
the index, and then computing the index) allows
The methodology used influences the results as the
noticing differences of behavior: mapping directly the
comparison between the two sorts of maps shows.
index tends to smooth the local differences and
But the most satisfactory way to obtain a map
prevents to know what is the meteorological variable
(interpolation of meteorological variables) from an
which dominates the others.
intellectual point of view is also the longer.
On the opposite, the risk map resulting from
Thus, it is easier to interpolate directly the value of
interpolation of meteorological variables shows easily
the risk index.
some “spots” the risk of which is obviously different
In fact the validity of the result depends upon the
(blue spots of fig 35 and 36).
performance of the coefficient of correlation.
They are due to the interpolation method of wind
This one is generally better for water soil reserve
speed which is not an environmental one because the
than for humidity, and better for 50 m definition than for
logic of spatial variation of wind does not obey
1 km.
principally to topographic conditions.

D-08-06.doc 43

7.1.6 Figures



4 2
3 Grasslands
Valencia Shrublands

Name (Elevation)
1 Cabañeros (700m)
2 Atazar (900m)
3 Alberche (800m)
4 Segovia (1200m)
5 Ibérica (900m)
7 6 Pirineo (1000m)
7 Cádiz (400m)

Figure 24: Location of the sampling plots

Grasslands Alberche-Ataz ar y = 0.841x + 6.908

R = 0.8815


E stimated





0 50 100 150 200 250 300 350

y = 0.9525x - 3.8232
Grasslands Ávila-S egovia R = 0.9053



Estimate d





0 50 100 150 200 250 300 350
Ob se rv e d

Figure 25: Observed and estimated FMC values for grasslands in Alberche-Atazar and Ávila-Segovia
Validation sites (2001 and 200)

D-08-06.doc 44

y = 0.7376x + 29.912
Cistus all sites
R2 = 0.7241


0 20 40 60 80 100 120 140 160 180

Cabañeros Cádiz Alberche-Atazar

Figure 26: Observed and estimated FMC values for C. ladanifer in Alberche-Atazar, Cabañeros and Cádiz.
Validation sites (2001 and 2002)

Rosemary Cabañeros (2001-2002)


3-4 29-5 16-7 2-9 9-5 27-6 13-8 30-9

Observed Estimated

Figure 27: Observed and estimated FMC values for R. officinalis in Cabañeros.Validation sites (2001 and 2002)

Figure 28: Live fuel moisture content map from Euro-Mediterranean countries (August, 10th; 2004).
Based on CHUVIECO et al., 2004a, processed by A. CAMIA, JRC

D-08-06.doc 45

150 0.2
ME 0.4
100 0.6
50 1

Figure 29. Scheme to convert fuel moisture content (FMC) to ignition potential (IP) (example of dead fuels)

Figure 30: The topography of the Department of Alpes-Maritimes: altitude.DTM with Pixel width: 50 m.

D-08-06.doc 46

Figures 31 and 32: Aspect of each 50 m pixel (left) and 1 km (right)

Figures 33 and 34: High definition (50m) Meteorological risk index map on 28 August 2003 at 06 and 12 UTC.

Figures 35 and 36: High definition (50m) Meteorological risk index map on 28 August 2003 at 06 and 12 UTC.

D-08-06.doc 47

Figures 37 and 38: Low definition (1 km) Meteorological risk index map on 28 August 2003 at 06 and 12 UTC.

Figures 39 and 40: 1km definition risk index maps for 31st August 2003 at 03:00 and 06:00 UTC

Figures 41, 42 and 43: 1km definition risk index maps for 31st August 2003 at 09:00, 12:00 and 15:00UTC

D-08-06.doc 48

7.2 PROBABILITY OF IGNITION (HUMAN In fact, the temporal and spatial data required to
FACTORS) evaluate the human factors in fire danger rating simply
do not exist (MARTELL, OTUKOL, and STOCKS 1987),
7.2.1 Introduction
while other variables of fire danger such as
Human activity is the main factor of forest fires temperature or relative humidity are routinely
ignition worldwide, having a special impact on generated.
Mediterranean countries of Europe. Moreover, human activities are very dynamic on
From different statistical sources we know that most time and space, which difficult the estimation of specific
of the wildfires in Mediterranean Europe (above 90%) spatial patterns, or rather these are more difficult to
occur as a consequence of human activities that can determine, as in the case of pyromania or specific
directly act as fire ignition sources or indirectly create deliberate motivations and attitudes.
the conditions that favours fire ignition and/or fire However, for other causes, such as those relating to
propagation. recreational activities or agricultural burnings are easier
to model using spatial variables.
For example, according to data published by the
Spanish Forest Service (DGCN) for the period 1988- The complexity of dealing with the human factors of
1999, 96% of the fires in Spain were produced, either fire danger have frequently led to investigators and fire
directly or indirectly, by human causes, which indicate managers to either leave them out of their risk models,
the close link between fires and human activities. or consider them only marginally.
The only natural cause of fires is lightning, which in Most commonly, when those factors are considered,
Spain accounts for a small percentage of only 4% the estimations are based on indirect assessments of
according to statistics of the past 15 years. human risk activity, namely indicators of activities which
This is probably due to the fact that storms are not are the usual cause of fire and which are generally
as important in Spanish forested areas as in other parts structural in nature, i.e. population density or urban-
of the planet. wildland interface.
In fact, this may well be the only meteorological The first studies on human factors of fire danger
phenomena in this climatic area that does not favour were based on indirect variables, obtained mainly from
fires ignition. censuses and survey sources (ALTOBELLIS 1983; BAIRD
Curiously enough, this distribution of fires according FOLKMAN 1971; COLE and KAUFMAN 1963; DOOLITTLE
to its cause is quite different to other typical 1979; DOOLITTLE and WELCH 1974; FOLKMAN 1965,
Mediterranean ecosystems such as the Californian
1973; JONES, TAYLOR, and BERTRAND 1965).
one, where lightning may cause up to 40% of fires. Later on, in the eighties the human factor is
There are other non Mediterranean parts of the analysed spatially, and it is frequently based on
planet where lightning is of high importance such as in cartographic aspects, that considered some human
Scandinavia and Russia (20 to 30%), Western North variables along with natural features, such as slope or
America (40 to 60%) and in the Australian region of
fuel types (Aerial Information Systems Inc. 1981;
Victoria (10 to 30%). CHUVIECO and CONGALTON 1989; DONOGHUE and MAIN
Consequently, it is clear the importance of including 1985; DONOGHUE, SIMARD, and MAIN 1987; LYNHAM and
human factors in any comprehensive fire danger index. MARTELL 1985; PHILLIPS and NICKEY 1978; MARTELL,
Nevertheless most current operational indices of fire BEVILACQUA, and STOCKS 1989; YOOL et al. 1985).
danger rely on physical variables (mainly weather The number of such studies has increased in the
data), and human factors are not ordinarily considered
last years due to the greater availability of digital
(Chuvieco et al., 2003b). cartographic and statistical information managed
Modelling human factors associated to fire ignition is through Geographical Information Systems (GIS)
very difficult, since economic and recreational activities (ABHINEET et al. 1996; CHOU 1992; CHOU, MINNICH, and
linked to fire are very disperse in time and space. CHASE 1993; CARDILLE, VENTURA, and TURNER 2001;
An alternative to model human factors of fire danger CHUVIECO and Salas 1996; CHUVIECO et al. 1999b;
is the estimation based on indirect indicators that relate LEONE et al. 2003; SALAS and CHUVIECO 1994;
fire occurrence with spatial variables linked to human THOMPSON 2000; VASCONCELOS et al. 2001; VEGA-
activities. GARCÍA et al. 1995; VEGA GARCÍA et al. 1996).
These recent studies commonly use variables
Human ignition danger may be defined as the related to recreational activities in forested areas,
probability of a fire occurring as a result of the presence proximity to roads and trails, population density,
and activity, either directly or indirectly, of human distance to human settlements, forest property types,
beings. etc.
The evaluation of human activity as an agent of fire These variables are easier to spatialize, but they
ignition is a complex task, since data on human may have marginal importance in some areas (VEGA-
activities on forested areas are rarely available (VEGA GARCIA et al., 1995).
GARCÍA et al. 1995).
Despite the importance of these empirical works,
there are still many aspects of the fire-human
relationships that require further research.

D-08-06.doc 49

For instance, we do not have yet a clear This included information on the location and likely
understanding on the spatial consistency of those cause of fire ignition, as well as surface area burned
relationships between fire and human factors. (not used in this study).
Some variables may be critical in some regions, The “Prométhée” database provides less
while marginal in others, depending on the socio- comprehensive information for fires going back to 1973.
economic and environmental structure of the territory. Fire ignition data were complemented by a Digital
Additionally, the human factors require a Elevation Model (50 m resolution), digitized road
comprehensive assessment, which includes not only network from the “BD Carto” data base of the IGN
spatial-variables (distances, fragmentation, interfaces) (National Geographic Institute), and vegetation and
but also socio-economic ones (unemployment, rural land-use maps at a 1:25,000 scale from the Corinne
population age, population density…). Land Cover data base.
Recent conferences in forest fire research have Elevation, vegetation, land-use, and road networks
emphasised the importance of this holistic approach maps for the Alpes-Maritimes are shown in Figures 44
(Viegas 1998, 2002). to 47.
In the following sections two examples of human In summary, the Alpes-Maritimes department is
ignition danger assessment are showed; one case at particularly mountainous in the North of the department
regional scale and another one at European Scale. and hilly in the South along the coast.
We are aware that both aproaches need to be The Mediterranean climate has a pronounced dry
improved and adapted in order to integrate them period in the summer, which is accompanied by dry
properly in a euro-mediterranean risk index using the periods in the mountains during the winter.
previously specified resolution of 1 km2. Within a 10-15 km swath along the coast, the
Finally a fire occurrence map is presented as a vegetation is typically Mediterranean, especially West
possible alternative of Nice.
The population of about 1 million inhabitants is
7.2.2 Estimation of human ignition danger at
concentrated in the coastal area with a succession of
regional scale: the case of Alpes-Maritimes
large cities (Cannes, Antibes, Nice, Monte-Carlo...).
The area of Mediterranean vegetation is therefore
Natural ignition due to lightning in the Alpes- concentrated in the suburban fringe just North of the
Maritimes (SE France) is relatively rare since most urban coastal area where houses are mainly individual
storms are usually accompanied by rain due to the villas surrounded by forests
proximity of the sea which induces conditions of high
Forest fire ignition risk can be considered high in
atmospheric vapour.
this sector due to abundant dry vegetation in the
Approximately 90% of fire ignitions between 1991 summer and a relatively dense road network and a
and 2003 in the Alpes-Maritimes were of human origin, concentration of human activities.
and this figure is probably an underestimate of long
Forest fire ignition: During the 1991-2003 period, a
term averages since the data include the fires of 2003
total of 550 fires were recorded in the Alpes-Maritimes.
which was a particularly hot and dry summer with more
A precise ignition location is available for only 362
“dry” storms than usual during the month of August.
(about 66%) of these, and a known can be defined for
It is therefore reasonable to attempt to estimate the
only 225 fires, or 50% of the total number of fires
probability of fire ignition as a function of the location
(Figure 48).
with respect to human settlements since it is in
The spatial distribution of fire ignitions in the
proximity to human activities that most fires find their
department according to known cause is shown in
Figure 49, identifying 8 categories of fire ignition. Description of study area and database The remaining fires were classed as cause unknown

Site location: the department of the Alpes-Maritimes Modelling fire ignition risk : a first « intuitive »
was chosen since it includes both low altitude climatic approach.
conditions and Mediterranean vegetation with a
An initial approach to estimating fire ignition risk is to
dominance of summer fires and a high altitude
look at individual aspects related to human activities
mountain area where fires tend to occur at the end of
and vegetation characteristics separately.
In this case, the impact of urban development, road
The surface area of the department is about 4300 network density, and vegetation type were analysed
km2. individually.
Data period: a 13 year period, from 1991 to 2003 Urban development:
represents the time during which data on fire ignition
According to a study conducted by KALABOKIDIS
were systematically collected and recorded.
(2002) in Greece, urban areas can strongly influence
Description of the data: the French National Forest fire ignition up to a distance of about 300 m, and
Office (“Office National des Forets”, or ONF) provided beyond that distance, the impact decreases sharply.
the bulk of the fire data for the period defined above.

D-08-06.doc 50

We therefore quantified the number of fires Summing the variables creates a raster layer where
occurring within that distance for the different types of the cell values can range from -1 to 16 and where
urban setting shown in Figure 46. greater values indicate greater ignition risk: <0 = no
risk, 0 = very low risk, 0≤4 = low risk, 4≤8 =
For each urban setting, a coefficient was attributed
intermediate risk, 8≤12 = high risk, ≥12 = very high risk.
according to the number of fires recorded within the
300 m limit; coefficient values ranged from -1 to 3. Discussion
The value of -1 was integrated was used in the very
The fire ignition map in Figure 50 shows the
high density urban areas where the risk of a fire is
greatest risk (orange and red) in the urban fringe of the
considered zero, even though 2 fires were identified in
coastal and pre-coastal areas.
that sector.
Table 10 summarizes the number of fires occurring The highly urbanised coastal areas are classed as
within the 300 m limit of each urban type and for each no risk (green) due to the negative coefficient attributed
fire ignition cause. to this sector where forest fires are extremely unlikely.
The bottom of Table 10 shows the coefficients that
were attributed to these types The fire ignition risk distribution can be analysed
according to three types of human occupation. The road network. In the South, the dense urban area reduces the fire
ignition risk to zero.
The distance of 300 m from a house applied above
In the extreme North, there is a forested zone that is
in Table 10 was maintained for fires occurring in
too far from human activities to present much risk of a
proximity to a road as shown in Table 11, which shows
fire ignition that can be attributed to anthropogenic
that there are far more fires occurring near municipal
roads than any other type, and it is for this reason that
The low fire ignition risk is therefore associated with
they have been attributed a coefficient value of 3.
the low density of human activities.
A coefficient of -1 was attributed to autoroutes
Fire ignitions in this zone were due to lightning
because they tend to have large paved shoulders and
which strikes randomly.
few stopping areas in the Alpes-Maritimes, making
Finally, there is an intermediate zone where fire
them less likely to be sources of fire ignition.
ignition risk is at a maximum.
Fire ignition was observed to occur near an
autoroute at Aix-en-Provence,but this remains a rare In this sector, human activities are sufficiently dense
event and was ignored in this study. to create a high probability of fire ignition and yet low
enough for the surface not to be largely non-vegetated Vegetation risk.
and artificial, as in the urban centre.
As for the two preceding factors, a risk coefficient
There is sufficient vegetation to provide a serious
was attributed to the different types of vegetation
fire risk and the human activities are neither dense
according to the relative weight of fire ignitions in each
enough to create largely artificial surfaces
The number of fires and coefficients are described Despite the logical distribution of forest fire ignition
in Table 12. risk described above, the model has some faults.
Summing the coefficients of the three variables
The “moor” vegetation class had no impact on fire
should give a maximum value of 9 (3 variables times a
ignition and was eliminated from the analysis.
maximum coefficient value of 3), but the greatest value
On the other hand, bushes (“maquis”) were strongly
obtained is 16.
associated with fire ignitions and were attributed the
This indicates that some cells were counted more
greatest coefficient.
than once. In the dense road network of the urban
Shrub species favour fire ignition whereas tress
fringe, cells located less 300 m from two or more roads
species play an important role in fire propagation under
would be counted once for each road.
certain vegetation humidity conditions.
The fire ignition risk is therefore overestimated in The “structural” risk associated with fire this sector.
In addition, the choice of a 300 m limit is somewhat
The structural risk map shown below (Figure 50) arbitrary since this value, observed in Greece, may not
was elaborated working in raster mode with a cell size be the most appropriate in the Alpes-Maritimes region.
of 250 m, the resolution of the Corinne Land Cover.
Finally, the coefficient values varying from -1 to 3
The IGN BD Carto data base was in vector format
were selected intuitively and may not reflect the actual
and had to be rasterized.
weighted importance they represent.
Each raster cell was attributed the value of the
coefficient attributed to it above. The second modelling approach: statistical
model A
The final ignition risk map was obtained by summing
the individual coefficients of the variables defined After the initial intuitive approach, a statistical
above. method was used to estimate fire ignition risk.

D-08-06.doc 51

This second model is based on the number of fires - the road network: as it was the case in the urban
occurring with varying distances of the two main fringe, the number of fire ignitions decreases
factors, the road network and urban land-use. exponentially with distance from a roadway (Figure
In this approach, the number of fires occurring with 58). Distance from the road network is therefore a
successive bands of 25 m was calculated to estimate good predictor of fire ignitions
the range of impact of these factors on fire ignition.
For both the urban fringe and northern mountainous
The vegetation variable was not taken into account areas, the distance from a roadway is a good predictor
since its spatial distribution was not adapted to the of fire ignitions for a distance of up to 250 m from a
approach. road.
Urban development is a good predictor only in the
In addition, no data were available on the density
urban fringe for a distance of up to 500 m.
and height of the different vegetation layers (of the
In the northern sector, there is no apparent
shrubs in particular) which could have been useful
relationship between human settlements and fire
descriptors of the contribution of the vegetation to
ignitions, so this factor was eliminated in this sector in
ignition risk.
the following analyses. Analysis of the factors at the departmental The fire ignition risk map developed using
statistical method A.
For the entire Alpes-Maritimes department, the
The fire ignition risk map using this method was
number of fires per band of 25 m from urban areas was
elaborated in the following way.
counted (Figure 51).
A distance grid is created for each of the factors for
Unlike the distance from urban settlements, Figure the urban fringe and mountainous area.
52 shows a correlation between the number of fire It is important to note that these two geographical
ignitions and distance from the road network. entities are treated separately.
The number of fires decreases exponentially with
Once that has been done, a new raster data layer is
increasing distance with a coefficient of determination
2 created by applying the regression equations shown
(r ) greater than 0.95. Distance from roadways is above.
therefore considered a good predictor of fire ignitions. This results in three new raster layers: two for the
urban fringe where both urban density and road
In order to improve the model, the Alpes-Maritimes
network are significant factors, and one for the
department was separated into two sectors of different
mountainous area where only road network is taken
urban density.
into consideration. Separating the department into two sectors. Each layer shows the potential number of fires
estimated by the regression equations.
Before separating the department into two zones,
Classes of fire ignition risk are determined from the
the dense urban coastal area was removed from the
sum of the factors in the urban fringe and the fires
study since the risk of a forest fire ignition there is
predicted by the road network alone in the mountainous
The northern part of the department is mountainous
The results of Model 2, method A are shown in
and rural with low population and road densities as can
Figures 59 (risk map alone) and 60 (risk map and
be seen in Figures 53 and 54.
actual fire ignitions)
Between the northern part and the dense urban
coastal area, an urban fringe can be found.. Discussion. The urban fringe. The relationship between fire ignition risk and actual
- the impact of urban development : in this sector, the fires shown in Figure 60 appears satisfactory in the
number of fires decreases exponentially with the mountainous area since most of the fire ignitions occur
distance from human structures (Figure 55). The with the high risk zone.
limits of the sectors appear justified since there is However, in the urban fringe several of the high risk
now a good relationship between the number of fire zones have few or no actual fire ignitions.
ignitions and distance from human settlements- The validity of the model in this sector is therefore
- the road network : fire ignitions decrease put into question.
exponentially with distance from a road in the urban
One of the reasons for the poor correspondence is
perhaps because the road network and urban density The mountainous and rural northern sector are not independent since increasing urban density
of the department inevitably increases road density and vice versa.
- the impact of urban development: in the northern This covariance is not taken into account in the
sector, there is no apparent relationship between urban fringe where the results of the predictive
the number of fire ignitions and distance from equations are summed.
human settlements as can be seen in Figure 57.
In the mountainous zone, much of the surface area
Human activities are therefore not a good predictor
is in a low class risk, and some fires have occurred far
of fire ignitions in this area.
from both houses and roads.

D-08-06.doc 52

These can be attributed to either natural fires started In addition, the coefficients enable us to estimate
by lightning or to humanly caused fires associated with the probability of fire ignition as a function of distance
pastoralism. from urban activities or settlements (X1) and roadways
Shepherds sometimes ignite fires to maintain grass (X2).
quality and occasionally these fires escape their In order to obtain the erosion risk map, the equation
control. is applied to the database.
This hypothesis could be tested by determining the We then obtain P values which range from 0.26 to
location of these fire ignitions with respect to the sheep 0.008.
grazing areas, but this information is not available These values are then subdivided into four fire
through local agricultural authorities. ignition risk classes; added to these classes are the two
classes which include the cells not taken into account
Since the statistical approach used in the second
(zero and very low).
model proved deficient in the urban fringe, another
statistical approach was tested. Figures 61 and 62 show the spatial distribution of
fire ignition risk based on this method..
7.2.3 The third model: statistical approach B.
Nearly half the surface area is in the low risk class,
This third model is also based on the distance from
yet fires have occurred in these zones in the past. Most
urban activities and roadways but it uses logistic
of these fires can be attributed to either lightning, which
regression which takes into account all of the cells.
strikes randomly, or pastoralism, which tends to be
This type of regression uses binary type data:
located far from roads and houses.
absence or presence of a fire ignition between 1991
and 2003. The remaining area has a variable risk according
the distance from roads and urban dwellings as The method.
predicted by the logistic regression.
To optimize the calculations, the grid size was Much of the urban fringe is in a high risk class, as it
increased to 1 km2 squares. was for the first model.
The dependent variable in this regression is the Pressure from human activities in this zone makes it
number of fire ignitions between 1991 and 2003 in each particularly vulnerable to fire ignitions.
1 km2. The coefficient of determination (r2) for the logistic
For each cell the mean distance from the centre of regression model is only 0.063, but this does not
the cell to a road or settlement is calculated. invalidate the model since the r2 value is considered
Table 13 shows that the tendencies observed above less meaningful in logistic regression than in classical
are still valid, but the relationship is weaker since cells linear regression.
without a fire ignition are also included in the
regression. On the other hand, the probability associated with
Pearson’s Khi2 value is pertinent and the risk of
In logistic regression, dependent variable is normally erroneously attributing explicative value to the road and
binary (0 or 1), but since some cells may have been urban variables is less than 0.1%.
affected by several fires (>1), the Y variable is
transformed into a series of repetitions which each 7.2.4 Model comparison and conclusion.
correspond to a Bernoulli law with the same From a simple visual evaluation of the risk maps,
P parameter, so that Y follows a binomial law with the three models appear to give similar results apart
parameters n, and P, as shown in the equation below. from differences in the spatial resolution.
1 − P ( X 1, X 2) In all three cases, the coastal area along the sea
LogitP ( X 1, X 2) = has no risk due to its high urban density, a
P ( X 1, X 2) mountainous zone in the North where the risk is low,
and finally an urban fringe zone between the two others
= Exp-(a0 + a1X1 + a2X2) where the risk is particularly great.

