Вы находитесь на странице: 1из 34

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/305897804

Annual Report 2015 - Hypercity Project

Technical Report April 2016


DOI: 10.13140/RG.2.1.4099.8647

CITATIONS READS

0 273

10 authors, including:

Roeland Samson Frank Veroustraete


University of Antwerp University of Antwerp
203 PUBLICATIONS 2,317 CITATIONS 339 PUBLICATIONS 2,059 CITATIONS

SEE PROFILE SEE PROFILE

Luis Alonso Roland Valcke


University of Valencia Hasselt University
138 PUBLICATIONS 2,135 CITATIONS 111 PUBLICATIONS 1,079 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

DAISEX View project

Dry deposition and canopy exchange for temperate tree species under high nitrogen deposition View
project

All content following this page was uploaded by Frank Veroustraete on 05 August 2016.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
annual report April 2016 - HYPERCITY

HYPERCITY - April 2016


This report is an attachment to the ACTIVITY REPORT for the HYPERCITY project with
CONTRACT SR/00/303 and contains the project information (progress, results and
future work) and illustrative material.

Summary
1. INTRODUCTION........................................................................................................................................ 2
2. RESEARCH OBJECTIVES............................................................................................................................. 2
3. MATERIALS AND METHODS ..................................................................................................................... 4
3.1. SELECTION OF STUDY AREAS, TREE SPECIES AND SAMPLING POINTS ....................................................................... 4
3.2. FIELD CAMPAIGN ........................................................................................................................................ 4
3.2.1. 15HAS (2015 HYPERCITY Antwerp SIRM) campaign ..................................................................... 4
3.2.2. 15HAM (2015 HYPERCITY Antwerp Mapping) campaign ............................................................. 6
3.2.3. 15HAU (2015 HYPERCITY Antwerp Upscaling) campaign ........................................................... 13
3.2.4. Additional measurements ........................................................................................................... 17
4. FIRST RESULTS AND DISCUSSION ........................................................................................................... 18
4.1. 15HAS CAMPAIGN ................................................................................................................................... 18
4.2. 15HAM CAMPAIGN ................................................................................................................................. 21
4.3. 15HAU CAMPAIGN .................................................................................................................................. 30
4.3.1. Leaf level ..................................................................................................................................... 31
4.3.2. Canopy level ................................................................................................................................ 31
5. FUTURE STEPS ........................................................................................................................................ 32
5.1. VALENCIA 2016 ....................................................................................................................................... 32
5.1.1. 16HVS (2016 HYPERCITY Valencia SIRM) campaign ................................................................... 32
5.1.2. 16HVM (2016 HYPERCITY Valencia Mapping) campaign............................................................ 32
5.1.3. 16HVU (2016 HYPERCITY Valencia Upscaling) campaign ........................................................... 32
5.2. ANTWERP 2016....................................................................................................................................... 32
5.3. VALENCIA 2017 ....................................................................................................................................... 32
6. REFERENCES ........................................................................................................................................... 33

POLITIQUE SCIENTIFIQUE FEDERALE - FEDERAAL WETENSCHAPSBELEID


RESEARCH PROGRAMME FOR EARTH OBSERVATION STEREO III
CONTRACT SR/00/303

1
annual report April 2016 - HYPERCITY

1. Introduction
A significant proportion of Europes population lives in urban areas where exceedances of air quality
commonly occur and pose serious health risks. A major contribution to the urban air pollution is
provided by burning of fossil fuels in road transport, with the emission of particulate matter (PM).
PM pollution is causing tremendous costs to society due to a significant increase in health problems
and morbidity. To be able to make reliable risk assessments and take adequate urban management
decisions, it is therefore important to get a detailed insight in the spatial distribution of air pollution.

As conventional air pollution monitoring stations only provide coarse-scale spatial information on
exposure to pollutants, there is a growing interest in monitoring and modelling urban air pollution to
obtain information with a high spatial resolution. A possibility for high spatial resolution monitoring is
biomonitoring of urban vegetation.