The intensity of the risk is related to the Despite the similarities, some differences between
the models persist.
exp-(a0+a1*X1+a2*X2) term. The first two models have spatial resolutions of 250
m whereas the third has a resolution of 1 km. Results and risk maps. The loss in spatial detail is compensated for by a
The solution to the above equation gives the results more objective approach and valid statistical
summarized in Table 14. parameters.
Models 1 and 3 are the preferred options: the first
The results of the logistic regression show that the for its simplicity and ease of application, the third for its
values of the coefficients are statistically significant. rigour.
The values of their critical probabilities are always <

D-08-06.doc 53

The three models agree with Figure 63 which shows In this study, non-stationarity is the main assumption
the general risk behaviour associated with human made for modeling the driving factors behind wildland
pressure on the environment: in the first segment fire occurrence.
(mountainous), the risk is low due to low human To overcome the constraints of traditional
pressure, the main cause of fire ignition. regression modeling which assumes stationary
In the third segment (coastal), the risk is also low, processes we applied geographically weighted
but in this case it is due to the scarcity of the vegetation regression (GWR) analysis.
in this densely urbanised environment. GWR considers that relationships vary in space
In the second segment (urban fringe), the risk is at a according to their location and allows local variations to
maximum due to an abundant supply of combustible be taken into account.
vegetation and the greater likelihood of fire ignition due
7.2.5 Study Area and Wildland Fire Database
to a high concentration of human activities.
The geographical extent of our study area includes
Improving fire ignition risk modelling can only be
Portugal, Spain, Southern France, South-Eastern
done with better data acquisition techniques.
Switzerland, Italy and Greece (see Figure 64).
Both fire ignition location and cause must be
Within the study area census data at provincial level
investigated and recorded systematically in order to
were acquired from national statistics.
provide a fuller and more reliable data base.
In total, 153 geographical units were identified, 18 in
In Spain, for example, forest fires causes are
Portugal, 48 in Spain, 14 in South France, 20 in Italy, 2
defined for the vast majority of fires, leaving very few
in South-Eastern Switzerland and 51 in Greece.
For most of the countries these geographical units
Example of human ignition danger estimation at
correspond to NUTS-3 level (Nomenclature of
european scale
Territorial Units for Statistics) except for Italy where
Continuing the work developed in Megafires Project they correspond to NUTS-2 level and for Portugal
(see section 2.1) a geo-statistical analysis to model where they correspond to their national district level.
wildland fire occurrence was carried out at provincial Unfortunately, this not fully homogenized division
level in Southern Europe from socio-economic and among the countries, which resulted from limitations on
demographic indicators of 1960 and 1990 together with data availability, could eventually influence negatively
variables that describe land cover and agricultural the regression analysis.
The basic data used in our study come from the
This work is has been presented in Koutsias et al
MEGAFiReS project 1999) with an update for including
(2005) and is extended and inproved in a new paper
the Southern part of Switzerland together with the
actually in revision (KOUTSIAS et al. 2006).
number of wildland fires occurred within the period
Here is presented main results obtained in the first
The database is composed by socio-economic and
In this study the classical OLS linear regression demographic indicators together with variables that
modeling has been used along with geographically describe land cover and agricultural activities.
weighted regression modeling to explain long-term In total, 77 variables were established for 1960 and
wildland fire occurrence patterns at Southern Europe. 1990.
By applying the global OLS linear regression we Differences of some variables (mainly for population
identified critical underlying causal factors for wildland characteristics) between the two date sets were also
fire occurrence from a set of socio-economic, computed showing changes in population
demographic, land cover and agricultural statistics. characteristics associated to urbanization and land
Traditional regression modeling assumes that the
These selected variables were considered of being
relationship between the dependent variable and the
potential underlying causal factors for explaining long-
explanatory variables is constant regardless of their
term fire occurrence patterns.
geographical location.
This assumption, which is known as stationarity, is Besides the census data, the mean annual number
often violated in real world situations. of wildland fires, having occurred between 1992-2000
However, this constraint is overcome by another at provincial level, was also computed from national
approach, known as geographically weighted forest fires statistics.
regression (GWR), which considers that such This variable is used as depedent variable.
relationships vary in space according to their location For Portugal, Spain, Southern France and Italy
and that allows local variations to be taken into wildland fire observations were provided using the
account. community centroids for the period 1992-2000.
In wildland fire occurrence modeling, especially For Switzerland the x and y coordinates of the
when the geographical extent of the study area is large ignition points were used having occurred in the same
(i.e. the whole Southern Europe), it would be more period.
reasonable to find varied rather than constant Finally, for Greece, wildland fire observations were
relationships. provided using the x and y coordinates of the ignition
points for the period of 1985 to 1995.

D-08-06.doc 54

Out of this original database, forest fire data were 7.2.7 Results and discussion
aggregated at provincial level so that to much the Global versus GWR linear model
aggregation scheme used for the census statistics.
Wildland fire occurrence was expressed using the A stepwise ordinary least squares (OLS) regression
mean annual number of fires to overcome the was carried out for developing a model with the most
inconsistence observed between the time period of significant variables.
Greek data and the rest of the countries. Eight of the original variables have been chosen by
A qualitative visualization of the spatial distibution of the model which explains 53% of the variation of the
the dependente variable reveals critical regions for fire dependent variable.
occurrence. Among them the density of livestock (i.e., sheep) in
In future studies the mean annual number of fires is 1990, the density of agricultural employees in 1990, the
going to be estimated for two different fire size classes percentage of forested area, and the difference of the
(burned area above 1 ha and above 100 ha). youth index between 1960 and 1990 are the most
important variables based on the criterion of
7.2.6 Methodology: Geographically Weighted standardized coefficients of the model.
Regression (GWR), the linear and logistic The GWR provided significantly better results than
case the global regression model, since the variance of the
GWR tries to capture the spatial variation by dependent variable explained increased to 68.65 %.
calibrating a multiple regression model so that at This is caused by the consideration of the non-
different points in space different relationships between constant relationships between the dependent and the
variables can be found (ZHANG and SHI 2002). explanatory variables, which change throughout the
A regression model is fitted at each data point by EUMed basin.
weighting all observations from that point as a fuction of The residuals in GWR (Figure 64) are significantly
distance. less than those of OLS regression, indicating the better
Consequently, the neighbors sampled around the performance of the former model.
point influence the regression coefficients more Based also on an ANOVA test, statistical significant
strongly than the observations farther away. improvement of the GWR model over the OLS linear
GWR estimates the parameters at each point in the regression model was found.
study area which then can be mapped using for A Monte Carlo significance test for the parameters
instance geographic information systems (GIS) to of the model found significant spatial variability on the
investigate local spatial variation in the regression intercept and on the coefficients of the difference for
relationships (FOTHERINGHAM et al. 2002). the youth index between 1960 and 1990, the number of
In geographically weighted regression the agricultural employees in 1990, and the density of
relationship between y and x can be expressed as: sheep in 1990.
Based on the same test, non-significant variability
p was found on the % of forested area, the % of the
y i = β 0 (u i ,ν i ) + ∑ x ij β j (u i ,ν i ) + ε i difference of the active population between 1960 and
j =1 1990, the density of agricultural employees in 1990, the
where % of difference in size of agriculture exploitations
- β0(ui, νi), and βj(ui, νi) are estimated coefficients as between 1960 and 1990, and the number of sheep in
a function of location, and 1990.
- ε is a random error term. Global versus GWR logistic model
Besides the linear regression model we developed
Besides the OLS linear regression model we
also a global and GWR logistic model.
developed also a global and GWR logistic model. As
In addition to linear regression, where the outcome
mentioned, logistic regression presupposes a binary
variable is supposed to be continuous, logistic
dependent variable that takes the value 1 in case of an
regression presupposes the existence of a binary or
event and 0 otherwise.
dichotomous dependent variable.
Thus we created a new variable which is based on
This main difference between the two regression
the reclassification of the kernel density surfaces (two
models renders logistic regression very popular,
classes based on the equal area criterion for each
especially when the experimental question can be
expressed in a dichotomous way.
Then, the median value (1 or 0) found inside each
Besides this, the independent variables in logistic
polygon (= provinces) was attached as the dependent
regression modelling can be a mixture of continuous
variable (Figure 65). We used the same independent
and categorical variables.
variables as in the case of linear regression.
Consequently, the assumption of multivariate
normality is not presupposed in logistic regression.
Previous use of LR in fire occurrence at provincial
level in Southern Europe can be found in Chuvieco et
al. 1999b).

D-08-06.doc 55

Three out of the four most significant variables of

7.2.8 Conclusion
linear regression (the density of agricultural employees
in 1990, the percentage of forested area, and the In conclusion, GWR seems to be a valuable
difference of the youth index between 1960 and 1990) approach for exploring and modeling non-stationary
were considered also by logistic regression as the most relationships between the response and explanatory
significant ones based on the Wald statistic. The GWR variables and thus to better understand the spatial
logistic model improved the deviance (-2LL) from processes in wildland fire occurrence.
178.35 to 116.54. The reduction of Moran’s I spatial autocorrelation of
Also, the GWR logistic model classified 84.97 % of the residuals indicates that GWR models were more
the observations successfully, which is significantly appropriate than the classical global ones.
higher than 66 % of the ordinary global logistic model.
The residuals in GWR logistic model, as shown in
Figure 65, are less than those of ordinary logistic
regression indicating the better performance of the
former model. Spatial Autocorrelation of Residuals

To further explore the advantages of the GWR over

the classical global approach we calculated the
correlograms of the residuals which show how they are
distributed in space (Figure 66).
A reduction of Moran’s I spatial autocorrelation is
evident in lag distances beyond 100 Km especially in
the GWR linear model.
This reduction indicates that the model is
The existence of spatial autocorrelation in OLS
regression violates the assumption of independent
errors (LICHSTEIN et al. 2002).

D-08-06.doc 56

7.2.9 Figures

Figures 44 and 45: Topographic map (Source: BD Carto IGN) and Vegetation map (Source: Corinne LC)

Figure 46 and 47: Landuse types map (left) (Source Corinne LC) and road network map (right) of Alpes-Maritimes.
(Source: BD Carto IGN)

D-08-06.doc 57

Figure 48: Recorded fires between 1991 and 2003 in the Alpes-Maritimes

Left: Figure 49: Spatial distribution of fires occurring during the 1991-2003 period according to cause

Right: Figure 50: Map showing the structural forest fire ignition risk
The black points show actual fire ignitions, and the colours indicate the calculated risk

Figures 51 and 52: Number of fires versus distance to habitats (left) and to road network (right)

D-08-06.doc 58

Figures 53 and 54: Urban density (left) and road network density (righ) in the Alpes Maritimes deparment

Figures 55 and 56: Number of fires versus distance from homes (left) and from a roadway (right) in the seashore

Figures 57 and 58: Number of fires versus distance to settlements (left), and to a roadway (right) in northern sector

D-08-06.doc 59

R= ∑ F(Xi)
ie Rlito = 45,545*Exp(-0,0126*X1) + 23,65*Exp(-0,0053*X3) (1)
Ra-pays = 38,10*Exp(-0,0101*X2) (2)
1: Littoral road network X2 : Northern sector road network X3 : Littoral distance to habitations

Figures 59 and 60: Fire ignition risk map, model 2 (left) with 1991 - 2003 ignitions shown in white (right).

Figures 61 and 62: Fire ignition risk map obtained using the logistic regression model (left)
with fire departures between 1991-2003 (right)

D-08-06.doc 60

7.2.10 Tables

Table 10: Number of fire ignitions according to urban density and ignition cause
for fires less than 300 m from a residence.
Discontinuous Agricultural with Natural areas with
Origin Dense urban Single dwellings
urban isolated dwellings isolated dwellings
Unknown 2 11 17 23 22
Voluntary 0 1 3 7 2
Other 1 0 2 10 6
BBQ 0 0 0 3 0
Gain 0 0 0 0 1
Electrical 0 0 0 0 0
Pastoralism 0 0 0 0 0
Lightning 0 0 0 0 1
Construction 3 0 4 12 9
% 5 10 18 34 32
Coefficients -1 1 2 3 3

Table 11: Number of fire ignitions according within 300 m of a road for each ignition cause.
Origin Autoroute National route Departmental Municipal
Unknown 3 13 47 112
Voluntary 0 2 4 21
Other 0 5 0 5
BBQ 0 6 10 21
Gain 0 0 2 0
Electrical 0 0 0 0
Pastoralism 0 0 1 15
Lightning 0 0 1 0
Construction 0 4 9 41
% 0.08 8 20 62
Coefficients -1 1 2 3

D-08-06.doc 61

Table 12: Number of fire ignitions according to vegetation type and ignition cause.
Origin Coniferous Broadleaf Mixed Shrubs Grass Bush
Unknown 37 43 3 13 0 53
Voluntary 4 3 0 3 0 10
Other 1 1 0 0 0 0
BBQ 10 8 3 0 0 10
Gain 0 3 0 2 0 3
Electrical 1 0 0 0 0 1
Pastoralism 5 1 0 1 0 6
Lightning 5 5 0 0 0 1
Construction 11 8 1 1 0 7
% 27 27 3 7 0 35
Coefficients 2 2 1 1 0 3

Table 13: Correlation matrix

Fires Urban distance Road distance
Fires 1 -0,060 -0,109
Urban distance -0,060 1 0,119
Road distance -0,109 0,119 1
Bold font indicates a significant corrrelation at α=0.05 (bilateral test)

Table 14: Results of the logistic regression and equation giving the fire ignition probability
as a function of two factors, X1 and X2.
Estimated model parameter values (maximum likelihood) :
Parameter Estimated value Standard deviation Khi² Pr. > Khi²
Constant -3,515 0,107 1073,719 < 0,0001
Urban distance (X1) 0,000 0,000 10,999 0,001
Road distance(X2) -0,005 0,001 50,738 < 0,0001

P(X1, X2) = 1 / (1+ Exp-(-3,515 - 0,000158*X1 - 0,005*X2))

D-08-06.doc 62

7.3 HUMAN IGNITION DANGER IN SOUTHERN As a non-parametric smoothing technique, the

EUROPE BASED ON FIRE OCCURRENCE kernel density estimation (SILVERMAN 1986) is well
MAPS suited for converting fire ignition point data into
continuous “fire occurrence distribution surfaces”.
The objective of this and next Eufirelab deliverables
Kernel density estimation is based on the estimation
is to elaborate a fire risk index at European scale.
of the density at each intersection of a grid
In order to do it, and taking into account the
superimposed on the data, after placing a probability
resolution of satellite images and the size of the study
density (kernel) over each point event.
area, it was decided that the ideal resolution of the pixel
Depending on whether constant or multiple adaptive
for the GIS layers must be 1 square kilometre.
values are used for the smoothing parameter, kernel
In the case of the human ignition factors it is very density estimation is distinguished into the fixed and
difficult to get and have access to available and reliable adaptive method, respectively. (SILVERMAN 1986,
data at this resolution or scale, for the whole or at least Levine 2002).
big part of the Euro-mediterranean area. KOUTSIAS et al. (2004) applied the kernel approach
The main problem we find with the spatial variables to assess fire occurrence patterns at landscape level by
(distances, fragmentation, interfaces) derived from addressing some of the inherent positional inaccuracies
cartographic layers like roads, urban areas, forest and of the fire ignition locations.
land use covers, is that this original layer are very DE LA RIVA et al. (2004) used also this approach to
generalist at this global scale, with very poor detail. In express fire occurrence patterns at municipality level by
the cases of socio-economic factors (unemployment, using fire ignition observations.
rural population age, population density…) estimated
The current study proposes an adaptive kernel
from census data, the main inconvenience is that they
density interpolation approach applied to community
are only available in these moments for provinces.
centroids where the number of fires per community is
This data unavailability at the desired scale
used as the intensity variable for the kernel density
(municipality level) for the whole area forced us to
discard the socio-economic variables because we don’t
The final aim is to define large-scale fire occurrence
consider appropriate to rasterize at 1 km2 the
patterns and to identify ‘fire occurrence hot spot areas’
information derived from provinces.
in southern Europe, using fire ignition observations
The same problem regarding vector layers, like
aggregated at community level. In our study the
roads, with scales larger than 1: 1.000.000.
number of fires form the period 1992-2000 have been
Because of these present difficulties to have an estimated from national fire statistics and expressed at
operational human danger ignition layer, one possibility community level using the community centroids (Figure
can be generating it from historical data of fire 67).
occurrence. The geographical extent of the study covers the
From the location of ignition points and using the European Mediterranean countries including Portugal,
kernel density approach, that is a well known and Spain, Southern France, Italy, Greece, as well as
explored interpolation method, we can generate a South-Eastern Switzerland (Canton of Grison and
probability surface, as is showed in section 2.2 about Ticino) from central Europe (Figure 67).
Spread Project review.
Motivated by the non-homogeneous spatial
Occurrence maps provides a global view of fires,
distribution of community centroids (Figure 67) we
which are mainly caused by human factors in Europe.
decided to choose the adaptive kernel density
Visual exploration of spatial distribution of point estimation mode instead of the fixed one.
observations is popular in point pattern analysis and The adaptive mode allows for the adjustment of the
takes place prior to any further analysis (BAILEY and bandwidth size in relation to the concentration of the
GATRELL 1995, Fotheringham et al. 2000). community centroids (WORTON 1989).
By visually examining point observations, one may For the adaptive kernel approach different
acquire a prior idea about their spatial distribution, bandwidth sizes have been tested (from 1 to 40 at
suggest hypotheses about explanatory factors, and various steps) that are indicative of typical resulting
decide for further statistical analysis and modeling patterns of under- to over-smoothing.
(SADAHIRO 2003). Here we present four representative steps (2, 5, 10,
In the following section is explained deeply the For a homogeneous process with no spatial
generation of this map, according with the information dependence, the expected number of events within a
of Deliverable D143 “Fire Risk and Human Factors (III)” distance d of a randomly chosen event equals ?pd2
from Spread Project. where ? is the density (GATRELL et al. 1996).
7.3.1 Fire Occurrence Hot Spot Areas Thus, if the expected number of events corresponds
to the number of centroids used in the adaptive kernel,
For the analysis of large-scale fire occurrence the estimated distance d can be used for defining the
patterns, we often need to transform fire ignition data bandwidth size in the fixed mode.
that come in different formats and accuracies (e.g. x The corresponding bandwidth sizes in the fixed
and y locations or number of fires per area unit) into mode are 3622, 5726, 8099 and 11453.
continuous surfaces.

D-08-06.doc 63

We applied both approaches, the fixed and the Thus kernel densities have been reclassified to 10
adaptive kernel density interpolation, at national level in classes based on the equal area criterion within each
Portugal in order to decide about the most proper country (Figure 69), presupposing equivalence for fire
approach and the most appropriate bandwidth size hot spot areas among the countries.
(Figure 68).
The effect of small bandwidth size to kernel density
surfaces is evident in the corresponding images in
Figure 68.
Low bandwidth sizes for both, the adaptive and the
fixed kernel density, (e.g. comprising 2 and 5
community centroids or 3622 and 6726 meters),
created surfaces in which the local variation of the
density surfaces is enhanced around the interpolated
This effect is more evident in the fixed kernel density
approach because the bandwidth size remains
constant throughout the whole extent of the study area
while the concentration of the interpolated points
(community centroids) varies significantly.
This variation in the concentration of community
centroids is taken into account by the adaptive kernel
mode since the bandwidth size varies locally so each
time the same number of community centroids to be
Based on visual inspection, among the different
bandwidth sizes used the one that corresponds to 10
community centroids is considered of being the most
appropriate to avoid under- or over smoothing.
To evaluate the density surfaces three grids of 2, 5
and 10 Km resolution were overlaid over the density
Subsequently the number of fires and the sum of
density kernel values were estimated.
The correlation coefficients (Table 15) were slightly
higher between the number of fires and the adaptive
kernel density surfaces.
This is another indication that the adaptive kernel
density interpolation performs better compared to the
fixed kernel density as the former allows for locally
varying bandwidth sizes.
The same methodology described previously and
applied at national level in Portugal has been applied to
the rest of the countries.
The absolute values of kernel densities for each
country depend on the number of fire ignition points as
well as on the spatial distribution of community
centroids and wildland fire observations.
The total number of wildland fire ignition
observations varies greatly between the different
countries and consequently kernel density values are
also very diverse.
The consequence of this variability found among the
countries results in an underestimation when the kernel
density surfaces are joined together under a common
classification scheme as shown in Figure 69.
This problem can be overcome if a reclassification
of kernel density surfaces is applied before merging all

D-08-06.doc 64

7.3.2 Figures

Figure 63: Fire ignition risk evolution according to vegetaion and human activity characteristics. (Carrega, 1992)

Figure 64: Observed and predicted values of OLS and GWR linear models and the distribution of their residuals.

Figure 65: Observed and predicted values of ordinary and GWR logistic models and their residual distribution.

D-08-06.doc 65

Figure 66: Correlograms of the residuals

Figure 67: Community Centroid

Figure 68: Kernel density estimates using the fixed (upper images) and
the adaptive kernel approach (lower images).

D-08-06.doc 66

Figure 69: Large-scale fire occurrence patterns at European level by kernel density estimation

7.3.3 Tables

Table 15: Correlation coefficients between number of fires and kernel densities at 2, 5 and 10 km grid resolutions

D-08-06.doc 67



The propagation danger component hast to cover
the potential hazard that a fire propagates in space and
time. In order to obtain a global view of risk associated to
This index is associated to estimating the behaviour fuel loads, terrain characteristics and wind flows, a
of the fire. global simulation analysis was performed.
Fire behaviour is mainly described with reference to This analysis tried to obtain average values of rate
the propagation rate and the intensity of the flame front. of spread and flame length, considering different wind
and topographic conditions for the estimated fuel maps
Not too many studies include this type of index. At
of the whole EUMed area.
global scale, a case applied to Euro-Mediterranean
This attempt should be considered as a general
basin has been included in the Spread project,
overview of average expected fire behaviour at global
described in chapter 4.
scale, in order to rank different danger levels according
The propagation danger index (PD) is composed by
to the combination of fuel and terrain spatial patterns.
two subindex: the average rate of spread (RoS), and
The estimation of the average RoS was based on
the flame length (FL).
several simulations performed by the Autonomous
Both variables are calculated using BEHAVE
University of Barcelona for different fuel types, slope
ranges and wind flows.
The input data is obtained from CORINE and a
global Digital Terrain Model. As a simulation kernel the wildland simulator
The results is a map taken as static, it will not proposed by COLLINS D. BEVINS, which is based on the
change throughout the fire season. fireLib library (COLLINS, 1996) were used.
FireLib is a library that encapsulates the BEHAVE
Taking into account the characteristics of the index
fire behaviour algorithm (MORGAN et al., 2001).
described in Spread project (input variables, spatial and
In particular, this simulator uses a cell automata
temporal resolutions), we propose this index as the
approach to evaluate fire spread.
“Propagation Danger Index” ot he Euro-Mediterranean
The terrain is divided into square cells and a
Wildland Fire Risk Index (EM-WFRI) defined in this
neighbourhood relationship is used to evaluate whether
a cell will be burnt and at what time the fire will reach
The definition and estimation of this index are also
the burnt cells.
described in the same chapter 4.
As inputs, this simulator accepts maps of the terrain,
vegetation characteristics, wind and the initial ignition
The output generated by the simulator consists of
two maps of the terrain.
In the first one, each cell is labelled with its ignition
time; in the second one, each cell is labelled with its
flame lenght.
This information must be used to calculate the rate
of spread and an average from among all flame lenght.
To calculate the rate of spread, the distance
between the ignition point and each particular cell in the
terrain is divided by the ignition time of that particular
This calculation is repeated for each cell in the
terrain to determine the maximum value of the rate of
This maximum value is used as the rate of spread
for that particular situation.
To provide the propagation danger map, a set of
prototype plots was created, considering all the fuel
models from Rothermel classification and a certain
slope percentage (from 0 to 100%, with a step of 5%).
The total number of plots was 273. Each plot
consists of a grid of cells with 11 columns x 11 rows
(each cell measured 328.083 x 328.083 feet).
The ignition point was located in the middle of the

D-08-06.doc 68

For each plot, many input parameter combinations 8.2 PROPAGATION DANGER (PD)
were used to simulate the wildland fire behaviour and
It derives from the combination of the two
the average rate of spread and flame lenght were also
intermediate products previously described: RoS and
The parameters considered for variation were: 1-hr
The results of the simulations were mapped at
dead fuel moisture, 10-hr dead fuel moisture, 100-hr
EUmed scale using CORINE land cover (reclassified
dead fuel moisture, live herbaceous moisture and wind
into fuel models) and slope maps
speed and direction.
The maps of FL and RoS were then normalized
Average values of rate of spread (RoS, in m/min)
using linear fitting and multiplied to produce PD:
were computed for the different fuel types.
Finally, the RoS values were scaled into a 0-1 PD = [(RoSi – RoSmin)/(RoSmax - RoSmin)+0.001] *
range, by normalizing the values between the [(FLi – FLmin)/(FLmax - FLmin)+0.001]
maximum and minimum values (figure 70).
Fuel types were derived from the Corine land cover. A small constant (0.001) was added to avoid zero
The resulting map is considered static, since no multiplication in case of minimum values.
RoS and FL were considered in this formula of
specific conditions are simulated (wind or FMC), but
only general patterns of propagation rates. equal importance, although this could be tuned up in
future improvements, according to further experience or
Same process as above was performed to compute suggestions.
the average flame length (FL), measured in metres and
normalized into a 0-1 scale (figure 71). This map is taken as static, i.e. it will not change
throughout the fire season.


Figure 70:Estimated average Rate of Spread (Normalised values from 0 to 1)

Figure 71: Estimated average Flame Length (Normalised values from 0 to 1)

D-08-06.doc 69


9.1 POPULATION VULNERABILITY Risks to these populations, as pointed out by CASE et

al. (2000), is expressed as a greater difficulty of
9.1.1 Introduction
successful evacuation, greater susceptibility to health
A vulnerability analysis is realized to obtain impairment, loss of possessions (which could be
information about the susceptibility of individuals, irreplaceable for people in two of these special groups),
property, and the environment to the adverse effects and loss of jobs and income.
of a given hazard in order to develop appropriate
In brief, houses which are dependent upon domestic
prevention strategies.
water supplies and fuel-wood, which lie in remote areas,
The analysis of this information helps to determine
and which have difficult routes of ingress and egress are
who is most likely to be affected, what is most likely
at greater risk of damage or destruction in the event of
to be destroyed or damaged, and what capacities
exist to cope with the generated effects.
People who are very young or very old are more likely
To perform a proper and detailed vulnerability to experience adverse health consequences in the event
analysis it should be necessary: of wildfire, and people who have very low incomes are
- to identify that part of population which is more likely to suffer irreversible economic and social
vulnerable because of proximity or exposure consequences. When these groups of people suffer from
factors; wildfires, they have less individual means of recovery,
- to assess vulnerability of buildings, infrastructure and require more help from the communities they live
and other aspects of the physical environment within (CASE et al., 2000).
due to factors such as site, materials used,
To identify those sections of a community most likely
construction technique, and maintenance;
to be affected by a particular hazard and to determine
- to evaluate transportation systems,
areas that require strengthening to prevent or mitigate the
communications systems, public utilities (water,
effects of the hazard, data on several variables must be
sewage, power), and critical facilities (i.e.,
collected (among which, information on the size, density,
hospitals) for weaknesses; to identify most
location, and socio-economic status of the at risk
vulnerable population (e.g., older adults, children,
single-parent families, the economically
disadvantaged, the disabled); Vulnerability assessment is one of the least
- to estimate the level of poverty, jobs that are at investigated tasks, due partly to the lack of relevant
risk, and the availability of local institutions that detailed socio-economic data and the difficulty of their
may provide social support; effective spatial representation for integration with
- to determine the community’s potential for physical environmental data on hazards (CHEN et al.,
economic loss and recovery following a disaster; 2003).
- to identify existing measures and resources that
Vulnerability assessment is however an important task
can reduce the impact of a given hazard.
in risk assessment and has social significance for a
The degree to which populations are vulnerable to hazard-prone vulnerable community.
hazards is not solely dependent upon proximity to the
Assessing vulnerability in spatial terms requires a
source of the threat or the physical nature of the
wide range of physical and socio-economic knowledge
hazard, social factors also play a significant role in
and expertise.
determining vulnerability (CUTTER et al., 2000).
In this context, GIS can be used for database
The primary social effects of wildfires are physical: establishment, analytical modelling, and decision support
houses and other kinds of human structures will burn; in a decision-making process.
people may be injured or may die.
GIS can play an important and integral role in
However, it should be noted that death and injury
lessening the adverse impacts of natural hazards on
depend upon unique sets of circumstances
surrounding each fire event: these cannot be
Having a wide range of spatial analysis techniques
modelled here.
and tools, they are helpful for identifying, measuring, and
For instance, as noted by CASE et al. (2000), in a
assessing many aspects of natural hazards and their
geographic information system, we can only describe
consequences (CHEN et al., 2003).
the total number of people at potential risk, and the
GIS spatial analysis in particular, applying various
number of people having special characteristics
methods and techniques, has the ability to employ
which place them in jeopardy.
physical environmental and socio-economic data for risk
Some individuals belong to special populations
and vulnerability analysis.
because they have characteristics, which lead them
to suffer effects disproportionate to the general High-quality GIS databases support subsequent risk
population. assessment and rational decision-making in a spatial and
Special populations include families with very temporal context, which can help risk managers and the
young children, the elderly, and households whose public understand how complex hazards and their
members have incomes below the poverty line. consequences will affect vulnerable communities.

D-08-06.doc 70

Spatial information system can be used to Additionally, some crucial social effects depend very
estimate the physical risk to people and houses, and much on the exact nature of the circumstances
to draw useful inferences about the nature of surrounding each fire or upon problematic factors that
secondary, less tangible, and less predictable social cannot be adequately described in any known model
effects resulting from wildfire (CASE et al., 2000). (CASE et al., 2000).
To reach these tasks geo-referenced data layers Some general and basic census data can be utilized
must be collected from a variety of sources. to develop indicators of relative risk or vulnerability to
human populations.
When we have to assess vulnerability to wildfire,
The elements chosen for this analysis can be the total
and in general in all the natural hazards context, the
population and population density within each unit (for
required data can be divided into three categories -
instance a 1 km2 grid) and populations that might be
physical environmental data, socio-economic
placed at additional risk due to their circumstances
environmental data, and management-related data
(children, elderly, low-income).
(CHEN et al., 2003).
These factors can be chosen as examples, not as an
Socio-economic environmental data are in exhaustive listing of relevant social information; however
particular used to assess community vulnerability and they can give some information about the ways in which
include population and housing census data and data people experience wildfire, and how they may be affected
on utilities and access. by it, providing decision makers with useful information in
making judgments about prioritising mitigation and
Because of the increasing emphasis on protection programs (CASE et al., 2000).
community vulnerability assessment, collecting socio-
economic data is essential. We could construct a density map of these special
According to Granger 1998) detailed information populations, reasoning that proportionately greater risk
on setting, shelter, sustenance, security, and society occurs where people and houses with these particular
is required. characteristics occur in greater numbers (CASE et al.,
For example, data on shelter include construction 2000).
materials of the walls, roofs, and floors, and the ages
Another crucial data set is represented by the number
of buildings.
of ignitions per 1 km2 grid over a certain historical period.
Data on utilities (e.g., water, electricity,
On the assumption that areas experiencing the most
telecommunication, gas), data on security facilities
ignitions in the past will continue to experience the
(e.g., hospitals, police stations, emergency
highest relative ignition rates, areas most likely to
operations centres), and data on access (e.g., roads,
experience future wildfire ignitions can be identified
bridges, tunnels, railways) are also required,
(NEUENSCHWANDER et al. 2000).
wherever possible.
The importance of assessing population vulnerability
The concepts previously expressed can be
is particularly relevant when we have to face the problem
usually applicable to a local scale, as a consequence
of wildland-urban interface (WUI) areas.
of the great amount of required input data and their
very refined spatial resolution. Wildland-urban interface in the European environment
and more specifically in the Mediterranean area is a very
The possibility to obtain a refined result is typically
complex spatial context with many interrelated social,
dependent on the chosen working scale and is strictly
natural resource and wildfire issues.
related to the availability and the resolution of data.
The problem of wildfires in the wildland-urban
The scale and resolution of the data that can be
interface has recently become quite relevant because of
used (typically available at the state scale) limit the
the increasing number of dwellings near to or inside
detail that can be produced.
natural areas and of the increasing number of wildfires
This choice can be conflicting with the desire of involving these sites.
scientists to be precise, and the needs of land
As previously underlined, Geographic Information
managers to identify specific project activities and
Systems (GIS) can be useful tools for WUI management,
locations (SAMPSON and NEUENSCHWANDER, 2000),
through their capability of handling in an integrated
but is consistent with the objective of producing a
environment multi source and multi resolution spatial data
strategic assessment, at least concerning population
At the 1:1,000,000 scale, corresponding to the
one selected to develop the Euro-Mediterranean
Wildland Fire Danger Rating System, the social data
sets above all are necessarily not detailed enough to
allow us to quantify the complete suite of social
effects we know can occur.

D-08-06.doc 71

Italy is divided into 20 Regions, characterized by deep

9.1.2 Example of population vulnerability
environmental differences; the study was applied in
mapping in Piemonte Region (Italy)
Piemonte, a Region located in the north-western part of
As an example of GIS analyses that can be Italy, with a surface of about 2,5 million hectares.
performed in WUI areas to assess wildfire risk and It has a quite heterogeneous territory, characterized
vulnerability see CAMIA et al. (2004). by plains mainly interested by agriculture, hilly areas and
The analyses described in this work were mountains among the highest of Europe.
performed in the framework of WARM Project.
There are 1.209 municipalities, whose population
They were firstly aimed at identifying and typifying
ranges from about 1.000.000 inhabitants to less than 100
settlements, providing a first description of the
surrounding environment and of fire suppression
In the country areas the secondary houses are quite
frequent and their number has been strongly increasing in
As previously mentioned, the study area is the last years.
characterized by small urban settlements, presenting
The fire season is typically a winter-early spring one,
both situations of classical interface and intermixes.
which is mainly characterized by surface fires.
Actually, an entire set of conditions, ranging from
Crown fires are mostly due to the presence of
the densely inhabited village with boundaries facing
evergreen coniferous stands, while most forests are
wildland and forests until the isolated houses are
made up by broadleaved species.
Most wildfires break out and spread in the mountain
The authors wanted to identify and classify the
belt from 200 m to 1000 m a.s.l..
different situations according to the degree of
urbanization within the wildland area, and therefore Data about wildfires are collected and recorded with a
the general wildland-urban mixing conditions for the spatial resolution of 1-km2 (i.e. on a UTM kilometric grid
different settlements, being clear the usefulness of with square cells of 1000 m size).
such a characterization from a fire management point For each wildfire, the cell where the fire started is
of view. recorded, while the adjacent cells that could have been
This task was approached using density analysis interested by the fire spread are not reported.
techniques to the house layer of the database, and
Only recently, information on fire perimeter was
applying thresholds to the derived house density
introduced among data collected following a fire event.
layer and identify in this way settlement boundaries.
Settlement types were then characterized by Among the available data there is no information
computing the number of houses in each polygon concerning possible infrastructures damaged or
discriminated, performing a spatial join between the threatened by fire.
settlements and the houses layers. WUI wildfires are therefore not identifiable from
The presence of specific fuel types, as well as the historical series.
topographic arrangement of the area, provides useful
Considering these constraints due to data deficiency
indications about the fire hazard and risk conditions
around the settlements, but also about the possibility or resolution, a methodology that analyses past events
and strategies for fire fighting resources. examining and assessing their territorial background was
In addition to such key features, from a fire
suppression point of view, it is important to know the The geographical units of the analysis correspond
accessibility of the settlements to be protected, as necessarily to 1-km² cells.
well as the relative location of fighting resources (fire
stations) and water supplies. To locate WUI fire prone areas in Piemonte territorial
elements that contribute to create the WUI environment
Although a road density map can provide useful were identified.
information on the accessibility of the sites, much
more informative is the map derived from the network Areas where natural vegetation and infrastructures are
analysis that allows to classify each settlement both present and intermingle were looked for using and
according to the time needed to reach it from the overlaying layers containing urban areas and woods, with
nearest fire station or the time from the nearest water predefined criteria.
supply. Thus, working mainly with GIS facilities, proper layers
Working at a global scale within a 1-km² grid, an derived from wildfires, urban areas and forest databases
example of what can be obtained with simple and were overlaid applying thresholds and neighboring
necessarily coarse data, is reported in Camia et al. constraints in order to extract those cells that are
(2002). expected to contain WUI environment prone to wildfires.

In this work, a methodology was developed to Data processing that eventually led to the selection of
analyse the problem of WUI wildfires and to territorial contexts where WUI areas are potentially
characterize its spatial distribution at the scale of the threatened by wildfires, allowed to assess the size of the
Italian Regions. problem at regional scale, its spatial distribution and to
find the most affected areas.

D-08-06.doc 72

The first GIS layer analysed is represented by These two values of burned surface were chosen
wildfires data. since 30 ha correspond to the burned surface threshold
An historical series of 20 years, from 1980 to that typifies a critical fire according to the Regional Fire
1999, was considered. Management Plan (BOVIO et al., 1999), while 15 ha have
From these data the number of wildfires occurred been prudentially added as an in-between attention
in this period within each grid cell was obtained. threshold for WUI fires in Piemonte.
With this respect, it is important to underline once
On the base of WUI wildfires frequency, a zoning of
again that it is possible to know only the cell in which
municipalities has been realized.
the wildfire started and not the neighboring cells
Piemonte municipalities were classified, identifying
involved, if any.
5 frequency classes, according to the number of wildfires
To include also those cells that, although not
potentially interesting WUI areas between 1980 and
directly recorded as sites where fire actually started,
were expected to be affected by fires started in
neighbouring cells, some further criteria were 9.1.3 Estimation of population vulnerability at Euro-
introduced. Mediterranean scale
The second input layer is represented by the Vulnerability is not directly measurable.
territory urbanization. However, it is possible to induce vulnerability using
Urban areas have been extracted from the land information about the characteristics of the geographical
use map (1:100,000) of Piemonte Region zone considered.
These areas were intersected with the cells As previously underlined, to realize a Wildland Fire
Danger Rating System operative at an Euro-
containing wildfires in order to select those cells
characterized both by houses and by wildfires. Mediterranean scale, we must face the problem of using
In this step all cells containing exclusively urban appropriate data at relevant scale and precision for the
expected aim, currently available for all the countries
area and not burned wildland were excluded.
In this way, starting from all wildfires occurred in involved.
Piemonte from 1980 to 1999, cells with wildfires Thus the question is: What do we have and how
occurred next to urbanized areas were identified. reliable is it?
As mentioned, wildfire starting within cells with no The difficulty when trying to approach the risk at global
urban area but adjacent to cells with urban area scale is to find relevant and available indicators allowing
presents a potential menace. a comparison between all countries.
Therefore, we added to the previous selection,
cells less than 1,000 m away from other cells Simple indices relying on good data and with stated
characterized by the presence of urban areas and limitation and subjectivity might be much more efficient
with more than 3 wildfires or more than 10 ha of than complex ones that cannot be computed because of
burned surface in the 20 considered years. the lack of (reliable) data.
This step allows to include not only directly Extrapolations from local researches to global scale
affected areas, but also those urban areas are rarely applicable as data may not be of comparable
considered potentially prone to WUI wildfires. formats or simply not available.
If a model requests a large amount of inputs, the
The third input layer is related with forests. chances that such model will never be used, by lack of
The forested surface of each cell was calculated data or by too fuzzy data, are significant.
from the land use map, selecting between the On the other hand, a model based on too few
previously extracted cells only those typified by a parameters will lead to large gap between observed facts
forestry cover higher than 5%. and expected figures.
Considering starting data and analysis scale, it Since no more detailed and precise data are currently
isn’t possible to assert these wildfires directly available and/or easily accessible for all the Euro-
interfered with human dwellings, but just that they Mediterranean countries, it was decided to work with the
occurred in environments characterized by the CLC2000 geographic data layer (EC JRC-IES, 2005).
presence of WUI areas.
CORINE land cover (CLC) is a geographic land
To characterize areas where WUI wildfires are cover/land use database encompassing most of the
particularly critical, all cells between those previously countries of the European Community and the majority of
selected, in which events characterized by a burned the Central and East European countries (Figure 72).
surface higher than 30 ha and 15 ha occurred, have
been identified. CLC 2000 is the year 2000 update of the first CLC
database, which was finalised, in the early 1990s as part
of the European Commission programme to COoRdinate
INformation on the Environment (Corine).
It also provides consistent information on land cover
changes during the past decade across Europe.
The CLC2000 database currently covers 32 countries.

D-08-06.doc 73

CLC describes land cover (and partly land use) SUB-INDEX 1
according to a nomenclature of 44 classes organized
hierarchically in 3 different levels (Figure 73). This sub-index qualifies each grid cell according to the
The first level (5 classes) corresponds to the main presence within the cell itself of urban land covers and
categories of the land cover/land use (artificial areas, potentially combustible ones. The presence of these land
agricultural land, forests and semi-natural areas, covers is quantified with this sub-index independently
wetlands, water surfaces). from their spatial location and topology within the cell.
The second level (15 classes) covers physical and
physiognomic entities at a higher level of detail (e.g. Third level CLC2000 classes were used.
urban zones, forests, lakes), while level 3 is The surface occupied inside the cell by the following
characterized by 44 more detailed classes. land cover classes is calculated in a GIS environment:
The nomenclature has been developed in order to - Areas occupied by forests and semi-natural areas,
map the whole Community territory, including the shrubs and herbaceous plants; recently burned areas
foreseen extension to other eligible countries. are also included (CLC2000 codes: 2.4.4; 3.1.x; 3.2.2,
The use of the CLC nomenclature with 44 classes 3.2.3, 3.2.4, 3.3.4 ; see Table 16);
at three hierarchical levels is mandatory. - Areas characterized by a major presence of houses
Additional national levels can be mapped but and consequently population (CLC2000 codes: 1.1.x;
should be aggregated to level 3 for the European 1.2.x; see Table 16);
data integration.
No unclassified areas should appear in the final This sub-index is meant to take into account those
version of the data set. situations characterized by a significant presence of both
combustible vegetation and population potentially at risk.
CLC was elaborated based on the visual
interpretation of satellite images. The total surface occupied within each cell by the two
Ancillary data (aerial photographs, topographic land cover categories (a and b) is then classified in ten
and vegetation maps, statistics, local knowledge) percentage classes.
were used to refine interpretation and the assignment The value of the sub-index, going from 0 to 5, is
of the territory into the categories of the CORINE land obtained through a bi-dimensional matrix (Figure 74),
cover nomenclature. combining the classes of the two land cover categories
present in a cell.
The smallest surfaces mapped (mapping units)
correspond to 25 hectares. In the matrix higher values are assigned to those cells
Only area elements (polygons) are identified. in which there is a major presence of both urban land
Areas smaller than 25 ha are allowed in the covers and potentially combustible ones, in order to
national land cover database as additional thematic identify those situations characterized by a higher
layers, but should be aggregated/generalized in the probability to have wildland-urban interface areas or by a
European database. higher probability to have a widespread human presence,
independently from the precise location of settlements.
Linear features less than 100 m in width are not
considered SUB-INDEX 2
The scale of the output product was fixed at This sub-index is based on the assessment of the
1:100.000. spatial configuration of specific land covers in the cell.
Thus the location precision of the CLC database The linear development of those zones in which urban
is 100 m. land covers are in contact with potentially combustible
From the CLC2000 layer, information on both land covers is measured.
potentially combustible land covers and populated Considering the scale and precision of CLC2000 data,
ones could be derived, providing a common base for this contact is more probably to be considered as a
all the involved countries, from which to indirectly neighbourhood relationship between the two categories
obtain information on population vulnerability to of land cover.
wildfires. The land covers selected for this analysis are:
An index to assess population vulnerability was - Areas occupied by forests and semi-natural areas,
defined, made up by three main sub-indices. shrubs and herbaceous plants; recently burned areas
These three sub-indices are calculated on a are also included (CLC2000 codes: 2.4.4; 3.1.x; 3.2.2,
1-km2 grid and their values within each cell are 3.2.3, 3.2.4, 3.3.4 ; see Table 16);
summarized to give the final index. - Areas characterized by urban fabric, continuous or
discontinuous (CLC2000 codes: 1.1.x; see Table 16);
The main aim of this sub-index is to identify those
situations in which, given a certain surface occupied
within a cell by the two categories of urban and potentially
combustible land uses, as assessed by sub-index 1,
there is a high probability that people, buildings or
infrastructure are in proximity of fuels susceptible to burn.

D-08-06.doc 74

Please note that the susceptibility and fire-prone SUB-INDEX 3
nature of these fuels should be assessed though
other components of the Euro-Mediterranean Cells with a certain level of possible vulnerability, as
Wildland Fire Risk Index proposed in this deliverable. resulting from the previous steps, are then further
As a consequence, the effective population differentiated according to the presence of a particular
exposition to fire risk comes from the integration of urban land cover, i.e. the “discontinuous urban fabric”
the Vulnerability index here discussed and the (CLC2000 code 1.1.2). This land cover identifies areas
Ignition and Propagation Danger indices previously where most of the land (between 30 to 80 %) is covered
discussed. by structures; building, roads and artificially surfaced
The sub-index value is assigned according to the areas are however associated with vegetated areas and
linear development of probable WUI areas measured bare soil, which occupy discontinuous but significant
in the cells and classified following Table 17. surfaces. This land cover, frequently corresponding to the
external part of settlements, is considered a more
Only those cells having at least one of the two vulnerable urban cover because of the concurrent
above described sub-indices different from 0 are presence of structures and artificial green areas or
selected to proceed with the analyses. wildlands, often associated in a complex spatial pattern.
This means that cells with no or little presence of
urban and potentially combustible land uses and cells The surface occupied by discontinuous urban fabric
with no or little linear development of WUI areas are inside cells previously selected is then calculated,
considered as not vulnerable areas, given the low assigning to each cell a value corresponding to a surface
detail of initial data and the working scale. class as defined in Table 18.
The final vulnerability index was obtained as the sum
of the three sub-indices in the cell, ranging from a
minimum of 1 to a maximum of 14.
Three vulnerability classes were identified, as reported
in Table 19.
An example of each sub-index and of the final
population vulnerability index, calculated according to the
methodology above described, is reported in Figure 75.
The described methodology was applied to the Italian
province of Torino (North-Western Italy, Piemonte
Region), corresponding to a NUTS 3 level. The CLC2000
data layer of the province and the DTM are shown in
Figure 76.
Resulting values of sub-index 1, sub-index 2, sub-
index 3 and the final vulnerability index are instead
reported in Figure 77.

D-08-06.doc 75

9.1.4 Figures

Figure 72: Example of CLC2000 geographic data layer. Yellow lines correspond to the boundaries of CLC classes
(see figure 73 for CLC nomenclature); red lines correspond to 1-km2 grid cells.

Figure 73: CLC2000 nomenclature (from EC JRC-IES, 2005)

D-08-06.doc 76

urban areas
1 2 3 4 5 6 7 8 9 10

91 - 100
11 - 20

21 - 30

31 - 40

41 - 50

51 - 60

61 - 70

71 - 80

81 - 90
0 - 10
surface %

0 - 10 0 0 0 0 0 0 0 0 0 0

11 - 20 0 1 1 1 2 2 2 1 1

21 - 30 0 1 2 2 2 3 3 3

combustible areas
31 - 40 0 2 2 3 3 4 4

41 - 50 0 2 2 3 4 5

51 - 60 0 2 3 4 5

61 - 70 0 2 3 4

71 - 80 0 3 4
81 - 90 0 3

91 - 100 0

Figure 74: Bi-dimensional matrix to determine sub-index 1 values within each 1-km² grid cell.

Figure 75: Example of each sub-index and of the final population vulnerability index calculated according to the
methodology here described

D-08-06.doc 77

Figure 76: DTM and CLC2000 data layer of the Italian province of Torino (North-Western Italy), where the
population vulnerability index here proposed was tested

D-08-06.doc 78

Figure 23: Sub-index 1, sub-index 2, sub-index 3 and the final population vulnerability index computed for the
province of Torino (Italy)

D-08-06.doc 79

9.1.5 Tables

Table 16a: Corine land cover nomenclature of the classes selected for the analyses (Populated Land Covers)
CLC code CLC description
Areas mainly occupied by dwellings and buildings used by administrative/public utilities or
authorities, including their connected areas (associated lands, approach road network,
1.1.1 Continuous urban fabric: Most of the land is covered by structures and the transport
1.1 Urban fabric network. Building, roads and artificially surfaced areas cover more than 80 % of the total
surface. Non-linear areas of vegetation and bare soil are exceptional.
1.1.2 Discontinuous urban fabric: Most of the land (Between 30 to 80 %) is covered by
structures. Building, roads and artificially surfaced areas associated with vegetated areas
and bare soil, which occupy discontinuous but significant surfaces.
Areas mainly occupied by industrial activities of transformation and manufacturing, trade,
financial activities and services, transport infrastructures for road traffic and rail networks,
airport installations, river and sea port installations, including their associated lands and
access infrastructures. Includes industrial livestock rearing facilities.
1.2.1 Industrial or commercial units
Artificially surfaced areas (with concrete, asphalt, tarmacadam, or stabilised, e.g. beaten
earth) without vegetation occupy most of the area, which also contains buildings and/or
1.2 Industrial,
commercial and Particularity of class 1.2.1: Agricultural farms: Areas of other than housing buildings, in-
transport units door spaces, stables, garages, workshops, lay-by and storing areas, often also bad land
with ruderal vegetation, part of farms. The farms are often located in outskirts or close to
rural settlements with agricultural function. Concentration of agricultural buildings in areas
of various sizes was associated with collectivisation of agriculture. The quoted areas
smaller than 25 ha are included in class 1.1.2.
1.2.3 Port areas: Infrastructure of port areas
including quays, dockyards and marinas.
1.2.4 Airports
Airports installations: runways, buildings and associated land.
1.4.1 Green urban areas:
1.4 Artificial, non-
Areas with vegetation within urban fabric, includes parks and cemeteries with vegetation,
and mansions and their grounds
areas 1.4.2 Sport and leisure facilities: Camping grounds, sports grounds, leisure parks, golf
courses, racecourses, etc. Includes formal parks not surrounded by urban areas.
2.4.2 Complex cultivation patterns: Juxtaposition of small parcels of diverse annual crops,
pasture and/or permanent crops.
2.4 Heterogeneous Particularity of class 2.4.2: Complex cultivation patterns with scattered houses Alternation
agricultural areas of small plots (smaller than 25 ha) of arable land with annual or permanent crops with
scattered garden huts or scattered houses. They are usually situated in proximity of rural or
urban settlements and are used for growing agricultural crops, fruit, and vegetable for the
particular households.
Areas occupied by forests and woodlands with a vegetation pattern composed of native or
exotic coniferous and/or deciduous trees and which can be used for the production of
timber or other forest products. The forest trees are under normal climatic conditions higher
than 5 m with a canopy closure of 30% at least. In case of young plantation, the minimum
cut-off-point is 500 subjects by ha.
3.1 Forests 3.1.1 Broad-leaved forest: Vegetation formation composed principally of trees, including
shrub and bush understoreys, where broad-leaved species predominate
3.1.2 Coniferous forest: Vegetation formation composed principaly of trees, including shrub
and bush understoreys, where coniferous species predominate.
3.1.3 Mixed forest: Vegetation formation composed principally of trees, including shrub and
bush understoreys, where neither broad-leaved nor coniferous species predominate.

D-08-06.doc 80

Table 16b: Corine land cover nomenclature of the classes selected for the analyses (Combustible Land Covers)
CLC code CLC description
3.2.2 Moors and heathland: Vegetation with low and closed cover, dominated by bushes,
shrubs and herbaceous plants (heather, briars, broom, gorse, laburnum, etc.)
3.2 Shrubs and/or 3.2.3 Sclerophylous vegetation: Bushy sclerophyllous vegetation, includes maquis and
herbaceous garrigue. In case of shrub vegetation areas composed of sclerophyllous species such as
vegetation Juniperus oxycedrus and heathland species such as Buxus spp. or Ostrya carpinifolia with
associations no visible dominance (each species occupy about 50% of the area), priority will be given to
sclerophyllous vegetation and the whole area will be assigned class 3.2.3.
3.2.4 Transitional woodland/shrub: Bushy or herbaceous vegetation with scattered trees.
Can represent either woodland degradation or forest regeneration/recolonisation.
Natural areas covered with little or no vegetation, including open thermophile formations of
sandy or rocky grounds distributed on calcareous or siliceous soils frequently disturbed by
erosion, steppic grasslands, perennial steppe-like grasslands, meso- and thermo-
3.3 Open spaces
Mediterranean xerophile, mostly open, short-grass perennial grasslands, alpha steppes,
with little or no
vegetated or sparsely vegetated areas of stones on steep slopes, screes, cliffs, rock fares,
limestone pavements with plant communities colonising their tracks, perpetual snow and
ice, in land sand-dune, coastal sand-dunes and burnt areas.
3.3.4 Burnt areas: areas affected by recent fires, still mainly black.

Table 17: Sub-index 2 values according to the linear development of probable WUI areas (m). Natural breaks
(Jenks) were selected to identify the six classes corresponding to the index values.
Sub-index 2 Linear development of probable WUI areas (m)
0 0
1 1-500
2 501-1000
3 1001-1700
4 1701-2700
5 2700-4700

Table 18: Sub-index 3 values according to the surface occupied in the cell by CLC2000 code 1.1.2.
Natural breaks (Jenks) were selected to identify the classes corresponding to the index values.
Sub-index 3 CLC 112 surface (m2)
0 0
1 1-62.000
2 62.001 -200.000
3 200.001 - 380.000
4 380.001 - 650.000
5 > 650.000

Table 19: Final vulnerability index values and corresponding vulnerability classes.
Classes Vulnerability index
Low 1-5
Medium 6-10
High 11-14

D-08-06.doc 81

9.2 VULNERABILITY RELATED TO The study of landscape structure/composition by

ENVIRONMENTAL VALUE means of spatial statistics helps us characterize the
territory and understand the spatial-temporal
9.2.1 Presentation
relationship among the different elements that compose
Ecological and landscape assessments are a landscape.
essential in the design of an integrated fire risk index Landscape ecology has developed many indices
which takes into account aspects related to the that measure spatial textures and shapes, as well as
vulnerability of forested areas. the spatial structure of the landscape (MCGARIGAL et al.
Such assessments are of great value in sustainable 2002; RIITTERS et al. 1995; MCGARIGAL y MARKS, 1995):
land management and help managers in decision patch density, size, compactness, fractal dimension,
making processes within a multipurpose scenario, dispersion, diversity, etc.
ranging from natural environment conservation policies Metric assessment and analysis of landscapes,
and biodiversity promotion strategies to urban and using a GIS (MCGARIGAL et al. 2002; BERRY, et al.
regional planning. 1999; IGIS, 1997; MCGARIGAL y MARKS, 1995), allows
For instance, given the fact that the budget available to characterise the structures and changes in forested
for preventive silviculture is usually limited or land occupation and land use within a territory, and
insufficient, is it reasonable to apply these tasks to also helps us find the environmental involvement of its
areas with the highest ecological and landscape value. activity (MALDENOFF y BAKER, 2000).
These areas are, therefore, the most vulnerable to
The studies carried out by Martinez et al., 2006
this line of events.
propose a simple method to map the landscape and
In addition, when a forest fire is spreading, it would
ecological value of an area, by integrating several
also seem reasonable for a fire manager to allocate fire
indicators in an exclusively ecological approach,
suppression resources according to a certain scale of
without considering visual or aesthetic aspects.
The results show that the methodology can be
This scale would bear in mind the presence of
usefully applied on a regional/global scale.
population, property (buildings, etc.) and also
In addition, it is a valuable source of information for
outstanding natural environments, such as natural
fire fighting authorities in decision making processes
parks, reserves, special protection zones, preserved
related to preventive forestry and to the distribution of
landscapes, etc.
extinction resources and manpower when a forest fire
In some previous studies, this problem has been is spreading and threatening the areas which are most
addressed by using GIS tools. vulnerable from an ecological point of view.
However, in some cases (Comunidad de Madrid,
We consider it would interesting to include the
2000) the only parameter which is taken into account to
landscape assessment methodology proposed by
measure priorities in forest fire fighting, is the ecological
these authors in the section concerning the vulnerability
value of the area according to the quality and the
of an integrated risk index on a European level.
degree of protection of the forested ecosystems.
The methodology starts by integrating different
There is no assessment of the landscape and no
indicators, under an exclusively ecological approach,
landscape ecology indices are applied. In other cases
and seeks to work out the intrinsic ecological value of a
(GULINCK et al. 2001), the method used to asses a
given landscape is based on land use data, which does
Figure 78 shows the methodological flow chart
not include information from other supplementary data
proposed by Martínez et al. (2006):
The main data source used to obtain the landscape
In this context, Landscape Ecology provides the
value is the land cover map from the CORINE-Land
appropriate conceptual framework that meets the
Cover project from the year 2000 (CLC2000) available
demand for information which regional planners
in digital format for all of Europe.
Two criteria groups are considered.
It is well known that landscape ecology is a
multidisciplinary science which mainly aims at solving i) Vegetation and land use
the problems to do with land management and
The first group of indicators is based on the intrinsic
development on a local and regional scale (NAVEH and
characteristics of vegetation, such as the degree of
proximity to climax vegetation.
This new science facilitates territorial analysis by
In addition, it analyses the importance of vegetation
attempting to understand and compare different spatial
types in its regional context (rarity) and in its global
patterns by means of patches with different shapes,
context (representativity).
quantities, classes, etc. (HONG et al. 2000; RIITTERS et
al. 1995). ii) Landscape Ecology Indicators
Landscape is under the influence of ecological and As well as the variables from the previous group,
human processes, both on a regional and local scale, other landscape ecology indicators are used, especially
which imprint changes in its structure and composition. indices related to diversity, connectance and
juxtaposition or interspersion, which take into account
the spatial distribution of the patches within a territory.

D-08-06.doc 82

MARTINEZ et al. (2006) propose a selection of Values range from 0 (consisting in just one match or
indices (Simpson's Diversity Index, Interspersion when none of the matches are connected) to 100,
Juxtaposition Index and the Connectance Index) with when all the patches in the analysis window are
which to get to know the diversity, homogeneity or connected.
fragmentation of a given landscape.
The Landscape Integrated Value (LIV) is the simple
Simpson’s Diversity Index (SIDI) assesses the average of the combination of the six components
number of different land cover types and the mentioned.
distribution of the area proportional the types of land An example of this product can be seen in figure 79.
Regarding the ecological value, the authors suggest
Values near 0 indicate that there is only one patch
considering whether the territory analysed is inside a
or land cover (non diversity) whereas values close to 1
preserved area of some kind: protected natural
(high diversity) indicate that the different types of land
environments, special protection areas for birds, sites
covers as well as the proportional distribution of their
and habitats of community interest, preserved
areas is more balanced.
woodlands and public woodlands, etc.
The Interspersion Juxtaposition Index (IJI) allows us This information can be easily accessed on a
to understand the spatial configuration of the patches national/European scale.
as well as its contiguity and its degree of interspersion.
In view of the results obtained in this work and
Low values (0) represent landscapes where patches
although the workload capacity of GIS could include
are distributed randomly, whereas high values (100)
other elements for improving landscape and ecological
correspond to landscapes which are distributed in an
assessment, we believe that the methodology
equal contiguity.
proposed by the authors is straightforward and easy to
This index shows whether patches are grouped or
distributed homogeneously in space.
In addition, it has basic data requirements, available
The Connectance Index (CI) is defined by the on a European scale, that guarantee the applicability of
number of functional unions among all the patches the method and its integration, together with other
within the same type. factors, into a synthetic fire risk index.
It show where each pair of patches are connected
according to a 500 meter analysis window.
9.2.2 Figures

Figure 78: Landscape assessment methodological flow chart

D-08-06.doc 83

Figure 79: Landscape integrated value map for the Madrid region (MARTÍNEZ et al., 2006)

D-08-06.doc 84

9.3 POTENTIAL SOIL EROSION Research indicates that fires are likely to increase
runoff rates and sediment yields relative to undisturbed
9.3.1 Introduction
forested land (AHLGREN and AHLGREN, 1960; LAIRD and
Soil erosion is a widespread threat to European HARVEY, 1986; May, 1990; SCOTT and VAN WYK, 1990;
soils. SOTO et al., 1991; SOLER and SALA, 1992; SWANSON,
The main consequences of erosion are not only on- 1981).
site soil degradation, declining fertility, desertification
A great number of pan-European soil erosion risk
and reduced infiltration and water storage capacities,
assessment efforts have been made with a variety of
but there are also off-site impacts that include
approaches such as GLASOD approach (Global
eutrophication of rivers and lakes, destruction of wildlife
Assessment of the Current Status of Human-Induced
habitats, siltation of dams, reservoirs, rivers, as well as
Soil Degradation), INRA (Institut National de la
infrastructure and property damage by muddy floods
Recherche Agronomique, 1988) approach, HOT-
(RUBIO et al., 2006).
Some authors have highlighted its impact on global
food security (CROSSON, 1997; LOMBORG, 2001).
Widely used models include RUSLE (Renard et al.,
In addition, soil erosion results in emission of soil
1997) and MUSLE (SMITH et al., 1984).
organic carbon to the atmosphere in the form of CO2
Another model, Soil Erosion Model for
and CH4, thereby enhancing global warming (LAL,
Mediterranean regions (SEMMED, DE JONG, 1994a)
was developed with the objective of using satellite data,
Global warming in turn is expected to increase
similarly to the Thornes model (THORNES, 1985),
erosion rates (NEARING et al., 2004).
Agricultural Nonpoint Source Pollution model (AGNPS:
European Research Programmes initiated by the
YOUNG et al., 1989) and the Areal Nonpoint Source
European Commission demonstrate that although the
Watershed Environment Response Simulation model
Mediterranean region is historically the most severely
(ANSWERS: BEASLEY et al., 1980).
affected by erosion, there is growing evidence of
Furthermore, important European policies and
significant erosion occurring in other parts of Europe
directives, such as the Water Framework Directive (EC,
(e.g. Austria, Czech Republic and the loess belt of
2000), the European Commission Soil Thematic
Northern France and Belgium).
Strategy (COM, 2002), and instruments of the Common
Soil erosion can therefore be considered, with
Agricultural Policy, such as agro-environmental
different levels of severity, an EU-wide problem
measures (EC, 1999), address the issues of soil
(MONTANARELLA et al., 2003).
Soil erosion occurs under diverse conditions and is
driven by interactions of many factors such as land use, Although there are many approaches in relation to
climate, soil conditions, and topography that are difficult soil erosion risk assessment, only a few can be found
to quantify. in the literature that take into account the effects of
A proper assessment of erosion problems is greatly wildfires on vegetation and soil conditions and are
dependent on their spatial, economic, environmental, suitable for burned woodland areas.
and cultural context (WARREN, 2002). Most models and strategies address soil erosion
As a result, a comprehensive understanding of soil problems in agricultural areas only.
erosion is still very difficult.
9.3.2 Objectives
Forest fires are a further factor that can lead to
In accordance with the aim of this deliverable, a
excessively high rates of erosion and contribute to
vulnerability index of soil erosion will be created for
environmental degradation.
Mediterranean countries that suffer high intensity and
Every year there are more than 50,000 forest fires in
frequency of forest fires.
Europe, affecting over 500,000 ha of forest and other
The questions and objectives that were set
woodlands, the majority of which occurs in the
regarding potential soil erosion in the previous
Mediterranean Region (European Commission, 2002).
deliverable (D-08-03) are the following:
The continuous impacts of fires, with progressively
- To what extent can planners prepare for the
shorter recovery periods, together with torrential rains
consequences of a forest fire before it occurs in
characteristic of the Mediterranean climate, favour the
order to respond more quickly to the potential runoff
intensification of erosive processes (GIOVANNINI et al.,
and erosion risks
- To elaborate a pre-fire strategy that would shorten
Wildfires and prescribed fires affect the vegetation,
the post-fire reaction time
soils, wildlife, and water resources of watersheds.
- To evaluate the use of the model quoted in D-08-03
They impose a wide range of effects which depend
(ANNEX I) at different spatial scales and to answer
on the mosaic of fire severity and post-fire hydrologic
how model scale affects the objectives and the
events (NEARY, 2004).
operational use of the model.

D-08-06.doc 85

As a result, this method has to be modified in order

From the above questions, two approaches can be
to meet the following needs:
put forward.
- to be applied in pre-fire time and thus help us to
In the first, a “what if” logic underlies the model in
acquire an ‘a priori’ knowledge of the most soil
order to define potential erosion risk areas.
erosion prone areas
The underlying question is therefore “what would be
- to provide European scale information of the
the post-fire soil erosion risk if a forest fire were to take
potential erosion risk.
place? (sub-chapter 9.3.3)”.
In this case, scenarios can be tested at the The way the four erosion factors were estimated
European scale to define potential soil erosion risk. and how they can now be addressed in order to predict
The focus here is on gaining a spatial perspective of potential soil erosion at the European scale in pre-fire
the distribution of high-risk zones. time is discussed below.
In a second approach, an operational model is Slope
elaborated to be used in a Mediterranean context in the
case of an actual fire. This factor can be considered static since it does not
Here, the question is “how can Mediterranean change over time.
countries best respond to a fire in order to deal A 25-m DEM was used for slope estimation, and a
efficiently with the erosive consequences? (sub-chapter 50-m resolution was considered too coarse to be
9.3.4)”. operational.
9.3.3 A European scale model to predict potential At the European scale, 30 to 50-m DEMs are
erosion risk available and are considered sufficient of the objectives
of this approach.
The aim of this approach is to perform a pre-fire Slope needs to be estimated at the European scale
prediction of post-fire soil erosion risk. only once since it does not change significantly over
Post fire soil erosion risk is strongly related to the time.
effects of fire on vegetation and soil conditions. FOX et al. (2006)formed five categories, according to
In other words, before a fire actually occurs, we slope magnitude and coefficients were ascribed to each
must firstly predict the kind and intensity of fire effects category as follows:
on soil properties and vegetation; secondly, we must be
able to estimate the consequences of these changes Pre-fire Vegetation density
on soil erosion rates.
This factor was estimated from the standing charred
One approach to addressing this problem is to trunks during a field survey and three categories were
modify the method quoted in D-08-03 deliverable included.
(ANNEX I) in order to be able to acquire information Coefficients were ascribed to each factor level as
about the post-fire erosion risk that may arise before a follows table 20:
Wildfire takes place.
This ‘a priori’ knowledge would give European level It is obvious that at the European scale, vegetation
data cannot be collected manually.
planners a vision of the spatial distribution of post-fire
erosion risk and help in the preparation of strategies at However, CORINE land cover or remote sensing
the European program scale to deal with this potential data could be used to estimate pre-fire vegetation.
Vegetation cover provides protection of the soil
against erosion processes. Discussion and suggestions about the This happens not only before fire (INBAR et al.,
method described in ANNEX I 1997), but also after fire because in low intensity and
medium intensity fires the surface organic matter is
The method quoted in D-08-03 deliverable takes increased providing protection from soil erosion (fallen
into account four factors to produce a map of soil needles, twigs, etc).
erosion risk at the catchments scale (FOX et al., 2006). In addition to the mulch effect provided by fallen
These factors are: pine needles (SHAKESBY et al., 1993; SHAKESBY et al.,
- Slope 1994), standing dead vegetation may favour infiltration
- Pre-fire vegetation (near stems and in burned roots) and slow runoff
- Fire severity velocity, so the net expected effect is lower erosion
- Soil erodibility where vegetation stands are denser.
These factors were estimated after a wildfire in the ANDREU et al. 1996 demonstrated that organic
Massif des Maures area near St Tropez in southern matter content strongly decreases immediately after
France which means that this method poses temporal medium to intense fires, with a subsequent increase
difficulties in the typical Mediterranean climatic context with time.
where there is a limited time between a summer or However, in zones that suffered a low to medium
autumn fire and the following rainy winter period. fire intensity its value increases.
This could be due to the accumulation of plant
residues, not completely burnt, according to the studies

D-08-06.doc 86

It was found from the literature that for land cover Fire severity
estimation, the following methods were applied:
- in the IMAGE/RIVM approach, Olson’s land cover It must be defined as potential fire severity.
database and statistical information from FAO was
used Fire severity was defined as either variable or
- in the CORINE approach, the CORINE landcover intense and coefficients were ascribed to each level as
database was used following: Table 22
- in the USLE/ESB approach, an exponential scaling “Variable severity” were sites where either the fire
function was linked to the Normalized Difference had passed quickly and charred the trunks and
Vegetation Index (NDVI) extracted from NOAA underbrush, but left the crown more or less intact, or
images the intense combustion of the standing vegetation was
- in the PESERA approach, a plant growth model was confined to a small area (dozens of m2).
used initially, but in monitoring mode it was replaced The “intense severity” category consisted of sites
by NDVI from Earth Observation Sensors NOAA- where the fire had consumed the underbrush and
AVHRR or SPOT-VGT (also applied by Zhang et al., crown entirely over an extensive area.
2004) According to this classification, this factor could also
- an NDVI (TUCKER et al., 1985) was also extracted be defined as fire type because the former category
from Landsat TM data (Bayaramin et al., 2006) in corresponds more or less to the description of a
the semi-arid area of Beypazari in Ankara in order to surface fire, whereas the latter to the description of a
create a layer of vegetation cover which was used in crown fire. In order to estimate fire severity, the main
a study of soil erosion risk assessment. NDVI was factors contributing to the outbreak of a crown fire
grouped as “fully protected” or “not fully protected” should be taken into account.
based on ground truth information with the global
positioning system. The map created can be seen The primary factors influencing crown fuel ignition
further down (Figure 80). The pixel size from NOAA- are:
AVHRR is of 1 km spatial resolution so in order to - Height of the under-story vegetation and
improve the spatial resolution, LANDSAT remote - The vertical distance between the ground/surface
sensing data could be used in local scale to fuel strata and the lower boundary of the crown fuel
calculate Vegetation Indices (VIEDMA et al., 1997: layer (Miguel et al., 2006).
Santos et al., 2000: DELGADO et al., 2001) in which These two factors are the most significant for
pixel size is 30 m. This can easily be used with a predicting danger of crown fire ignition.
30-m DEM. Although, they cannot be estimated easily at the
Resolution of 30 by 30-m pixel size can be European scale from remote sensing data, because
considered appropriate because of the great fluctuation methods that can be applied for their estimation, are
in vegetation density that can be found inside a 1 by 1 usually, if not always, by means of LIDAR data
km pixel size of NOAA-AVHRR. (ANDERSEN et al., 2005: DRAKEA et al., 2002:
MORSDORF et al., 2004: RIANO et al., 2003a).
Although LANDSAT remote sensing data could be This system is airborne and an inventory effort at
used for NDVI calculation, it is not convenient for such a scale would be practically and financially
European scale projects because of its swath width unfeasible.
(185 km).
The Disaster Monitoring Constellation (DMC) can be Instead of this method, it is proposed to initially
used for NDVI calculation. reclassify vegetation types into the two aforementioned
This system records reflected electromagnetic categories (variable or intense fire severity) based on
radiation over the same spectral range as Landsat- the following:
ETM 2, 3 and 4 bands, it offers 32 m spatial resolution, - vegetation types from the CORINE database in
daily coverage and a very wide swath of 600-km. NDVI relation with the potential existence of a dense
data can be used in combination with thresholds in understory below specific dominant tree species
order to classify the pixels into the three - the flammability of species (Dimitrakopoulos and
aforementioned categories of land cover density. Papaioannou, 2001)
In addition, a fourth category, high vegetation It is widely known that some dominant tree species
density, can be added to the table above. are associated with specific densities and structures of
This class was not encountered in the Massif des understory vegetation, as is the case for Pinus
Maures catchment described in deliverable D-08-03, halepensis where we find very dense underbrush in
but would be found at the European scale. contrast to Fagus species under which there is limited
Finally, it is worth mentioning that the DMC is the understory vegetation.
first earth observation constellation of five low cost Flammability of species on the other hand can help
satellites providing daily images for applications us to predict the intensity and severity of a forest fire
including global disaster monitoring. before it occurs.

D-08-06.doc 87

The periodicity of estimation for this factor could The disadvantage of this approach is that the
follow that of the Propagation Danger Index described information derived is at a 50-km spatial resolution,
elsewhere in this deliverable since they are highly which renders it difficult to interpret at sub-national
correlated. scales (GOBIN et al., 2003).
Other factors that can lead to high severity wildfires Soil type can be considered a very important factor
are the high density and favourable horizontal structure affecting erodibility since some soil types present
of vegetation cover which may result in extensive higher susceptibility under the same conditions.
spreading, low live and dead fuel moisture contents, For example, PAPAMIXOS 1996 states that terrarosa
intense winds, high temperatures and low relative soils present higher soil erosion risk in comparison to
humidity. other soil types, especially for high slope values.
Studies of GIOVANNINI et al. 1990 indicate that,
The vegetation cover density could be assessed as
depending on the fire intensity; CaCO3 could suffer a
stated earlier with NDVI calculation whereas
break-up and change to CaO or Ca+2 (ANDREU et al.,
information about the live and dead fuel moisture
content from the corresponding indices developed in
This new constitution is far more erodible and can
other sections of this deliverable.
be washed away by water more easily.
Wind speed, temperature, and relative humidity can
GOFAS (2001) refers also to.
be estimated either from historical meteorological data
or from meteorological stations across Europe. It has been mentioned in the D-08-03 deliverable
that high surface temperatures volatise organic Soil erodibility
materials and create gases that move downward in
Three classes of soil conditions were defined and response to a temperature gradient and then condense
coefficients were ascribed to each class as following: on soil particles causing them to become water
Table 23 repellent (LETEY, 2001).
This causes an increase in surface runoff and
Shallow stony soils are assumed to generate more therefore soil erosion rate.
runoff but surface material is less mobile due to the Many researchers focus on the changes in soil
high stoniness, and in many cases, the bare bedrock. properties as the essential element affecting erosion
In contrast, deep soils with fewer stones may have rates (IMESON et al., 1992; KUTIEL and INBAR, 1993;
higher infiltration rates, but they are more erodible, SEVINK et al., 1989).
particularly with respect to the weak aggregate stability GIOVANNINI and LUCCHESI 1991 stress the
and possible presence of hydrophobicity. importance of the soil conditions as the main factor in
Therefore, it was decided to increase the value of the erosion processes.
the coefficient with decreasing stone content. In soils with a high percentage of organic matter,
In order to estimate this factor, without altering the intense fire may lead to complete destruction of the
logic upon which it was built and its functionality, it is organic layer and the exposure of a hydrophobic
proposed to use the method applied in the CORINE mineral soil layer, which acts as a repellent mantle
approach: erodibility is estimated from soil texture, (DEBANO, 1981).
depth and stoniness, extracted from the soil map of the This increases runoff and erosion rates (SEVINK et
European Communities (CEC, 1985). al., 1989).
The CORINE model, which has been applied to IMESON et al. 1992 state that for fires of medium and
many countries in the European Community, is high intensity, the degradation of the vegetation cover
constituted by combining 4 parameters: soil erodibility, and soil organic matter produce the surface
erosivity, topography and vegetation cover (DENGIZ and accumulation of hydrophobic substances that reduce
AKGUL, 2005). infiltration and increase runoff (ANDREU et al., 1996).
The following diagram (Figure 81) shows the three This situation causes increases in runoff production
parameters used to estimate erodibility and how they and soil removal, which favour the loss of nutrients.
can be integrated into four classes. The severity of water repellency depends on the
Figure 82 (taken from KIRKBY, 2001) shows the combined interactions of soil properties and the soil-
overall flow chart of CORINE model for risk heating regime developing during a fire.
assessment of soil erosion The longevity of fire-induced water repellency
depends on some of the same factors that affect its
In the IMAGE/RIVM approach, terrain erodibility was formation.
estimated based on soil type and landform. Water repellency produced by low to moderate
Landform was classified into types by using the severity fires is usually of shorter duration than that
difference between minimum and maximum altitudes produced by high severity fires (DEBANO, 2006).
for each grid cell whereas soil type: The results of preliminary field observations suggest
- was derived from the FAO Soil Map of the World that water repellency might well be an important factor
and responsible for the accelerated erosion experienced
- is composed of soil depth, soil texture, and bulk during the first few years following wildfires (KRAMMES
density (GOBIN et al., 2003). and DEBANO, 1965).

D-08-06.doc 88

An initial laboratory study showed that water

9.3.4 An operational model for the Mediterranean
repellency could be intensified by heating a soil -
organic matter mixture in a muffle furnace at different
temperatures for different lengths of time (DEBANO and The post-fire effects of a wildfire are usually
KRAMMES, 1966). temporally and operationally problematic because the
It was hypothesized that a more efficient coating of first winter following a summer forest fire is the most
mineral soil particles occurred at lower temperatures vulnerable period, (SOTO and DÍAZ-FIERROS, 1998;
and for shorter periods of heating than in the case of Vacca et al., 2000), and runoff and erosion
longer periods of heating at higher temperatures, which management strategies have to be implemented in the
destroyed the organic substances responsible for the first weeks after a fire.
water repellency. The techniques used and their spatial location must
High intensity fires, of 1,200 °C or more, do not often be decided upon urgently and in an administrative
always result in high severity impacts on the soil if their context where the sources of funding, funding channels
duration is short, but low intensity fires of just 300 °C and agency responsibilities are all evolving over a
that smoulder for a long time in roots or organic matter period of days to months (FOX et al., 2006).
can produce large changes in the nearby soil (NEARY,
2004). In deliverable D-08-03, an operational model was
The variability of erosion is associated with fire presented to respond to post-fire runoff and erosion
severity. Fire severity has been reported (INBAR et al.,
1998; ROBICHAUD, 2000) as being one of the main A full description of the model can be found in
reasons for causing erosion variability. Appendix 1.
Initial testing of the model proved satisfactory, but
Fire effects on soil depend mainly on fuel type, fire an additional procedure has been introduced to
severity/intensity, topography and soil conditions improve it further still.
(PAPAMIXOS, 1996). The model itself will not be revised here since it has
Nevertheless, the way soil erodibility is addressed in been the subject of discussion in the preceding section
this method can be considered as simplified since and can be consulted in full in the appendix.
some of the factors affecting susceptibility (soil type The following section deals only with the
and wildfire effects on soil properties) are not taken improvement to the model and it resides in its capacity
fully into account. to estimate the evolution in erosion rate over the first
However, extracting this type of data before an years following a fire.
actual event, and even after an event without direct The change in erosion rate depends principally on
measurements, is unfeasible at this time. vegetation re-growth, and this in turn is closely related
to soil properties. Conclusions
In the Mediterranean environment, both soil
In the Mediterranean, heavy rains follow the fire properties and vegetation characteristics are strongly
season and this means that intervention measures influenced by topography.
have to be undertaken upon urgently. The objective was therefore to predict temporal
changes in erosion rate over a 6 year period based on
Up until today there is a number of factors, which the impact of topography on soil properties and
are related to soil erosion risk assessment that are vegetation growth.
difficult to estimate. The proposed addition to the model was developed
The proposed methodology is based on the use of from a detailed study of the relationships between
factors that are easy to estimate either directly or topography, landuse, soil properties, and post-fire
indirectly. vegetation re-growth in the same catchment used to
It constitutes an approach for “fast and coarse” risk test the initial model.
assessment of potential soil erosion after a fire event. Although the relationships from this single
It should be mentioned that Soil Erosion is closely catchment cannot be extrapolated to the entire
related to the effects of fire on vegetation and soil. Mediterranean region, they are typical of many
The proposed methodology still needs to be Mediterranean environments.
evaluated by conduction of extensive field surveys after
a fire event.

D-08-06.doc 89

In this study, two soil properties were taken into Landuse versus topography
consideration: soil depth, measured using a hand-held
In Mediterranean regions, there is a distinct soil corer, and grain size distribution.
relationship between landuse and topography. Textural analyses were carried out on surface (top 5
The climate is conducive to the production of olives, cm) samples for each topographic section and these
vines and other crops. were compared to sediments accumulated in sediment
In Mediterranean coastal areas of high population traps put in place after the 2003 fire described in the
density, suitable land, possessing both relatively low previous deliverable.
slopes and fertile soils, are shared between agricultural Thirty sites were selected, ten on each slope
production and human occupation. section.
This is clearly illustrated in Figure 83, where the Figure 86 shows the distribution of soil depths
valley bottoms are occupied almost entirely by according to slope section.
vineyards or urban development and the steeper North facing slopes clearly have deeper soils
upland areas are forested with some individual (α=0.05, r2=0.74 in an analysis of variance).
housing. Soil depths on these soils since measurements
The spatial distribution in landuse has implications were stopped at a depth of 1 m due to the difficulties in
for post-fire soil erosion risks. manipulating the hand held corer at greater depths.
In Mediterranean France and throughout much of Differences in soil depth arise essentially from long-
the Mediterranean region, forest fires occur term soil forming processes.
systematically on the steepest slopes of the catchment. Better moister conditions on North facing slopes
The post-fire context is therefore naturally highly increase the rate and duration of pedogenic processes
erosive where slopes tend to be both steeper and and favour vegetation growth (described below).
longer (Figure 84): slope length in the field is Greater rooting depths and vegetation densities, in
determined essentially by field length, whereas in turn, contribute to more active soil formation.
forested areas it is not uncommon to find slopes that In addition to differences in soil depth, we can also
extend from the crest down into valley bottoms where note changes in surface grain size distributions.
slope inclination levels off. The stacked histogram of Figure 87a shows that the
The impacts of slope inclination and length on fine fraction (< 0.2 cm) content on the South facing and
erosion rates are well known and are not elaborated convex slopes is less than half the content on the North
upon here. facing slopes.
In addition to the impact of slope inclination and The sediments trapped in sediment traps (Figure
length on erosion rate, the steeper slopes also 87b) positioned in a stream channel draining a burned
influence sediment deposition. slope (see FOX et al., in press) can explain this
Sediments detached from slopes have a low tendency.
probability of being deposited before reaching a Log Debris Dams (LDD) are located in the channel
tributary, and, once in the tributary, there is little and tend to trap coarser sediments.
likelihood of deposition before they reach the main Sedimentation basin (Basin) sediments represent
channel. the fraction of sediments moved through the channel
As can be seen in Figure 85, the area that could after initial settling behind the LDDs.
serve as a potential zone of deposition is much more Figure 87b clearly shows selective erosion where
restricted in the forested zone. the finer fraction is removed preferentially from burned
This is quite different from an agricultural context slopes.
where fields along river channels can serve as sinks for The results suggest that post-fire erosion processes
sediments eroded upslope and where channel occur longer and/or more frequently on South facing
deposition can be important. and convex slopes.
Therefore, the relationship between landuse and Finally, it should be noted that the gravel size
topography affects not only erosion rate, but it also has fraction (0.2-0.5 cm) represents a greater fraction of the
an impact on sediment redistribution. surface texture on convexities than on South facing
slopes. Topography versus soil characteristics In this case, runoff on the low inclination convexities
has sufficient velocity remove the finest fraction, but not
The relationships between topography and soil the coarse (> 0.2 cm) sediments, as is the case on the
characteristics were determined by an analysis of the South facing slopes.
impacts of topographic units on soil properties.
Slopes were divided into three categories: top-slope
convexities, North-facing slopes and South-facing
slopes. In many Mediterranean environments, water is
a limiting factor and soils tend to be deeper and better
developed on North facing slopes where evapo-
transpiration rates are lower.

D-08-06.doc 90

Actual erosion rates depend on random rainfall Topography versus vegetation
effects: for the same erosion index, erosion rates can
For each of the 3 plots identified, vegetation be significantly different according to the total rainfall
characteristics were measured in two ways. and number of high rainfall intensity events.
Vegetation height was measured on a random Hence, an erosion index of 90 (maximum value in
sample of 20 plants located in a 2 m by 2 m perimeter. Figure 90) can correspond to an actual erosion rate of
This was carried out by hand using a tape measure. from a few T ha-1 to dozens of T ha-1 according to post-
Vegetation density was quantified by photographing fire rainfall patterns.
a 50 cm by 50 cm grid (Figure 88) and then by noting In addition to the impact of rainfall described above,
the percentage of points occupied by vegetation in a several processes contribute to determine the evolution
random sample of 50 points. in post-fire erosion.
For each x,y coordinate in the 50-points sample, the At least three can be described here:
surface was noted as either vegetated (1) or non- - The initial rates of erosion depend on the magnitude
vegetated (0). of the first post-fire rainfall events as described
This value was then converted to a percentage. above.
Three photos were taken at each of the 30 plots for - The initial decrease in erosion rates can be
a total sample size of 90-grid pictures. attributed to the loss of the most easily mobilised
As expected, both vegetation height and density sediments. As the finer fraction is eroded, the
were greater on North facing slopes than South facing increasing stone cover at the surface progressively
slopes and top-slope convexities (Figure 89). decreases sediment detachability and erosion rates.
Furthermore, soil depth, vegetation height and - Erosion rates decrease further due to the
surface cover were all positively correlated. progressive establishment of a vegetation cover on
Slope was significant in analyses of variance the slopes. This occurs faster on North facing slopes
(α=0.05) for both vegetation height (r2=0.74) and than South-facing slopes.
surface cover (r2=0.63). The trends cited above are shown in figure 91.
The greater vegetation growth on North-facing The decrease in the erosion index trend is different
slopes supports the hypothesis evoked above that for North-facing and South-facing slopes since
selective erosion is more intense on South facing vegetation dynamics proceed at different rates.
slopes than North facing slopes since the latter take
more time to recover after a fire. The change in erosion index over time was
In addition, South-facing slopes have always had estimated using non-linear decaying exponential curves
more human occupation, as is shown by old terraces for each of the slope types (Table 24).
and newer housing, so the likelihood of fire ignition has The two extremes, North- and South-facing slopes,
always been greater on South facing slopes. were defined so that the index reached approximately
This, combined with drier conditions, suggests that zero within three years for the North-facing slope and
South-facing slopes may burn more often than North within six years for the South-facing slope.
facing slopes. The 3-year limit for the North-facing slope was
determined in the field by observations of the filling of a Implications of topographic effects for erosion sedimentation basin.
modelling The 6-year limit for the South-facing slope is an
estimation based on the current vegetation cover.
The model described in Annex gives an Convex slopes were treated with the same equation
instantaneous spatial distribution of post-fire soil
as South-facing slopes and West- and East-facing
erosion risk. slopes were given decay values intermediate to the two
However, post-fire erosion, though at its maximum extremes.
during the first year, can continue over a period of
several years. Slope orientation was defined for each cell in
Faster vegetation re-growth on North facing slopes Figure 90.
suggests that erosion will be decrease more rapidly Top-slope convexities were defined as slopes above
here than on South-facing slopes or top-slope 150-m with an inclination of less than 10%.
convexities. This proved to be a simple but effective method of
This is supported by both the vegetation data identifying these topographic features.
gathered in-situ and the soil and textural results. Once the orientation was simplified into the five
To complete the model, it is therefore necessary to categories in table 24, the appropriate regression
consider this temporal evolution. equation was applied to each of the cells, year by year
for 6 years, in order to predict its evolution over time.
Figure 90 shows the initial calculation of the index of For each year, the erosion index value was
soil erosion risk.
converted to an erosion risk class, as in the initial
The initial model re-classed these values into method (see Annex).
categories according to erosion risk. The result is a dynamic erosion risk spatial
It should be noted that these values represent a
distribution showing the potential erosion risk for each
post-fire structural risk. year during the 6-year period (Figures 92a to 92f).

D-08-06.doc 91

The map for the sixth and final year is not shown Conclusions
since the entire surface is in the low class category.
In the initial year after the fire, the North facing The Mediterranean environment is subject to regular
slopes initially in the very high-risk class have dropped forest fires and it is possible that the number and
to the high and intermediate classes. intensity of fires will increase in coming years.
By the third year, all of the North-facing slopes are Reacting quickly to post-fire erosion risk is therefore
in the low risk class. an issue that must be dealt with by European countries
Many of the South-facing slopes persist in the very in particular.
high-risk class after the first year, despite the The above deliverable demonstrated two
exponential decrease in erosion risk, and many of approaches to dealing with post-fire erosion risk.
these slopes remain in the intermediate class even In the first, a method for determining pre-fire
three years after the fire. potential risk at the European scale is described.
Showing the temporal evolution in erosion risk is an In this context, the issue addressed is how to
added factor in determining where erosion control acquire apriori knowledge about post-fire erosion risk
measures should be put into place after a fire. before an actual fire occurs.
The second approach refines a post-fire erosion
mapping method that was presented in deliverable D-
The major improvement concerns the integration of
post-fire erosion risk evolution over a period of 6 years
after the fire.
In this approach, it was demonstrated that
topography plays a major role in determining post-fire
vegetation recovery and soil depth, and these, in turn,
have a major influence on post-fire erosion.
Field data support the approach and it has
considerable potential for testing in a number of
Mediterranean post-fire contexts.
9.3.5 Figures

Figure 80: Map of NDVI produced by Bayaramin et al. (2006) for their study area.

D-08-06.doc 92

Soil texture (ST) Soil Depth (SD) Stoniness (SS)

0 for Bare rock 1 for > 75 cm 1 for >10 %

1 for C, SC, SiC 2 for 25-75 cm 2 for <10 %
2 for SCL, CL, SiCL, LS, S 3 for 25 cm
3 for L, SiL, Si, SL

Soil erodibility(K)

0 for ST *SD *SS=0

1 for 0 < ST *SD *SS < 3
2 for 3< ST *SD *SS < 6
3 for ST *SD *SS > 6

Figure 81: Methodology for CORINE soil erodibility assessment (CORINE, 1992).

Soil Texture, ST
0 for Bare rock
1 for C, SaC, SiC
2 for SsCL,CL,SiCL,Lsa
3 for SaL, L, SiL, Si

Erodibility, K
Soil Depth, SD 0 for ST.SD .SS =0
1 for >75 cm 1 for 0< ST.SD .SS <3
2 for 25 – 75 cm 2 for 3< ST.SD .SS <6
3 for <25 cm 3 for ST.SD .SS >6

Soil Stoniness, SS
1 for >10%
2 for <10%

Fournier Index, F Erosivity, R

1 for Σpi2/Σp <60 1 for F.B<4
2 for 60<Σpi2/Σp<90 2 for 4<F.B<8
3 for 91<Σpi2/Σp<120 3 for F.B>8
4 for Σpi2/Σp>120
Potential Soil Erosion
Bagnouls-Gaussen Risk, EP
Aridity Index, B 0 for K.R.S = 0
1 for Σ(2Ti-pi) =0 1 for 0 < K.R.S < 5
2 for 0<Σ(2Ti -pi)<50 2 for 5 < K.R.S < 11
3 for 50<Σ(2Ti-pi)<130 3 for K.R.S > 11
4 for Σ(2Ti-pi)>130

Slope angle, S
1 for <5%
2 for 5-15% Actual Soil Erosion
3 for 15-30% Risk, E A
4 for >30% 0 for EP.V = 0
1 for EP.V = 1-2
2 for EP.V = 3-4
Land Cover, V
3 for EP.V >=5
1 for fully protected
2 for not fully protected

Figure 82: Methodology for CORINE Soil Erosion assessment

D-08-06.doc 93

Hilly metamorphic
uplands Alluvial plain


Vineya rds
Ot her

Figure 83: Steeper uplands in the coastal area of the Massif des Maures (France) are occupied by forest, the
lowland plains are occupied by vineyards and urban development.

a) Slope inclination b) Slope length

Mean (%) Median (%) Mean (m) Median (m)

Forest: 23.3 25.5 Forest: 357.1 400.0
Non-forest: 10.1 6.6 Non-forest: 162.7 162.0
35 400

) 28
%( 300
y )
T 21 m(
e ht
s g 200
u n
d e
n 14 L
a Forest e
L ol
r Non-forest
p S
a 7
0 0
0 20 40 60 80 100 Forest Vineyard
Slope Inclination (%)

Figure 84: Highly erosive topographic conditions with few possibilities for sediment deposition in the fire zone

D-08-06.doc 94



Soil Depth (cm)

30 60
Area (%)

20 40

10 20

0 0
Forest Non-forest North South Convex

Left Figure 85: Percentage area in the forest and non-forest zones with a slope inclination less than 4%.
Right Figure 86: The North-facing slopes have a greater depth than the South-facing and convex slopes.

a) Slope grainsize b) Sediment trap

characteristics grainsize characteristics
100 100
0.2-2 cm

80 ) 80
) %(
%( n
n t
oi 60 u 60
t bi
u rt
bi si
si D
D e
40 si 40
s ar
ar G
G 20 ai 20

0 0
North South Convex BASIN LDD
< 0.2 cm < 2.0 mm
0.2-0.5 cm • Selective erosion of the finer fractions (< 2 mm) from slopes.
2-20 mm
0.5-1 cm > 20 mm
1-2 cm • Increasing stone cover with erosion – South and Convex slopes.
>2 cm
• Intermediate fraction (0.2-0.5 cm) detached from South facing
slopes but not convexities

Figure 87: North-facing slopes have finer textures than South facing and convex slopes (a)
This results from selective erosion of the finer fraction (measured in sediment traps - b)).

D-08-06.doc 95

Figure 88: Example of a grid photograph used to measure vegetation density.

a) Undergrowth vegetation height b) Undergrowth vegetation cover

120 100

80 R2=0.63
) R2=0.74
c( )
80 %(
t r 60
g e
i v
e c
h n
n oti
oi at
t 40 e
at g
e 20

0 0
North South Convex North South Convex

Correlations: Vegetation height/area: r=0.82; Soil depth / Vegetation height:

r=0.87; Soil depth / Vegetation area: r=0.75

Figure 89: Vegetation height (a) and density (b) were both greater on North facing slopes than on South facing
slopes and top-slope convexities.

D-08-06.doc 96

Theoretical range: 1-135; Actual range: 1-90

Figure 90: Raw values for soil erosion risk calculated from the initial soil erosion model.

Peak erosion = f(rainfall events)

Initial decrease = f(sediment

availability, increased stoniness
through selective erosion)
Initial Erosion Index


Greatest decrease = f(vegetation




0 1 2 3 4 5 6 7
Time (years)

Figure 91: Theoretical curves for soil erosion index trends over time: rates decrease and the greatest decrease
arises due to differences in vegetation growth, which are significantly different on North and South facing slopes

D-08-06.doc 97

Figure 92a and b: Initial soil erosion (left) and one year after the fire risk maps

Figure 92c and d: Erosion risk maps two (left) and three (right) years after the fire

Figure 92e: Erosion risk maps four (left) and five (right) years after the fire

D-08-06.doc 98

9.3.6 Tables

Table 20: Coefficients assigned to each slope class

Slope (%) Coefficient
0-5 1
5-10 2
10-20 3
20-30 4
>30 5

Table 21: Coefficients assigned to each vegetation density

Pre-fire vegetation density Coefficient
Medium 1
Low 2
Bare 3

Table 22: Coefficients assigned to each fire severity level.

Fire severity Coefficient
Variable 1
Intense 3

Table 23: Coefficients assigned to each soil condition class

Soil erodibility Coefficient
Shallow soil with high stone content (depth < 20 cm) 1
Intermediate depth and stoniness 2
Deep soil with few stones (depth > 50 cm) 3

Table 24: Non-linear equations used to predict the evolution in the erosion index
Slope orientation Non-linear equation

North EI = IV × e −2 yr
South EI = IV × e −0.8 yr
Convex EI = IV × e −0.8 yr
East EI = IV × e −1.2 yr
West EI = IV × e −1.6 yr

D-08-06.doc 99


10.1 GENERAL PRESENTATION In any case, several straightforward solutions can

be raised to make this conversion, using for instance
The basic tools for the application of the proposed
linear functions from the maximum and minimum
EM-WFRI for the regional wildfire management are
values adjustment.
available once the basic structure has been defined
The inclusion of scaled sub-indices to probability
and the different components of this index have been
units will make easier the use of the different
integration methods.
Next, the methodology to integrate these
components and finally the most suitable procedure to The hierarchical structure of the index, as well as
spread the information should be defined. the probability units format of the different sub-indices,
Concerning this, the multiple suggested possibilities will allow the inclusion of new variables or the
in the D-08-05 must be remarked. substitution or removal of some of the proposed ones,
Given the diversity of available methods, and the in function of the available information of the territory in
variety of potential applications and users, whether of which will be used.
the total of a part of the identified risk components, a Moreover, it will allow introducing modifications in
unique integration methodology has not been function of the work scale, as given some of the
considered to propose. included variables have been presented in this
Instead of this, the index structure and the obtained deliverable (climate, population, etc.)
input variables are offered to the users, leaving them to
In the definition of the risk index structure, the
determine the most suitable integration method
proposed input variables have a spatial nature and a
according to the index destination and the availability of
digital format, as far as possible according to the
combination of its components with additional
available information at the European Union level.
information that can be useful for the management.
Remote sensing information (with a very suitable
In fact, a survey that was performed in Spain to
spatial and temporal component for this kind of studies)
potential users of this index (forest managers and
was considered, as well as point data (meteorological
technicians) in the framework of the FIREMAP
variables, census data, etc.) that would be spatialized
research Project (
by means of any spatial interpolation algorithm, and
reveals the preference of managers and technicians to
cartographic information available in digital format
have the basic components of the index (variables and
(Corine Land Cover maps, digital terrain models,
sub-indices), instead of the final index value.
national cartographic databases, etc.).
Thus it enables the use of the information and its
All this information leads to a suitable series of
integration in the prevention systems available for each
spatialized and geo-referenced information layers in
digital format for using in a Geographical Information
In order to make easier the integration of variables System (GIS)
according to the user interests, the final risk index is
In this context, GIS is a one of the most suitable
generated from three sub-indices: the ignition danger
tools for obtaining a risk wildfire index with the
index, the propagation danger index, and the
proposed characteristics in this deliverable, allowing
vulnerability index.
the management, manipulation, analysis, modelling
These indices are obtained separately, in this way
and representation of the input data as well as the
each one offers specific information about each need in
output indices.
terms of wildfire prevention (ignition, propagation, and
Moreover, they make considerably easier the
wealth), and can be used independently for each
updating processes of the information.
management task.
In this updating process, the existing connection
The three indices combination will offer the global
between the GIS and the remotely sensed images
risk value, and depending on the user, the indices
processing should be considered, thus this is the
weighting can be varied according to the importance of
source of an important part of the dynamic information
each factor in the territory or in the considered
used in the risk index (mainly fuel moisture content).
management task.
On the other hand, the disposal of this information in
It would be recommended for the integration of the a GIS means that it can be used in forest management
variables that all indices that are taken into account for for different facets other than risk indices.
the global risk -not only the three intermediates, but
However, the use of this tool, that considerably
also those that have been used to generate the three
facilitates the process, does not guarantees the
ones- were scaled in probability units (0-1 or 0-100%).
goodness of the obtained index, which reliability will
The way to achieve the conversion from the original
depend on the selection and accuracy of the input
index value to probability units has been included in
variables, as well as the success in the weighting risk
this deliverable for some of the indices (i.e, fuel
assigned to each variable in the model.
moisture content).
Undoubtedly, GIS offer a wide range of possibilities
to define the fire risk integration model:

D-08-06.doc 100

- qualitative methods, where arbitrary weights are The Fire-Climate-Society Strategic Model is part of
based on the judgment of an expert; the Model Wildfire Alternatives (WALTER), located at
- quantitative indices, based on multi-criteria the University of Arizona.
evaluation or other survey system; WALTER is an interdisciplinary research initiative
- regression techniques, where statistical estimation aimed at improving our understanding of the processes
methods are applied to explain fire occurrence; and consequences of interactions among wildfire,
- neural network, similar conceptually to the climate and society.
regression models, although with different fitting FCS-1 Fire-Climate-Society (FCS-1) is an online,
procedures. All these methods were analysed in spatially explicit strategic wildfire-planning model with
detail in the deliverable D-08-05. Its use allows, with an embedded multi-criteria decision process that
a good enough effective cost and speed, the facilitates the construction of user-designed risk
creation of new models, the model application at assessment maps under alternative climate scenarios
different scenarios or the modification of risk classes and varying perspectives of fire probability and values
limits, among others. Because of it, it seems at risk.
unquestionable the use of this tool for the
The model is generated by integrating the following
generation of the risk index proposed in this
variables or sub-indices: fuel moisture stress index, fire
return interval departure, large fire ignition probability,
In these last years an important number of GIS lightning probability, human factors of fire ignition,
applications have appeared in the web, with few use recreational value, species habitat richness, property
requisites, that make considerably easier the access to value, personal landscape value.
this application from everywhere. One of the main advantages of this system is that it
The simplest ones only allow the visualization or can be used with a standard model, where the weights
download of a variable of interest, as the more complex of the different variables are assigned according to the
ones allow the user to interact with the proposed model defined by the experts who created the system,
models. or the user can decide the risk weights of each input
Several examples of both situations are found in the variables, as shown in the figure 93.
field of wildfire risk assessment. This system seeks to capitalize on advances in geo-
For instance, in the United States we can obtain in spatial, analytical, and web delivery technology to
real time the National Fire Danger Rating System provide access to scientific and management activities.
estimation or any of its sub-indices that is composed
Bearing in mind the required structure of the Euro-
(USDA Forest Service;
Mediterranean Wildland Fire Risk Index EM-WFRI:
In the European Union, several risk indices by
- the nature of variables that compose it,
means of the Risk Forecast System can be obtained in
- the possibility of modifying and updating these
the European Forest Fire Information System (EFFIS;
- the integrating of different methods,
In contrast to this simpler situation, in which the user
- its use at different spatial scales, etc.,
only checks the generated information by the different
it is obvious:
institutions, situations that are more complex may
- to recommend generating this risk index by means
appear, where the user can interact with the index
of GIS and,
- to integrate the use of this tool in a web service.
One of the most interesting examples with this
option is the Fire-Climate-Society Strategic Model In this way a very flexible tool would be created, and
(FCS-1) developed by the Arizona University could be used for wildfire management in very different
(, facets of prevention and at different scales (local,
regional, national, supranational), allowing the user to
define the most suitable scenario at the time of use.

Figure 24: Analytic Hierarchy Process (AHP) weights to the Fire-Climate-Society Strategic Model
(Wildfire Alternatives, University of Arizona).

D-08-06.doc 101


1996. Forest fire risk modelling using remote Analysis. Longman Scientific and Technical, Burnt
sensing and geographic information system. Current Mill, Essex, England.
Science (70 10):928-933.
BAIRD AW. 1965. Attitudes and Characteristics of forest
AERIAL INFORMATION SYSTEMS Inc. 1981). Final report. residents in three Mississippi counties., 48:
Spatial risk method for predicting wildland fire Mississippi State University, Social Science
occurrence. Crestline, California: USDA, Pacific Research Center Report No. 29.
Southwest Forest and Range Experiment Station.
AGUADO I; CHUVIECO E; CAMARASA A; MARTÍN MP; Management problems in Mediterranean cork oak
CAMIA A 1998). Estimation of Meteorological Fire forests: post-fire recovery. Journal of Arid
Danger Indices from Multitemporal Series of NOAA- Environments 54, 565-569.
AVHRR data. Pages 1131-1147 in D. X. Viegas,
BARET F.; FOURTY T. 1997. Estimation of leaf water
editor. III International Conference on Forest Fire
content and specific leaf weight from reflectance
Research - 14th Conference on Fire and Forest
and transmittance measurements. Agronomie, 17,
Meteorology. ADAI, Coimbra.
444- 455.
AHLGREN IF; AHLGREN CE 1960. Ecological effects of
BARET F; GUYOT G. 1991. Potentials and limits of
forest fires. Bot. Rev. 26, 483–533.
vegetation indices for LAI and APAR assessment.
ALBINI FA 1976. Estimating wildfire behavior and Remote Sensing of Environment, 35, 161-173.
effects. USDA For. Serv. Gen. Tech. Rep. INT-30.
ALBINI FA; REINHARDT ED 1995. Modeling ignition and CORINE methodology to Assess Soil Erosion Risk
burning rate of large woody natural fuels. Int. J. in the Semi-Arid Area of Beypazari, Ankara. Turk J
Wildland Fire, 5: 81–91. Agric For, 30, 81-100.
ALEXANDRIAN D; GOUIRAN M 1990. “Les causes BEMMERZOUK AM. 1997. Evaluation of the CFFDRS
d'incendies : levons le voile.” Revue forestière and the KBDI in the Mediterranean-type
française - Espaces forestiers et incendies Numéro environment of Crete (Greece. M. Sc. Thesis,
spécial 1990. Chania, Greece.
KYUN I; MARTÍN MP; SALAS FJ 1996. Estimating Landscape Structure Programs with GIS.
temporal dynamics of fuel moisture content of
Mediterranean species from NOAA–AVHRR data.
BERTRAND AL; BAIRD AW. 1975. Incendiarims in
EARSEL Advances in Remote Sensing, 4(4), 9 –
Southern Forest: a decade of sociological research:
Report. USDA Forest Service, Southern Forest
ALTOBELLIS A 1983. A survey of rural population density Experimental Station and Social Science Research
and forest fire occurrence in the South, 1956-1970. Center at Mississipi University.
New Orleans: USDA Forest Service, Southern
BLACKMARR WH; FLANNER WB. 1968. Seasonal and
Forest Experimental Station. Report SO-294.
diurnal variation in moisture content of six species of
ANDERSEN HE; MCGAUGHLEY RJ; REUTEBUCH SE 2005. Pocosin shrubs. Research Paper: SE-33. pp. 11.
Estimating forest canopy fuel parameters using USDA Forest Service, Southeastern Forest
LIDAR data. Remote Sensing of Environment, 94, Experiment Station. Asheville, NC.
BLANCHI R ; JAPPIOT M et al. 2002. Forest fire
ANDERSON HE; SCHUETTE RD; MUTCH RW 1978. assessment and cartography - a methodological
Timelag and equilibrium moisture content of approach. Forest fire research and wild land fire
ponderosa pine needles. Research Paper SE-74, safety, Coimbra.
USDA Forest Service, Intermountain Research
Station, Ogden, Utah.
cover technical guide. Addendum 2000. European
ANDREU V; RUBIO JL; FORTEZA J; CERNI R 1996. Environment Agency - Technical report n° 40.
Postfire Effects on Soil Properties and Nutrient
BOVIO G; CAMIA A. 1998. An analysis of large forest
Losses. International Journal of Wildland Fire 6 (2),
fires danger conditions in Europe. Pages 975-994 in
pp. 53-58.
DX VIEGAS, editor. 3rd International Conference on
ANDREU V; IMESON AC; RUBIO JL 2001. Temporal Forest Fire Research, Luso-Coimbra, Portugal.
changes in soil aggregates and water erosion after a
BOVIO G; CAMIA A. 2000. Description Eudic software for
wildfire in a Mediterranean pine forest. Catena 44,
fire danger indices calculation.

D-08-06.doc 102

BOVIO G; SOL B; VIEGAS DX. 1994. Synthese des CARLSON JD; BURGAN RE. 2003. Review of users’
Travaux sur l'intercomparison des indices de danger needs in operational fire danger estimation: The
meteorologique d'incendie. Final Report of Minerve I Oklahoma example. International Journal of Remote
Project EC, Environment and Climate Programme. Sensing, 24(8), 1601– 1620.
BOVIO G; CAMIA A.; GOTTERO F. 1999. Piano Regionale CARREGA P. 1990. Climatology and index of forest fire
per la difesa del patrimonio boschivo dagli incendi hazard in Mediterranean France. Pages B.05-01/11
1999-2001. Regione Piemonte, Assessorato in International Conference on Forest Fire
Economia Montana e Foreste. Research, Coimbra, Portugal.
BOWMAN WD 1989. The relationship between leaf water CARREGA P. 1994. Topoclimatologie et habitat. U.F.R.
status, gas exchange, and spectral reflectance in Espaces et Cultures. Nice, Université de Nice -
cotton leaves. Remote Sensing of Environment, 30, Sophia Antipolis: 408.
249– 255.
CARTER GA. 1991. Primary and secondary effects of
BRADSHAW LJ; DEEMING JE; BURGAN RE; COHEN JD water content on the spectral reflectance of leaves.
1983. The 1978 National Fire-Danger Rating American Journal of Botany, 78, 916– 924.
System. Technical Documentation: GTR INT-169.
CARVACHO L. 1998. Evaluación de la estimación de
pp. 44. USDA Forest Service. (Ogden, Utah.
grandes incendios forestales en la cuenca
BURGAN RE. 1979. Estimating live fuel moisture for the mediterránea europea por redes neuronales y
1978. National Fire Danger Rating System. USDA regresión logística. Serie Geográfica 7:73-85.
For. Serv. Res. Pap. INT-226.
BURGAN RE; ROTHERMEL RC. 1984. BEHAVE: fire Chapter 7. Assessing Potential Wildfire Effects on
behaviour prediction and fuel modeling system. Fuel People. Co-published simultaneously in Journal of
subsystem. USDA For. Serv. Gen. Tech. Rep. INT- Sustainable Forestry, 11(1/2), 159-175; and:
167. Mapping Wildfire Hazards and Risks (ed: Neil
Sampson, R. Dwight Atkinson, and Joe W. Lewis),
BURGAN RE; HARTFORD RA. 1993. Monitoring
Food Product Press, an imprint of The Haworth
Vegetation Greenness with Satellite Data. GTR INT-
Press, Inc., 2000, , 159-175.
297, USDA Forest Service, Ogden, Utah.
BURGAN, R. E., KLAVER, R. W. and KLAVER, J. M. 1998.
moisture content in shrubs to predict fire risk in
Fuel models and fire potential from satellite and
Catalonia (Spain. Agricultural and Forest
surface observations. International Journal of
Meteorology, 116, 49– 59.
Wildland Fire 8:159-170.
CEC 1985. Soil Map of the European Communities,
BURGAN RE. 1999. Personal Communication. Not
1:1,000,000. CEC Luxembourg, 124pp., 7 maps.
GRÉGOIRE JM. 2001. Detecting vegetation leaf water
Meteorological fire danger indices and remote
content using reflectance in the optical domain.
sensing. Pages 39-59 in E. Chuvieco, editor.
Remote Sensing of Environment, 77, 22–33.
Remote Sensing of Large Wildfires in the European
Mediterranean Basin. Springer-Verlag, Berlin. CECCATO P; GOBRON N; FLASSE S; PINTY B;
TARANTOLA S. 2002. Designing a spectral index to
CAMIA A ; BOVIO G ; MARZANO R. 2002. Preliminary
estimate vegetation water content from remote
analysis of wildland-urban interface fire prone areas
sensing data: Part 1. Theoretical approach. Remote
in Italy. In: Viegas DX (ed.), Proc. IV Int. Conf. on
Sensing of Environment, 82, 188– 197.
Forest Fire Research and 2002 Wildland Fire Safety
Summit, Luso (Portugal), Millpress Science Pub., CECCATO P; LEBLON B; CHUVIECO E; FLASSE S;
Rotterdam, pp.9; ISBN 90-77017-72-0. CARLSON JD. 2003. Estimation of live fuel moisture
content. In E. Chuvieco (Ed.), Wildland fire danger
estimation and mapping. The role of remote sensing
Spatial analysis in European Wildland-Urban
data (pp. 63–90. Singapore: World Scientific
Interface environments using GIS. In: Xanthopoulos
G. (ed.), Proc. II International Workshop “Forest Fire
in the Wildland-Urban Interface and Rural Areas in CERDÀ A; IMESON AC; CALVO A. 1995. Fire and aspect
Europe: an integral planning and management induced differences on the erodibility and hydrology
challenge”; 15–16 May 2003, Athens (Greece. of soils at La Costera, Valencia, south-east Spain.
Mediterranean Agronomic Institute of Chania Catena 24, 289-304.
(MAICh), Graphic Art and Publishing Unit, Chania,
Crete, 185 – 196.
WILLIAMS D. 1983. Fire in Forestry. Forest Fire
CARDILLE JA; VENTURA SJ; TURNER MG. 2001. behavior and effects. John Wiley and Sons: New
Environmental and social factors influencing York.
wildfires in the Upper Midwest, United States.
Ecological Applications 11 (1):111-127.

D-08-06.doc 103


Integrated Approach to Natural Hazards. Risk Estimation of fuel moisture content from
Assessment Using GIS: With Reference to multitemporal analysis of Landsat Thematic Mapper
Bushfires. Environmental Management, 31(4), 546- reflectance data: Applications in fire danger
560. assessment. International Journal of Remote
Sensing, 23(11), 2145–2162.
CHLADIL MA; NUNEZ M. 1995. Assessing grassland
moisture and biomass in Tasmania. The application CHUVIECO E; AGUADO I; COCERO D; RIAÑO D. 2003a.
of remote sensing and empirical models for a cloudy Design of an empirical index to estimate fuel
environment. International Journal of Wildland Fire, moisture content from NOAA–AVHRR analysis in
5, 165– 171. forest fire danger studies. International Journal of
Remote Sensing, 24(8), 1621– 1637.
CHOU YH. 1992. Spatial autocorrelation and weighting
functions in the distribution of wildland fires. CHUVIECO E; ALLGÖWER B; SALAS FJ. 2003b.
International Journal of Wildland Fire 2 (24):169- Integration of physical and human factors in fire
176. danger assessment. In Wildland Fire Danger
Estimation and Mapping. The Role of Remote
Sensing Data, ed. E. CHUVIECO, pp: 197-218.
probability of fire occurrence in San Jacinto
Singapore: World Scientific Publishing.
Mountains, California, USA. Environmental
Management 17 (1):129-140. CHUVIECO E; AGUADO I; DIMITRAKOPOULOS A. 2004a.
Conversion of fuel moisture content values to
CHRISTENSEN NL. 1987. The biogeochemical
ignition potential for integrated fire danger
consequences of fire and their effects on the
assessment. Canadian Journal of Forest Research-
vegetation of the Coastal Plain of the southeastern
Revue Canadienne De Recherche Forestiere 34
United States. In: The role of fire in ecological
(11): 2284-2293.
systems, (L. Trabaud Ed.. SPB Academic
VEGA J., RIVA, J., and PÉREZ, F. 2004b. Combining
CHRISTIANSEN J; FOLKMAN W. 1971. Characteristics of
NDVI and surface temperature for estimation of live
people who start fires...some preliminary findings.
fuel moisture content in forest fire danger rating.
Berkeley, Cal.: USDA, Forest Service. Pacific
Remote Sensing of Environment, 92, 322-331.
Southwest Forest and Range Experiment Station.
ORUETA A; Prado E. (in press. Improving burning
CHUVIECO E; CONGALTON RG. 1989. Application of
efficiency estimates through satellite assessment of
remote sensing and Geographic Information
fuel moisture content. Journal of Geophysical
Systems to Forest fire hazard mapping. Remote
Research - Atmospheres
Sensing of Environment 29:147-159.
COHEN WB. 1991. Response of vegetation indices to
CHUVIECO E; SALAS FJ. 1996. Mapping the spatial
changes in three measures of leaf water stress.
distribution of forest fire danger using GIS.
Photogrammetric Engineering and Remote Sensing,
International Journal of Geographical Information
57(2), 195– 202.
Systems 10 (3):333-345.
COLE LW; KAUFMAN HF. 1963. Socio-economic factors
CHUVIECO E; SALAS FJ; VEGA C. 1997. Remote sensing
and forest fire in Mississippi counties. Mississippi.
and GIS for long-term fire risk mapping. Pages 91-
USA: Mississippi State University Social Sciences
107 in E. Chuvieco, editor. A review of remote
Research Centre. Preliminary Report 14.
sensing methods for the study of large wildland
fires. Departamento de Geografía, Universidad de COLL C; CASELLES V. 1997. A global split-window
Alcalá, Alcalá de Henares. algorithm for land surface temperature from AVHRR
data: Validation and algorithm comparison. Journal
of Geophysical Research, 102B14, 16697–16713.
RIAÑO D. 1999a. Short-term fire risk: Foliage
moisture content estimation from satellite data. In E. COLLINS DB. 1996. “FireLib User Manual and Technical
Chuvieco (Ed.), Remote sensing of large wildfires in Reference”.
the European Mediterranean Basin (pp. 17– 38.
COMUNIDAD DE MADRID 2000): Plan de Protección Civil
Berlin: Springer-Verlag.
de emergencia por incendios forestales en la
SILVA F. 1999b. Integrated fire risk mapping. In 4-28.
Remote Sensing of Large Wildfires in the European
Mediterranean Basin, Ed. E CHUVIECO. pp 61-84.
EUROPEENNES. 1987. Carte de la Végétation
Berlin: Springer-Verlag.
Naturelle des Etats membres des Communatés
Européennes et du Conseil de l’Europe. Office for
Official Publications of the European Communities,

D-08-06.doc 104

CORINE 1992. Soil Erosion Risk and Important Land DELGADO XP 2001. Spatial patterns of forest fires in
Resources in the Southern Regions of the European Catalonia (NE of Spain) along the period 1975-
Community. EUR 13233, Luxembourg. 1995. Analysis of vegetation recovery after fire.
Forest Ecology and Management, 147, pp.67-74 .
CROSSON P 1997. Will erosion threaten agricultural
productivity? Environment 39 (8), 4– 12. DELICHATSIOS MA, PANAGIOTOU T, KILEY F. 1991. The
use of time-to-ignition data for characterizing the
thermal inertia and the minimum (critical) heat flux
Predicting the ignition of crown fuels above a
for ignition or pyrolysis. Combust. Flame, 84: 323–
spreading fire. Part II: model evaluation.
International Journal of Wildland Fire, Vol. 15, pp.
61-72. DENGIZ O, AKGUL S, 2005. Soil Erosion Risk
Assessment of the Golbasi Environmental
Protection Area and Its Vicinity Using the CORINE
the Vulnerability of People and Places: A Case
Model. Turk J Agric For, 29, 439-448.
Study of Georgetown County, South Carolina.
Annals of the Association of American Geographers, DESBOIS N, DESHAYES M, BEUDOIN A. 1997. Protocol for
90(4), 713-737. fuel moisture content measurements. In 'A review of
remote sensing methods for the study of large
DATT B 1999. Remote sensing of water content in
wildland fires. (Ed E. Chuvieco) pp. 61-72.
eucalyptus leaves. Australian Journal of Botany, 47,
Departamento de Geografía, Universidad de Alcalá.
(Alcalá de Henares, Madrid.
2004. Mapping forest fire occurrence at a regional
Flammability assessment of Mediterranean forest
scale. Remote Sensing of Environment 92(3):363-
fuels. Fire Technology, 37, 143– 152.
Effects of fire and torrential rainfall on erosion in a
Mediterranean gorse community. Land Degradation
and Development 14, 203-213.
DEBANO LF, KRAMMES JS 1966. Water repellent soils
variability of soil hydrophobicity in fire-prone
and their relation to wildfire temperatures.
eucalyptus and pine forests, Portugal. Soil Science
International Association of Hydrological Science
163, 313-323.
Bulletin 2, 14-19.
DEBANO LF 1981. Water repellent soils: a state-of-the-
water repellency: its causes, characteristics and
art. USDA Forest Service General Technical Report
hydro-geomorphological significance. Earth-Science
PS W-46, Berkeley, CA.
Reviews 51, 33-65.
DEBANO LF. 2000a. Water repellency in soils: a
DONOGHUE LR, MAIN WA. 1985. Some Factors
historical overview. Journal of Hydrology 231-231,
Influencing Wildfire Occurrence and Measurement
of Fire Prevention Effectiveness. Journal of
DEBANO LF. 2000b. The role of fire and soil heating on Environmental Management 20 (1):87-96.
water repellency in wildland environments: a review.
Journal of Hydrology 231-231, 195-206.
Determining the Economic Relationship Betwen Law
DEEMING JE, BURGAN RE, COHEN JD. 1978. The Enforcement Activities and Arson Wildfires: a
National Fire-Danger Rating System - 1978. GTR Feasibility Study in Arkansas. Journal of
INT-39, USDA Forest Service, Ogden, UT. Environmental Management 25:377-393.
DEFRIES RS, TOWNSHEND JRG 1994. Global land- DOOLITTLE ML. 1979. Southern woods-burners: a
cover: Comparison of ground-based data sets to descriptive analysis: USDA Forest Service.
classifications with AVHRR Data. In G.M. Foody, Southern Forest Experiment Station. Research
and P.J. Curran (Eds.), Environmental remote Paper SO-151
sensing from regional to global scales ( pp. 84– 110.
DOOLITTLE ML, WELCH GD. 1974. Fire prevention in the
Chichester: John Wiley and Sons.
Deep South: personal contact pays off. Journal of
DE JONG SM, 1994. Applications of reflective remote Forestry 72:488-490.
sensing for land degradation Studies in a
DOUGUEDROIT A, JOLY C. 2002. Spatial distribution of
Mediterranean Environment. PhD thesis, Utrecht
the forest fires in lower Provence : the influence of
University, The Netherlands, 237 pp.
Marseille. Forest Fire Research and Wildland Fire
DELETIC A. 2001. Modelling of water and sediment safety, Coimbra, Milpress.
transport over grassed areas. Journal of Hydrology
248, 168-182.

D-08-06.doc 105

BLAIRD BJ, HOFTONA MA, CHAZDONE RL, and 10-hour timelag fuel moisture calculations for
WEISHAMPELF JF, PRINCEA SD. 2002. Estimation of fire danger rating. Research Note RM-207, USDA,
tropical forest structural characteristics using large- Forest Service, Fort Collins, Colorado.
footprint lidar. Remote Sensing of Environment, 79,
Moisture content calculations for 1000-Hour Timelag
DROUET JC, SOL B. 1993. Mise au point d'un indice Fuels. Forest Science 27:19-26.
numerique de risque météorologique d'incendies de
forêts. Forêt Mediterranéenne 14:155-162.
Quantitative Geography: Perspectives on Spatial
EASTMAN JR 1993. IDRISI. Clark University, Worcester. Data Analysis. SAGE Publications, London.
1997. Mapping city lights with nightime data from Geographically weighted regression: The analysis of
the DMSP Operational Linescan System. spatially varying relationships. John Wiley and Sons,
Photogrammetric Engineering and Remote Sensing New York, NY. 269 p.
ESPELTA JM, RETANA J, HABROUK A 2003. An economic Mapping erosion risk and selecting sites for simple
and ecological multi-criteria evaluation of erosion control measures after a forest fire in
reforestation methods to recover burned Pinus nigra Mediterranean France. Earth Surface Processes
forests in NE Spain. Forest Ecology and and Landforms, Vol 31, pp. 606-621.
Management 180, 185-198.
FUJIOKA FM. 1987. Fire weather network design: an
EUROPEAN COMMISSION 2002. Forest Fires in Europe – update. In '9th Conference on fire and forest
2001 fire campaign – Official Publication of the meteorology'. American Meteorological Society. pp.
European Commission, S.P.I.02.72 EN. 165-169. (San Diego, California.
JRC), Institute for Environment and Sustainability XANTHOPOULOS G 2001. Contribution of log erosion
(IES), 2005) - CORINE Land Cover updating for the barriers to soil protection and vegetation recovery
year 2000. IMAGE2000 and CLC2000. Products after wildfire in a Pinus halepensis forest, in Greece.
and Methods. Maria Vanda NUNES DE LIMA (Ed.. In Proceedings of the International Conference,
EUR 21757 EN, ISBN 92-894-9862-5. Forest Research: a Challenge for an Integrated
European Approach, Thessaloniki, Greece,
Radoglou K. (ed.), 1:311-316.
Resources, CD-Rom. European Environtmental
1996. Spatial point pattern analysis and its
EUROSTAT, 1999. Statistical references. OYce for
application in geographical epidemiology.
oYcial publications of the European Communities,
Transactions of the Institute of British Geographers
No.2/1999, Luxembourg.
FAO 1986. Wildland fire managemente terminology.
Report nº 70, FAO Forestry Paper, Roma. M-99.
Beneficial and detrimental effects of heating on soil
ISBN 92-5-0024207.
properties. In: Fire in ecosystem dynamics, (J.G.
F.A.O. 1988. Soil map of the world 1:500.000 scale. Goldamrner and M.J. Jenkins Eds. SPB Academic
F.A.O. Rome. Publishing 95-102.
WALSH RPD 1997. Sediment and solute yield in the primary factor controlling erosion in burnt soils?
forest ecosystems affected by fire and rip-ploughing In: Sala, M., Rubio, J.L. Eds. , Soil Erosion and
techniques, central Portugal: a plot and catchment Degradation as a Consequence of Forest Fires.
analysis approach. Physical Chemistry of the Earth Barcelona, p. 16.
22, 309-314.
FOLKMAN WS. 1965. Forest fires as accidents: an CIOMPI S, LLOVET J. 2001. Effects of land use and
epidemiological approach to fire prevention eventual fire and soil erodibility in dry Mediterranean
research, 15. Berkeley, California: Pacific Southwest conditions. Forest Ecology and Management 147,
Forest and Range Experiment Station. USDA. 15-23.
FOLKMAN WS. 1973. Fire prevention in Butte County, GOBIN A, KIRKBY C, GOVERS G 2003. Pan-European
California: their knowledge and attitudes regarding soil erosion risk assessment. European Commision
forest fire prevention., 32. Berkeley, California: funded fifth framework project.
USDA, Forest Service Pacific Southwest Forest
GOFAS A 2001. Εγχειρίδιο Δασοπροστασίας. Εκδόσεις
Range Experiment Station. PSW-98.

D-08-06.doc 106

GONÇALVES MZ, LOURENÇO L. 1990. Meteorological INBAR M, TAMIR M, WITTENBERG L. 1998. Runoff and
index of forest fires risk in the Portuguese mainland erosion processes after a forest fire in Mount
territory. Pages B.07-01 - B.07-14 in I International Carmel, a Mediterranean area. Geomorphology 24,
Conference on Forest Fire Research., Coimbra. 17-33.
GRANGER K. 1998. Developing an understanding of JACKSON RD, IDSO SB, REGINATO RJ, PINTER PJ. 1981.
urban geohazard risk. Australian Journal of Canopy temperature as a crop water stress
Emergency Management 13(4):13–17. indicator. Water Resources Research, 17, 1133–
framework for comparative landscape analysis and JACQUEMOUD S, USTIN SL, VERDEBOUT J, SCHMUCK G,
evaluation based on land cover data, with an ANDREOLI G, HOSGOOD B. 1996. Estimating leaf
application in the Madrid region (Spain), Landscape biochemistry using the PROSPECT leaf optical
and Urban Planning, 55 pp. 257-270. properties model. Remote Sensing of Environment,
56, 194–202.
HARDY CC. 2005. Wildland hazard and risk: problems,
definitions and contexte. Forest ecology and JACQUEMOUD S, BACOUR C, POILVE H, FRANGI JP. 2000.
management. Comparison of four radiative transfer models to
simulate plant canopies reflectance: Direct and
HARDY CC, BURGAN RE. 1999. Evaluation of NDVI for
inverse mode. Remote Sensing of Environment, 74,
monitoring live moisture in three vegetation types of
471– 481.
the Western U.S. Photogrammetric Engineering and
Remote Sensing, 65, 603– 610. JOLY C, SUSINI F et al. 1999. Analysis of the occurrence
of fire patterns in a limited area Forest fires: Needs
and Innovations, Athens, Greece.
Performance of grass filters under laboratory and
field conditions. Transactions of the American JONES AR, TAYLOR ML, BERTRAND AL. 1965. Some
Society of Agricultural Engineers 27, 1321-1331. human factors in woods burning, 28. Baton Rouge,
La.: La. State University Agricultural Experiment
Station Bulletin Nº. 601.
aggregation index (AI) to quantify spatial patterns of
landscape, Landscape Ecology, 15 pp. 591-601. KALABOKIDIS KD, KONSTANTINIDIS P, VASILAKOS C.
2002. GIS analysis of physical and impact on
wildfire patterns. In Proceedings of IV International Conference on Forest fire research and wild land
is.jsp?LIEUID=DEPT06 fire safety, Coimbra. (D.X. Viegas, Ed.) KALLURI SNV, TOWNSHEND JRG, DORAISWAMY P. 1998.
rofeux. A simple single layer model to estimate transpiration
from vegetation using multispectral and (base de meteorological data. International Journal of Remote
données Prométhée) Sensing, 19(6), 1037– 1053.
KEETCH J, BYRAM G. 1968. A Drought Index for forest
Strength and persistence of fire-induced soil fire control. Research Paper. SE-38, USDA Forest
hydrophobicity under ponderosa and lodgepole Service, Southeastern Forest Experiment Station,
pine, Colorado front range. Hydrological Processes
Asheville, NC.
15, 2877-2892.
KERDILES H. 1997. Software for processing AVHRR
ICONA. 1993. Manual de operaciones contra incendios data for the Communities of Europe (SAPCE):
forestales. 5.1/65, ICONA, Madrid.
Algorithms used in SPACE version 2. Part 1:
IGIS 1997. FRAGSTATS*ARC User’s Manual Calibration, atmospheric correction and cloud /fragstatsarc detection. Internal document at ARIS Unit, SAI, JRC
evolution of the NDVI as an indicator of forest fire KIRKBY M. 2001. MEDRAP Athens December 2001
danger. International Journal of Remote Sensing Physical Aspects of Soil Erosion Control. page 11.
IMESON AC, VERSTRATEN JM, VAN MULLIGEN EJ, occurrence patterns at landscape level: beyond
SEVINK J. 1992. The effects of fire and water positional ac-curacy of ignition points with kernel
repellency on infiltration and runoff under density estimation methods. Poster presentation in
Mediterranean type forest. Catena 19, 345-361. 2002 World Conference on Natural Resources
Modeling, “Modeling Natural and Biotic Resources
INBAR M, WITTENBERG L, TAMIR M. 1997. Soil Erosion in a Changing Planet”, 23-26 June 2002, Sigri,
and Forestry Management After Wildfire in a Lesbos.
Mediterranean Woodland, Mt. Carmel, Israel.
International Journal of Wildland Fire 7 (4), 285-294.

D-08-06.doc 107

occurrence patterns at landscape level: beyond BROWN. J.F. 1991. Development of a Land-Cover
positional accuracy of ignition points with kernel Characteristics Database for the Conterminous U.S.
density estimation methods. Natural Resource Photogrammetric Engineering and Remote Sensing
Modeling 17(4): 359-375. 57:1453-1463.
2005. Modelling Wildland Fire Occurrence in annual people-caused forest fire ocurrence in
Southern Europe by a Geographically Weighted Ontario. Paper read at 8th National Conference on
Regression Approach. Paper presented at 5th Fire and Forest Meteorology, April 29-May 2, at
International Workshop on Remote Sensing and Detroit, Michigan.
GIS Applications to Forest Fire Management: Fire
MALDENOFF, D. J. AND BAKER, W.L. 2000): Spatial
Effects Assessment, 16-18 June 2005, at Zaragoza,
Modelling of Forest Landscape Change, Cambridge
University Press.
MANDALLAZ, D., AND YE. R. 1997. Prediction of forest
(OLS) and Local (Geographically--Weighted
fires with Poisson models. Canadian Journal of
Regression) Models for Explaining Long-term
Forest Research 27:1685-1694.
Wildland Fire Occurrence Patterns in Southern
Europe. Journal of Geophysical Research (in press) Manuel d’utilisation de SYSTAT 8.0 1998. Régression
logistique : p 517 - 584
KRAMMES JS, DEBANO LF. 1965. Soil Wettability: a
neglected factor in watershed management. Water MARQUES, M.A., AND MORA E. 1998. Effects on erosion
Resources Research 1, 283–286. of two post-fire management practices: clear-cutting
versus non-intervention. Soil and Tillage Research
KUTIEL P, INBAR M. 1993. Fire impact on soil nutrients
45, 433-439.
and soil erosion in a Mediterranean pine forest
plantation. Catena 20, 129–139. MARTELL, D.L., OTUKOL, S. AND STOCKS, B.J. 1987. A
logistic model for predicting daily people-caused
LAIRD JR, HARVEY MD. Complex response of a
forest fire occurrence in Ontario. Canadian Journal
chaparral drainage basin to fire. In: Hadley, R.F. Ed.
of Forest Research 17:394-401.
, Drainage Basin Sediment Delivery, Vol. 159. IAHS
Publ., pp. 165–183. MARTELL, D.L., BEVILACQUA, E. AND STOCKS, B.J. 1989.
Modelling seasonal variation in daily people-caused
LAL R, 2004. Soil carbon sequestration impacts on
forest fire occurrence. Canadian Journal of Forest
global climate change and food security. Science
Research 19 (12):1555-1563.
304 (5677), 1623– 1627.
LE BISSONNAIS, Y. 1996. Aggregate stability and
J. 1998. Variations de la composition chimique des
assessment of soil crustability and erodibility : I.
eaux de drainage et de ruissellement et
Theory and methodology 47, 425-437.
modifications des caractères agronomiques des sols
LEBLON, B. 2001. Forest wildfire hazard monitoring après incendie de forêt dans le bassin versant du
using remote sensing: A review. Remote Sensing Rimbaud (Var, France. Géomorphologie 3, 215-232.
Reviews, 20(1), 1 –57.
LETEY, J. 2001. Causes and consequences of fire- Estimating human risk factors in Wildland fires in
induced soil water repellency. Hydrological Spain using logistic regression. II International
Processes 15, 2867-2875. Symposium on Fire Economics, Planning and
Policy: A Global Vision. Córdoba, Spain. 19-22 de
LEVINE, N. 2002. CrimeStat II: A Spatial Statistics
Abril, 2004. 15 pp.
Program for the Analysis of Crime Incident
Locations (version 2.0), Ned Levine and Associates MARTINEZ, J, ROMERO, R. AND ECHEVARRIA, P. 2006):
Annandale, VA and The National Institute of Justice Valoración paisajística y ecológica de la Comunidad
Washington, DC. de Madrid: su integración en un índice sintético de
riesgo de incendios forestales. XII Congreso
Nacional de Tecnologías de la Información
FRANZREB, K.E. 2002. Spatial autocorrelation and
Geográfica. Granada. CDRom.
autoregressive models in ecology. Ecological
Monographs 72 (3): 445-463. MATAIX-SOLERA, J., AND DOERR, S.H. 2004.
Hydrophobicity and aggregate stability in calcareous
LIGDI, E., AND MORGAN, R.P.C. 1995. Contour grass
topsoils from fire-affected pine forests in south-
strips: a laboratory simulation of their role in soil
eastern Spain. Geoderma 118, 77-88.
erosion control. Soil Technology 8, 109-117.
MCARTHUR, A.G. 1967. Fire behaviour in eucalypt
LOMBORG, B., 2001. The Skeptical Environmentalist:
forests. Leaflet 107, Commonwealth of Australia
Measuring the Real State of the World. Cambridge
Forest and Timber Bureau, Canberra, Australian
University Press.
Capital Territory.

D-08-06.doc 108

spatial pattern analysis program for quantifying McArthur's fire-danger meters expressed as
landscape structure. Portland, Department of equations. Australian Journal of Ecology 5:201-203.
Agriculture, Forest Service.
ORIEUX, A. 1979. Conditions météorologiques et
MCGARIGAL, K., MARKS, B., HOLMES, C. AND ENE, E. incendies di forêts en région mèditerranéenne. Note
2002): Fragstats 3.3. Spatial Pattern Analysis Technique du Service Météorologie Métropolitain
Program for Quantifying Landscape Structure, Ministère des Transports, Direction de la
Department of Natural Resources Conservation, Météorologie.
University of Massachusetts.
MONTANARELLA, L., 2003. Soil erosion risk in Europe. MARTELLACCI., C. 1992. Un indice meteorologico di
European Commision, Joint Research Centre, rischio per incendi boschivi. Bollettino Geofisico
Institute for Environment and Sustainability. T.P. 15:49-62.
280, 21020 Ispra (VA) ITALY.
PALTRIDGE, G.W., AND BARBER, J. 1988. Monitoring
grassland dryness and fire potential in Australia with
MORAN, M.S., CLARKE, T.R., INOUE, Y., AND VIDAL, A. NOAA/AVHRR data. Remote Sensing of
1994. Estimating crop water deficit using the relation Environment, 25, 381– 394.
between surface-air temperature and spectral
PAPAMIXOS, T., N., 1996. Δασικά Εδάφη, Σχηματισμός,
vegetation index. Remote Sensing of Environment,
Ιδιότητες, Συμπεριφορά. Aristotle University of
49, 246– 263.
Thessaloniki, Υπηρεσία Δημοσιευμάτων.
PAUSAS, J.G. 1997. Resprouting of Quercus suber in
M.G., AND LONG, D.G. 2001. Mapping fire regimes
NE Spain after fire. Journal of Vegetation Science 8,
across time and space: Understanding coarse and
fine-scale fire patterns. International Journal of
Wildland Fire, Vol. 10: 329-342. PAUSAS, J.G. 2004. Changes in fire and climate in the
eastern Iberian peninsula (Mediterranean basin.
Climatic Change 63, 337-350.
geometric reconstruction of boreal type forest PAUSAS, J.G, AND VALLEJO, V.R. 1999. The role of fire
stands at single tree level for forest and wildland fire in European Mediterranean Ecosystems. In Remote
management. Remote Sensing of Environment, 92, Sensing of Large Wildfires in the European
353-362. Mediterranean Basin, Chuvieco E. (ed.), Springer-
Verlag, 3-16.
1999. Modeling hydrology and sediment transport in PEW, K.L. AND LARSEN, C. 2001. “GIS analysis of spatial
vegetative filter strips. Journal of Hydrology 214, and temporal patterns of human-caused wildfires in
111-129. the temperate rain forest of Vancouver island,
Canada.” Forest ecology and management 140: 1-
NAVEH, Z. AND LIEBERMAN, A. 1994): Landscape
Ecology. Theory and Application, Nueva York,
Springer-Verlag. PHILLIPS, C.B., AND NICKEY. 1978. The concept of
"spatial risk" and its application to fire prevention.
NEARING, M.A., PRUSKI, F.F., O’NEAL, M.R., 2004.
Fire Management Notes 39 (4):7-8-19.
Expected climate change impacts on soil erosion
rates: a review. Journal of Soil and Water PINAYA, I, SOTO B, ARIAS M, AND DÍAZ-FIERROS F. 2000.
Conservation 59 (1), 43– 50. Revegetation of burnt areas: relative effectiveness
of native and commercial seed mixtures. Land
Degradation and Development 11, 93-98.
FOLLIOTT, P.F. 1999. Fire effects on belowground
sustainability: a review and synthesis. Forest PROSPER-LAGET, V., DOUGUÉDROIT, A., AND GUINOT,
Ecology and Management 122, 51-71. J.P. 1995. Mapping the risk of forest fire occurrence
using NOAA satellite information. EARSEL
NEARY, G. D., 2004. An Overview of Fire Effects on
Advances in Remote Sensing, 4(3), 30– 38.
Soils. Southwest Hydrology, September/October
2004. PROSSER IP, AND WILLIAMS L. 1998. The effect of
wildfire on runoff and erosion in native Eucalyptus
forest. Hydrological Processes 12, 251-265.
2000. Chapter 3. Indexing Colorado Watersheds to RENARD, K.G., FOSTER, G.R., WEESIES, G.A., MCCOOL,
Risk of Wildfire. Co-published simultaneously in D.K., YODER, D.C., 1997. Predicting soil erosion by
Journal of Sustainable Forestry, 11(1/2), 159-175; water: a guide to conservation planning with the
and: Mapping Wildfire Hazards and Risks (ed: Neil Revised Universal Soil Loss Equation. Agricultural
Sampson, R. Dwight Atkinson, and Joe W. Lewis), Handbook, vol. 703. U.S. Department of Agriculture
Food Product Press, an imprint of The Haworth 404 pp.
Press, Inc., 2000, , 35-55.

D-08-06.doc 109

USTIN, S. L. 2003. Modelling airborne laser 2. Characteristics of the Study Area, Data Utilized,
scanning data for the spatial generation of critical and Modeling Approach. Co-published
forest parameters in fire behavior modelling. simultaneously in Journal of Sustainable Forestry,
Remote Sensing of Environment, 86, 177– 186. 11(1/2), 159-175; and: Mapping Wildfire Hazards
and Risks (ed: Neil Sampson, R. Dwight Atkinson,
and Joe W. Lewis), Food Product Press, an imprint
l'élaboration d'un système d'information sur les
of The Haworth Press, Inc., 2000, , 15-33.
paysages du monde.” Cahiers des sciences
humaines 32: 823-842. SAN MIGUEL-AYANZ, J., ANNONI, A. AND SCHMUCK. G.
1998. The use of satellite imagery for retrieval of
information on wildfire damage in Mediterranean
landscapes. in ERIM’98, International Symposium
1995): A factor analysis of landscape pattern and
on Remote Sensing of Environment: Information for
structure metrics, Landscape Ecology, 10, pp. 23-
sustainability, Tromso.
RIVA, J.D.L., PEREZ-CABELLO, F. et al. 2004. “Mapping
LIBERTA, G. 2002. Towards a coherent forest fire
Wildfire occurrence at regional scale.” Remote
information system in Europe: The European Forest
sensing of environnement 92: 363-369.
Fire Information System (EFFIS. in D. X. Viegas,
ROBICHAUD, PR, HUNGERFORD RD. 2000. Water editor. IV International Conference on Forest Fire
repellency by laboratory burning of four northern Research and 2002 Wildland Fire Safety Summit.
Rocky Mountain forest soils. Journal of Hydrology ADAI, University of Coimbra, Coimbra, Portugal.
231–232: 207–219.
ROBICHAUD, P. 2004. Post-fire rehabilitation treatments: JOSÉ U. PAÚL, 2000. A comparative study of
are we learning what works ? Southwest hydrology, vegetation indices to assess land cover change after
Watersheds on Fire 3, 20-21. forest fires. Remote Sensing for Earth Science,
Ocean, and Sea Ice Applications, 3868, pp.232-240.
RODRIGUEZ Y SILVA, F. 2002. Visual Forest Fire
Weather Index (vffwi) a mathematical model for the SCOTT, D.F., VAN WYK, D.B., 1990. The effects of
prediction of forest fires weather danger in wildfire on soil wettability and hydrological behaviour
mediterranean ecosystems. In 'IV Conference on of an afforested catchment. Journal of Hydrology
Forest Fire Research'. ADAI, (Luso,Portugal. 121, 239–256.
ROTHERMEL, R.C. 1972. A Mathematical Model for SEBASTIÁN LÓPEZ, A. 2004. Modelización a escala
Predicting Fire Spread in Wildland Fuels. Res. Pap. europea del riesgo de incendio forestal, desde la
INT-115. pp. 40. USDA Forest Service. (Ogden, perspectiva dinámica y estructural. Universidad de
Utah. Valladolid, Valladolid.
SACKETT., S.S. 1986. Modelling moisture content of AYANZ, J. 2001. Two models for a structural forest
fine dead wildland fuels: input to BEHAVE fire fire risk index at a regional scale. in 21st EARSeL
prediction system. Research Paper INT-359, USDA Workshop Remote sensing and GIS applications to
Forest Service Intermountain Research Station, forest fire management: New Methods and Sensor,
Odgen, Utah. Paris. France.
1997. Soil profile characteristics influencing runoff BURGAN, R.E. 2002. Integration of satellite sensor
and soil erosion after forest fire: a case study data, fuel type maps and meteorological observation
(Valencia, Spain. Soil Technology 11, 67-78. for evaluation of forest fire risk at the pan-European
scale. International Journal of Remote Sensing
RUBIO, J. L., BOARDMAN J., HELMING K., 2006. Soil 23(13): 2713-2719.
erosion across Europe: Research approaches and
perspectives. Catena, Article in Press. SEVINK, J., IMESON, A.C., VERSTRATEN, J.M., 1989.
Humus form development and hillslope runoff, and
SADAHIRO, Y. 2003. Stability of the surface generated the effects of fire and management under
from distributed points of uncertain location. mediterranean forest in NE-Spain. Catena 16, 461–
International Journal of Geographical Information
Science 17:139-156.
SALAS, F.J., AND CHUVIECO, E. 1994. GIS applications TERRY, J.P., and WALSH, R.P.D. 1993. Wildfire
to forest fire risk mapping. Wildfire 3:7-13. impacts on soil erosion and hydrology in wet
Mediterranean forest, Portugal. International Journal
of Wildland Fire 3, 95-110.

D-08-06.doc 110

GONÇALVES, A.J.B., AND WALSH, R.P.D. 1994. 1995. An experimental study of the influence of
Limiting soil loss after forest fire in Portugal: the traditional slash-and-burn practices on soil erosion.
influence of different post-fire timber clearing Catena 24, 13-23.
practices. In Proc. 2nd Int. Conf. Forest Fire
Research II, 1161-1170.
1986. On the nature of models in remote sensing.
SHAKESBY, R.A., BOAKES, D.J., COELHO, C.O.A., Remote Sensing of Environment, 20, 121–140.
SWANSON, F.J., 1981. Fires and geomorphic processes.
Limiting the soil degradational impacts of wildfire in
In: Proceedings, fire regimes and ecosystems
pine and eucalyptus forests in Portugal. Applied
conference, Honolulu. Gen. Tech. Rep. WO-26
Geography 16, 337-355.
USDA, Washington, DC, pp. 401–420.
TADESSE, L.D. and MORGAN, R.P.C., 1996. Contour
The erosional impact of soil hydrophobicity: current
grass strips: a laboratory simulation of their role in
problems and future research directions. Journal of
erosion control using live grasses. Soil Technology
Hydrology 231-232, 178-191.
9, 83-89.
SILVA, F.C.O.R. 2002. "Visual forest fire weather index
TERRES, J.M. 1999. The Crop Growth Monitoring
(VFFWI)" a mathematical model for the prediction of
System implemented by JRC/ARIS unit for the
forest fire weather danger in Mediterranean
information needs of the EC DG VI - Agriculture.
ecosystems. Forest fire research and wild land fire
Proceedings of the international symposium
safety, Coimbra.
Modelling Cropping Systems - European Society for
SILVERMAN, B.W. 1986. Density Estimation for Statistics Agronomy. EX 18746/C. ORA/PRO 60194R
and Data Analysis. Chapman and Hall, London,
1999. Nutrient losses in eroded sediment after fire in
SIMARD, A.J. 1968. The moisture content of forest fuels. eucalyptus and pine forests in the wet
A review of the basic concepts. Information Report: Mediterranean environment of northern Portugal.
FF-X-14. pp. 47. Forest Fire Research Institute. Catena 36, 283-302.
(Ottawa, Ontario.
SMITH, S.J., WILLIAMS, J.R., MENZEL, R.G., COLEMAN, 2000. Solutes in overland flow following fire in
G.A., 1984. Prediction of sediment yield from eucalyptus and pine forests, northern Portugal.
Southern Plains grasslands with the Modified Hydrological Processes 14, 971-985.
Universal Soil Loss Equation. Journal of Range
THOMPSON, W.A. 2000. Using forest fire hazard
Management 37 (4), 295– 297.
modelling in multiple use forest management
SNEEUWJAGT, R.J., AND PEET, G.B. 1985. Forest fire planning. Forest Ecology and Management
behaviour tables for Western Australia. Perth: 134:163-176.
Western Australian Department of Conservation and
THORNES, J.B., 1985. The ecology of erosion.
Land Management.
Geography 70, 222–234.
SOL, B. 1990. Estimation du Risque Météorologique
TRABAUD, L. 1976. Inflamabilité et combustibilité des
D'incendies de Forêts dans le Sud-Est de la France.
principales espèces des garrigues de la region
Revue Forestière Française 17:263-271.
mediterranéenne. Oecol. Plant. 11: 117–136.
SOL, B. 1999. A European program improves the
operational fight against forest fires: MINERVE
African land-cover classification using satellite data.
Program and meteorological fire danger in
Science, 227: 367-374.
southeastern France, in DELFI Proceedings,
Athens, pp. 185-192. VACCA, A., LODDO, S., OLLESCH, G., PUDDU, R., SERRA,
G., Tomasi, D. and Aru, A. 2000. Measurement of
SOLER, M., SALA, M., 1992. Effects of fire and of
runoff and soil erosion in three areas under different
clearing in a Mediterranean Quercus ilex woodland:
land use in Sardinia (Italy. Catena 40, 69-92.
an experimental approach. Catena 19 3r4 , 321–
332. VALETTE, J.C. 1990. “Inflammabilité des espèce
forestières méditerranéennes: conséquences sur la
combustibilité des formations forestières.” Revue
1991. Influence of antecedent soil moisture on
forestière française - espaces forestiers et incendies
pedological effects of fire. In: Sala, M., Rubio, J.L.
numéro spécial 1990.
Eds., Soil Erosion and Degradation as a
Consequence of Forest Fires. Barcelona.
SOTO, B. and DÍAZ-FIERROS, F. 1998. Runoff and soil
erosion from areas of burnt scrub: comparison of
experimental results with those predicted by the
WEPP model. Catena 31, 257-270.

D-08-06.doc 111

VALETTE, J.C. 1992. Inflammabilities of Mediterranean VIEDMA, O., MELIA, A, J., SEGARRA, D., GARCÍA-HARO,
species. In Forest Fire Risk and Management: J., 1997. Modelling rates of ecosystem recovery
European School of Climatology and Natural after fires by using Landsat TM data. Remote
Hazards Course. Halkidiki, Greece, 27 May – 4 Sensing of Environment, 61(3), pp. 383-398.
June 1992. Edited by P. Balabanis, G. Eftichidis,
VIEGAS, D.X., ed. 1998. Proceedings of the III
and R. Fantechi. European Commission, Halkidiki,
International Conference on Forest Fire Research
Greece. pp. 51–64.
and 14th Conference on Fire and Forest
VALLEJO, R, AND ALLOZA J.A. 1998. The restoration of Meteorology. Held in Luso, Coimbra, Portugal, on
burned lands: the case of eastern Spain. Large November 16-20th: ADAI - Associaçao para o
Forest Fires, 91-108. Desenvolvimento da aerodinâmica Industrial - and
University of Coimbra.
VAN DER GOOT, E, 1997. Technical description of
interpolation and processing of meteorological data VIEGAS, D.X. 1998b. Fuel moisture evaluation for fire
in CGMS. Internal document at SAI, JRC Ispra behaviour assessment. In G. Eftichidis, P. Balabaris,
and A. Ghazi (Eds.), Advanced study course on
VAN WAGNER, C.E. 1982. Initial moisture content and
wildfire management. Final Report, Marathon (pp.
the exponential drying process. Canadian Journal of
81– 92. Athens: European Union.
Forest Research 12:90-92.
VIEGAS, D.X., ed. 2002. Proceedings of the IV
VAN WAGNER, C.E. 1987. Development and structure of
International Conference on Forest Fire Research
the Canadian Forest Fire Weather Index System.
and 2002 Wildland Fire Safety Summit. Held in
Technical Report 35. pp. 37. Canadian Forestry
Coimbra, Portugal, on 18-23 November: ADAI -
Service. (Ottawa, Ontario.
Associaçao para o Desenvolvimento da
VASCONCELOS, M.J.P., SILVA, S., TOMÉ, M., ALVIM, M., aerodinâmica Industrial - and University of Coimbra,
and J.M.C. PEREIRA. 2001. Spatial prediction of fire Portugal.
ignition probabilities: comparing logistic regression
and neural networks. Photogrammetric Engineering
Moisture content of fine forest fuels and fire
and Remote Sensing 67 (1):73-83.
occurrence in central Portugal. International Journal
VEGA GARCÍA, C., WOODARD, P.M. TITUS, S.J. of Wildland Fire, 2(2), 69–85.
ADAMOWICZ, W.L. AND LEE, B.S. 1995. A logit model
for predicting the daily occurrence of human caused
SOL. B. 1996. Testing Meteorological Fire Danger
forest fires. International Journal of Wildland Fire 5
Methods in Southern Europe. Pages 571-589 in
13th Conference on Fire and Forest Meteorology,
VEGA GARCÍA, C., LEE, B.S., WOODARD, P.M., AND Lorne - Australia.
TITUS. 1996. Applying neural network technology to
human-caused wildfire occurrence prediction. AI
2001. Estimating live fine fuels moisture content
Applications 10 (3):9-18.
using meteorologically-based indices. International
VELEZ, R et al. 2000 La defensa contra incendios Journal of Wildland Fire, 10, 223–240.
forestales: fundamentos y experiencias. McGraw
VINEY, N.R. 1991. A review of fine fuel moisture
Hill, Madrid
modelling. International Journal of Wildland Fire
VELEZ, R. AND MÉRIDA, J.C. 2002. Forest fire 1(4): 215-234.
management in Spain: Some examples of
systematic analysis of a comprehensive database to
RUFFINONI, C. 1995. Structure and function of buffer
improve effectiveness and efficiency. Forest fire
strips from a water quality perspective in agricultural
research and Wild land fire safety, Coimbra.
landscapes. Landscape and Urban Planning 31,
VELEZ, R., REANO, P., et al. 2002. Forest fire prevention 323-331.
with a target: the rural people. Forest fire research
WALLIS, M.G., AND HORNE, D.J. 1992. Soil water
and wildland fire safety, Coimbra.
repellency. In Advances in Soil Science, Vol. 20,
VERSTRAETE, M.M. AND PINTY, B. 1996. Designing Stewart BA (ed., Springer: New York, 91-146.
optimal spectral indexes for remote sensing
applications. IEEE Transactions on Geoscience and
TERRY, J.P. 1992. Effects of land use management
Remote Sensing, 34(5), 1254– 1265.
practices and fire on soil erosion and water quality in
VIDAL, A., PINGLO, F., DURAND, H., DEVAUX-ROS, C., the Agueda river basin, Portugal. Geoökoplus III,
AND MAILLET, A. 1994. Evaluation of a temporal fire 15-36.
risk index in Mediterranean forest from NOAA
thermal IR. Remote Sensing of Environment, 49,
296– 303.

D-08-06.doc 112


W., 1989. AGNPS: a nonpoint-source pollution
A.D. 1995. Post-fire land use and management and
model for evaluating agricultural watersheds.
runoff responses to rainstorms in northern Portugal.
Journal of Soil and Water Conservation 44 (2), 168–
In Geomorphology and Land Management in a
Changing Environment, McGregor DFM, and
Thompson DA (eds), Wiley,Chichester, 283-308. ZARCO-TEJADA, P.J., RUEDA, C.A., AND USTIN, S.L.
2003. Water content estimation in vegetation with
WARREN, A., 2002. Land degradation is contextual.
MODIS reflectance data and model inversion
Land Degradation & Development 13 (6), 449– 459.
methods. Remote Sensing of Environment, 85, 109–
P.R. 2001. The effects of log erosion barriers on
ZHANG, L.J., SHI, H.J. 2004. Local modeling of tree
post-fire hydrologic response and sediment yield in
growth by geographically weighted regression.
small forested watersheds, southern California.
Forest Science 50 (2): 225-244.
Hydrological Processes 15, 3053-3066.
WORTON, B.J. 1989. Kernel methods for estimating the
utilization distribution in homerange studies. Ecology
Deriving stand age distribution in boreal forests
YOOL, S.R., ECKHARDT, D.W., ESTES, J.E., AND imagery. Remote Sensing of Environment, 91, pp.
COSENTINO., M.J. 1985. Describing the brushfire 405–418.
hazard in southern California. Annals of the
Association of American Geographers 75 (3):417-

D-08-06.doc 113


UMR 6012 Espace CNRS, Equipe GVE, Department of Geography, University of Nice Sophia Antipolis
BP 3209 06204 Nice Cedex 3, France.


The Mediterranean climate is particularly conducive
to large-scale forest fires. The increases in runoff and soil erosion rates
Mean annual precipitation is sufficient to support a observed after a forest fire are due primarily to the
dense mixture of oak, pine, and other forest species destruction of the vegetation cover (PROSSER and
and the warm dry summer season makes both the live WILLIAMS, 1998; WOHLGEMUTH et al., 2001).
and accumulated biomass highly inflammable for up to However, the combustion of the standing vegetation
at least two months of the year (VALLEJO and ALLOZA, and litter layer provokes several changes in soil
1998; PAUSAS and VALLEJO, 1999. properties.
The abandonment of traditional underbrush clearing The study of low severity fires or traditional slash
practices combined with an increase in accidental and and burn methods has confirmed some of the short-
intentional fire ignitions has led to more frequent forest term benefits of forest fires on soil fertility (pH, nutrient
fires throughout much of the Mediterranean region status) and their influence on post-fire regrowth (SOTO
(MORENO and VALLEJO, 1999, in FERNANDEZ et al., et al., 1995; NEARY et al., 1999).
However, the effects are not all beneficial since the
Changing climatic conditions are likely to increase
combustion of the soil organic matter can also lead to
the number of fires and area burned in coming years
lower cation exchange capacities, and high severity
(PAUSAS, 2004).
fires can lead to significant nutrient losses through
The impacts of a major forest fire on runoff and
volatilisation and accelerated erosion (MARTIN et al.,
erosion are well-known, and these include greater peak
1998; NEARY et al., 1999; THOMAS et al., 1999).
flows and soil loss until the return of a vegetation cover.
Two further changes in soil characteristics, the
The first winter following a summer forest fire is the
development or enhancement of hydrophobicity and a
most vulnerable period and runoff and erosion rates
decrease in aggregate stability, have a direct impact on
generally decrease rapidly afterwards (SOTO and DÍAZ-
soil erodibility (ANDREU et al., 2001.
FIERROS, 1998; VACCA et al., 2000).
It is therefore essential that runoff and erosion The formation of soil hydrophobicity, or water
management strategies be implemented in the first repellency, has been described extensively elsewhere
weeks after a fire. (e.g. WALLIS and HORNE, 1992; DEBANO, 2000; DOERR
et al., 2000; DEBANO, 2000b; HUFFMAN et al., 2001), so
Several natural factors determine the impact of a fire
only a few aspects related directly to runoff generation
on post-fire runoff and erosion rates, and these include
and soil erosion will be considered here.
the following: pre-fire vegetation, topography, slope
Water repellency develops naturally in most
aspect, fire severity, changes in soil properties, and
Mediterranean forests (DOERR et al., 1998; MATAIX-
post-fire rainfall (WALSH et al., 1992; RUBIO et al., 1997;
SOLERA and DOERR, 2004).
INBAR et al., 1998; SOTO and DÍAZ-FIERROS, 1998;
During a forest fire, heat from the combustion
THOMAS et al., 2000; DELUIS et al., 2003).
vaporizes organic substances, some of which migrate
Although many of these factors (such as slope or
downwards into the soil where they condense at cooler
rainfall intensity) are common to all soil erosion
temperatures and coat mineral particles (DEBANO,
contexts and need little explanation, the changes in soil
erodibility brought about by the intense heat are
After a forest fire, it is therefore common to find a
particular to forest fire conditions.
thin layer of hydrophobic soil at the surface or within a
For this reason, these will be reviewed briefly before
depth of a few centimetres (DEBANO, 2000; MATAIX-
presenting the erosion control strategy.
SOLERA and DOERR, 2004).
Although the effect of hydrophobicity on runoff and
erosion has clearly been demonstrated at the plot
scale, it has been difficult to demonstrate its importance
for a catchment (SHAKESBY et al., 1993; SHAKESBY et
al., 2000).
One reason for this is the high spatial variability in
hydrophobicity: at the catchment scale, preferential flow
in decayed root channels, cracks, rodent burrows, and
hydrophilic patches may account for a large proportion
of the infiltrated water (IMESON et al., 1992; FERREIRA et
al., 1997; SHAKESBY et al., 2000).

D-08-06.doc 114

The impact of forest fires on soil aggregate stability 12.3 SITE DESCRIPTION
is more ambiguous than for hydrophobicity.
The study area is located in south-east France in
On the one hand, forest fires tend to produce
the Massif des Maures near St Tropez; the
greater water repellency in most forest soils, and this
approximate centre of the burnt area is 43°16’ N, 6° 28’
tends to reduce slaking, thereby increasing aggregate
stability (MATAIX-SOLERA and DOERR, 2004).
The Giscle catchment has a surface area of about
On the other hand, aggregate stability is positively
234 km2 and is composed of two major topographical
correlated with organic matter content, so the
destruction of the organic compounds present in the
The lower portion of the catchment (roughly 25% of
soil tends to diminish aggregate stability (CERDÀ et al.,
the total area) is an alluvial plain occupied by vineyards
1995; GIOVANNINI et al., 2001).
and urban development.
The net effect probably depends on initial organic
The upper portion (about 75%) of the catchment has
matter content, soil texture, fire intensity, and the size
forested hilly terrain dominated by a mixture of
range of aggregates being considered.
Mediterranean oak and pine species.
(MATAIX-SOLERA and DOERR 2004) showed that finer
A major forest fire which occurred at the end of
sieve fractions (< 0.25 mm) are more hydrophobic than
August, 2003, burned more than 2000 ha located at the
larger ones (> 0.25 mm), and that both hydrophobicity
head of the Giscle river and a few of its tributaries.
and aggregate stability are correlated with soil organic
Local authorities were therefore concerned about
matter for different sieve fractions, but this question
the impact of increased runoff on the urbanised areas
requires further research.
downstream as well as the possible increase in
Soils on south-facing slopes tend to have both lower sediment load entering the port.
aggregate stability values and greater erosion rates
Annual precipitation is about 950 mm with rainy
than soils on north-facing slopes (CERDÀ et al., 1995;
seasons in the autumn and spring.
MARQUÉS and MORA, 1998; ANDREU et al., 2001).
The hilly forested zone is underlain by metamorphic
Several reasons can explain this: south-facing gneiss and schists, and the soils are classified as
slopes in the Mediterranean climate tend to have Rankers in the FAO classification.
greater potential evapo-transpiration rates leading to Soil textures for the < 2 mm fraction are typically
sparser vegetation cover and therefore thinner soils about 75% sand, 10% silt, and 15% clay. Mean Weight
and lower soil organic matter contents. Diameter (MWD), measured using the method of Le
Bissonnais 1996), for soils in the valley bottom is 3.28
In addition, the drier conditions on south-facing
mm (std. dev. = 0.07), classifying these soils as highly
slopes may make them more susceptible to forest fires
with more frequent burns and exposure to soil erosion
processes; finally, post-fire recovery on south-facing
slopes is slower than on north-facing slopes (CERDÀ et
al., 1995; PAUSAS and VALLEJO, 1999), so erosion
processes continue for longer.

D-08-06.doc 115

12.4 METHODS - either the fire had passed quickly and charred the
trunks and underbrush but left the crown more or
12.4.1 Mapping soil erosion risk
less intact, or
Since undisturbed forests typically have very low - where the intense combustion of the standing
soil erosion rates, most soil erosion models were vegetation was confined to a small area (dozens of
developed for agricultural contexts and were not square meters)
suitable for the burned forest site. In intense severity zones, the fire had consumed the
Measuring soil erosion and elaborating a regression underbrush and crown entirely over an extensive area.
model were also out of the question due to the lack of
Soil erodibility: soil depth and stoniness distribution
within the catchment are controlled by slope inclination,
After an initial survey of the digital data available,
position within the catena, and aspect.
the dominant factors affecting soil erosion (described
Thin stony soils were generally found on steep
below) were mapped and combined to create an
slopes, convexities, and south-facing slopes, while
erosion map.
deeper less stony soils were found on gentler slopes,
The multiplicative method is described below after a
concavities, and north-facing slopes.
presentation of the factors.
Three classes were defined:
In the field, each layer was mapped on a 1:25,000
- thin and stony,
topographic map; some areas of the burned upper
- of intermediate depth and stoniness, and
catchment were inaccessible except via footpaths,
- deep with few stones.
much of the burned catchment was accessible only via
A soil corer was used along a dozen slope transects
unmade roads (fire prevention access roads and
and thin soils were designated as those that had
private lanes), and only the lower portion of the
depths of < 20 cm.
catchment could be accessed by car.
A soil was considered deep if it was > 50 cm in
Therefore, cycling throughout the catchment and
combining direct on-site observations with panoramic
views from viewpoints using binoculars for remote Ascribing coefficients and combining the soil erosion
sites, carry out most of the mapping. factors.
Roughly 2,000 ha had to be mapped in a few days, Mapping the soil erosion risk was based on the
so each factor was attributed a number of qualitative multiplication of coefficients (equ. 1) ascribed to the soil
classes that could be estimated quickly without time- erosion factors summarized in Table 1.
consuming measurements.
Erosion Index = Slope x Vegetation density x Fire
The methods used are outlined below after a
severity x Soil erodibility (1)
description of the soil erosion factors.
For each factor, a raster layer with a 25 m cell size
12.4.2 Soil erosion factors
was created using geographic information software.
Slope: the importance of slope inclination for soil Each cell within the layer was assigned a coefficient
erosion is well known, and a Digital Elevation Model (Table 1) and the layers multiplied (equ. 1) to provide a
(DEM) of the catchment with a 50 m grid was already theoretical range of values of from 1 to 135.
available, but the resolution quickly revealed itself to be
In the absence of a suitable, functional, and readily
insufficient, so a 25 m DEM was ordered.
available soil erosion model, values for the coefficients
Pre-fire vegetation: pre-fire density was estimated were attributed arbitrarily according to personal
from the standing charred trunks, and three categories experience and field observations.
were included – bare, low density, and medium density. Estimating coefficients for slope, vegetation and fire
Rock outcrops were devoid of any vegetation and severity was relatively straightforward.
the distinction between low and medium density was The soil erodibility coefficients were more
based mainly upon the ease of walking through the problematic: shallow stony soils generate more runoff
charred forest. but surface material is less mobile due to the high
In addition to the mulch effect provided by fallen stoniness, and in many cases, the bare bedrock.
pine needles described above (SHAKESBY et al., 1993; Deep soils with fewer stones may have higher
SHAKESBY et al., 1994), standing dead vegetation may infiltration rates, but they are more erodible, particularly
favour infiltration (near stems and in burned roots) and with respect to the weak aggregate stability and
slow runoff velocity, so the net expected effect is lower possible presence of hydrophobicity.
erosion where vegetation stands are more dense. Therefore, we decided to increase the value of the
coefficient with decreasing stone content, as can be
Fire severity: this is a major factor influencing runoff
seen in Table 1.
and soil erosion rates (ROBICHAUD and WALDROP, 1994,
cited in LETEY, 2001; RUBIO et al., 1997; PROSSER and
WILLIAMS, 1998).
Two severity categories were included: variable and
Variable severity sites were designated as areas

D-08-06.doc 116

12.4.3 Partial validation of the soil erosion risk 12.5 RESULTS

12.5.1 Distribution of the soil erosion factors
Once the soil erosion risk map was completed, the
Figures 1 to 4 show the distribution of the four
catchment was divided into sub-catchments and the
factors of soil erosion in the catchment.
percentage area in the high and very high risk
More than 30% of the slopes in the study area have
categories (Table 2) was calculated for each sub-
an inclination greater than 20% (Figure 1), showing the
natural vulnerability of the zone to erosion processes.
On 2 December, 2003, a rainfall event generated
In addition, many of the steeper slopes were those
significant runoff in the catchment: peak rainfall
most affected by the fire (Figure 2).
intensity values were not available but roughly 35 mm
fell in the morning, and about 35 mm fell in the Typically, pre-fire vegetation densities (Figure 3)
preceding five days (measured at about five km from were low on convexities or south-facing steep slopes
the fire site. with shallow stony soils (Figure 4).
Runoff samples from most of the major streams Conversely, deeper soils with a denser pre-fire
were obtained manually by plunging a 0.5 L bottle into vegetation cover tend to be concentrated in the
the stream during the storm. southeast corner of the map where slopes are gentle.
Replicate suspended sediment samples from the
main channel and selected tributaries were collected 12.5.2 The soil erosion risk map
within a period of about 10 minutes so rainfall/runoff Values resulting from the multiplication of the factors
conditions were considered comparable. ranged from 1 to 90, and were subdivided into soil
The amount of suspended sediment transported erosion classes according to breaks between peaks in
through the channel depends on both sediment the histogram distribution.
concentration and discharge, but the timing imposed by Table 2 presents the values for each class, and
sampling simultaneously and the difficult field Figure 5 shows the results.
conditions made it impossible to measure discharge Each class tends to correspond to a typical
accurately. environment (Table 2.
Therefore, we only measured suspended sediment
concentrations. As stated above, the erosion model used is
An order of magnitude based on visual estimates extremely simple and the subjective definition of the
would give discharges of about 10 m3 s-1 for the main coefficients is open to some debate.
channel and about 20-30% of that value for the But the objective was to provide a quick spatial
tributaries. representation of general trends and not a quantitative
It was therefore possible to compare the stream estimate of soil loss.
suspended sediment load to the percentage of the So, it is unlikely that a more time-consuming
catchment in the high and very high-risk categories. modelling approach would have significantly altered the
For comparison, we sampled a stream draining a spatial distribution of the erosion classes identified.
vineyard sub-catchment and an unburned forest area. The authors had been working in the catchment for
three years prior to the fire, and soil erosion in the
vineyards had been mapped using both the Revised
Universal Soil Loss Equation (RUSLE) and a
regression equation based on measurements carried
out in a reference area.
Neither of these models, designed for agricultural
contexts, could be applied to burned forest sites, and
the data requirements for more complex deterministic
models were unrealistic within the time frame allowed
for the post-fire management proposals.

D-08-06.doc 117

12.5.3 Partial model validation
Identifying critical sites for post-fire erosion control
Significant runoff occurred during the 2 December,
methods depends on an accurate spatial estimation of
2003, storm (Figs 6, 7, 8. Stream discharge samples (2
high-risk zones.
replicates) were collected for three of the four main
Several factors contribute to make this task
somewhat more difficult than in a typical agricultural
Sampling locations are shown in Figure 5: Tourré
- measured just West of the “Pont de Bois” before
These include the following:
entering the main Giscle channel
- The short intervention time between the forest fire
- the subcatchment surrounding the “Val de Gilly”
and the installation of post-fire erosion control
sampled just before entering the Giscle channel
methods. Identifying critical areas must be done
- the main channel sampled at the “Pont Gué”
within a period of weeks after the fire for the
(sampled upstream of other tributaries. The Périer in
methods to be efficient during the first winter rains.
the southern part of the catchment could not be
- The lack of spatial data on soil properties that
sampled in similar rainfall/runoff conditions, so it
contribute to enhance or reduce the erosion risk
was not analysed. In addition, samples for the
(soil depth, stoniness, hydrophobicity, aggregate
burned north-facing slope
stability…. Data for soil properties in agricultural
- draining into the Giscle main channel just South of
areas tend to be more abundant than for forests
the “Pont Gué”; samples of runoff from a vineyard
since soil productivity depends on these properties.
and an unburned forest were collected for
Data for forested areas are scarce.
comparison purposes.
- In the Mediterranean region, forested slopes tend to
For each of the three sub-catchments sampled and be particularly steep since flatter areas are occupied
the burned north-facing slope, stream suspended by agricultural activities, including vineyards, fruit
sediment concentration was plotted against the sum of and olive groves, and cereal crop. After a major
the percentage area in the high and very high-risk forest fire, slope angle is perhaps the single most
erosion classes (Figure 9). important factor determining erosion rates. DEM
The numbers in Figure 9 correspond to sampling spatial resolutions should be in the order of 25 m to
locations described in Figure 5. Sediment concentration accurately represent the terrain.
values represent the means of 2 samples. - Pre-fire vegetation density and fire intensity are
Differences in concentration between samples were spatially variable and along with slope angle they
never greater than 10% except for the Gilly, for which it play a major role in determining soil erosion risk.
was 31%.
The case study presented above describes an
Although there are only four points, there is a
operational strategy to map erosion risk at the scale of
general trend for the sediment concentration to
a large catchment.
increase with increasing area in the High and Very High
Work is currently underway to investigate two future
erosion classes.
directions of research.
The exception is the Giscle (sample 3) which is The first is to elaborate a pre-fire strategy that would
located immediately downstream of a long straight shorten the post-fire reaction time.
section with natural sediment deposition. To what extent can planners prepare for the
Therefore, the low suspended sediment consequences of a forest fire before it occurs in order
concentration of the Giscle was attributed to natural to respond more quickly to the potential erosion risk?
deposition in the stream bed upstream of the sampling The second is to determine how the spatial scale of
area. the data and model affect output.
Sediment concentrations for the vineyard (5) and Soil erosion risk maps have been produced to
unburned forest were 1.64 g l-1 and 0 g l-1, respectively. identify potential high risk areas at the European scale
for essentially arable land.
The question now asked is “Can a similar approach
be used for post-fire soil erosion risk?”.

D-08-06.doc 118

12.7 CONCLUSIONS In the future, soil erosion modelling should be

accompanied by the development of soil conservation
Major forest fires in the Mediterranean environment
models where high risk sites are evaluated for
occur regularly and might become more frequent in the
suitability to different erosion control methods.
future as the climate changes.
Increasing the efficiency of the approach and the
A method for choosing appropriate sites for the
effects of varying spatial scale are two further research
erosion control measures is based firstly on an
directions that are currently being pursued.
assessment of the spatial distribution of erosion risk.
It should be noted that although the soil erosion risk
map produced in the first stage of the method is useful
in identifying high risk areas, it does not represent in
itself a sufficient tool for developing a soil conservation

D-08-06.doc 119


Figures 1 and 2: Slope classes (left) and Fire Severity (right)

Figures 3 and 4: Estimated pre-fire vegetation density (left) and distribution of soil characteristics (right)

Figure 5: Soil erosion risk map showing the erosion risk categories.

D-08-06.doc 120

Figure 6: Runoff observed in the burnt forest during the 2 December, 2003, storm

Figures 7 and 8: Runoff flowing over a low bridge (Pont Gué) during the 2 December, 2003, storm (left)
and the same bridge (Pont Gué) at normal discharge (right)

Sediment Concentration (g L )

4 2


1 3 1 TOURRE
20 40 60 80
Area (%) in High & Very High Erosion

Figure 9: Relationship between sediment concentrations in storm runoff and proportion of area (%) defined as high
or very high erosion risk (numbers refer to site locations described in Fig. 5)

D-08-06.doc 121


Table 1: Coefficients ascribed to the four soil erosion factors used to estimate erosion risk: slope, pre-fire
vegetation density, fire severity, and soil erodibility
Slope (%) Pre-Fire Fire Severity Soil Erodibility
Vegetation Density
0-5 1 Medium 1 Variable 1 Shallow soil with high stone content 1
5-10 2 Low 2 Intense 3 Intermediate depth and stoniness 2
10-20 3 Bare 3 Deep soil with few stones 3
20-30 4
>30 5

Table 2: A brief description of the typical environments observed in the catchment for a range of erosion risk values
Erosion Area
Class Summary Description
Value km2 (%)
Valley bottoms with gentle slopes, low severity fire, dense pre-fire
Low ≤5 3.3 (16) vegetation and deep soils. Included are also some flat topslope sections
where soils are stonier and shallower.
Convexities and concavities upslope and downslope, respectively, of
major linear slope sections. These areas have gentler slopes than the
Intermediate 6 - 15 8.3 (40)
linear sections and are frequently found along the limits of the pre-fire
vegetation and fire serverity classes.
Steep slopes with bare or low pre-fire vegetation densities in the
severely burned area dominate. One significant exception is the north-
facing slope located south of the main Giscle channel. In this area, pre-
High 16 - 26 6.5 (31)
fire vegetation density and soil depth are greater than elsewhere in the
high erosion risk class, but the steep slopes make it vulnerable to
Severe fire conditions in low density pre-fire vegetation (with patches of
Very high ≥ 27 2.7 (13)
bare slope) on very steep slopes.

D-08-06.doc 122