When light reaches vegetation, it is absorbed, reflected or transmitted by the leaves. This process is
affected by the morphological and physiological characteristics of the leaves. Air pollution can induce
stress in tree leaves, altering the morphological and physiological characteristics of the leaves
(Balasooriya et al., 2009; Joshi & Swami, 2009; Doganlar & Atmaca, 2011). So, if these leaf
characteristics change due to air pollution, this will also change the reflectance signal of the leaf. This
makes the reflectance signal a good indicator for air pollution.
The reflectance signal can be measured using remote sensing techniques. Remote sensing is the
acquisition of information about an object, here plant leaves, without making physical contact with
the object. The advantage is that vegetation is not harmed and large surfaces can be monitored at
once.

Khavanin Zadeh et al. (2012) were able to estimate the habitat quality of urban trees based on RGB
reflectance measurements and the dorsi-ventral leaf asymmetry (the difference between the upper
and lower leaf side). When using hyperspectral data instead of RGB data, the potential of the dorsi-
ventral leaf asymmetry as air quality indicator can be higher. Until now the dorsi-ventral leaf
asymmetry and the resulting differences in hyperspectral leaf reflectance is hardly considered in any
(remote sensing) biomonitoring approach.

2. Research objectives
Various techniques exist for biomonitoring of urban vegetation. However, many techniques needs
further development. For example the use of hyperspectral reflectance and fluorescence. Also the
use of the dorsi-ventral leaf asymmetry (Khavanin Zadeh et al., 2012) and chlorophyll content
(Delegido et al., 2014) as indicator for air quality needs further development.

The overall objective of the HYPERCITY project is to develop, test and validate a plant-based passive
biomonitoring methodology based on hyperspectral observations and considering leaf asymmetry.
We make use of a dual approach, i.e.: (1) trees spatially distributed over the entire urban area for
mapping purposes, and (2) large solitary trees growing in various contrasting urban environments in
terms of air pollution used for scaling up exercises from leaf to canopy.

This overall objective is split up into the following objectives:

(i) Map the urban environmental air quality at high spatial resolution by magnetic leaf biomonitoring;

(ii) Map urban air quality based on physiological and reflectance based leaf characteristics and
validate these maps against the maps obtained in (i) and simulated air pollution maps;

2
annual report April 2016 - HYPERCITY

(iii) Scale up hyperspectral leaf reflectance to canopy level reflectance for trees growing in sites with
contrasting air quality, taking leaf asymmetry into account and based on a modelling approach;

(iv) Develop a hyperspectral Normalised Difference Asymmetry Index (NDAI) at leaf level for different
species growing at various levels of air quality as assessed in (i);

(v) Further develop and optimise the possibilities and applications of chlorophyll content mapping
based on advanced machine learning regression algorithms that deal efficiently with hyperspectral
data, and based on intensive physiological leaf measurements and additional validation data
obtained in (i) and (iv); and

(vi) Formulate a protocol to estimate urban air quality based on high spatial resolution leaf/canopy
data and/or airborne measurements.

The overall outcome of the project will be an optimisation of a biomonitoring methodology for the
monitoring of urban habitat quality (focussed on air quality) based on chlorophyll content,
chlorophyll fluorescence and hyperspectral leaf and canopy reflectance measurements.

3
annual report April 2016 - HYPERCITY

3. Materials and methods


In this section the materials and methods used in the 2015 campaign of HYPERCITY are explained.

3.1. Selection of study areas, tree species and sampling points


The first study area is the city of Antwerp (Belgium). During the summer of 2015 samples were taken
in Antwerp. The second city is the city of Valencia (Spain). Samples will be taken in Valencia during
the summer of 2016. Antwerp is located in a mild sea climate whereas Valencia is located in a
Mediterranean climate.

Because biomonitoring based on airborne or spaceborne remote sensing needs large vegetation
units, this project will deal with urban trees. The major selection criteria for urban tree species are
their abundance in the study area, contrast in leaf characteristics especially in terms of dorsi-ventral
leaf asymmetry, tree ecophysiological and architectural characteristics (fast or slow growing, size
crown) and a positive experience and bio-indicator potential gained in previous projects.

Plane tree Platanus x. acerifolia is selected as prior species of interest for Antwerp. Also in Valencia,
plane tree is selected, together with a second species, which will soon be selected. Platanus x.
acerifolia is selected because it is abundant in both cities (Valencia and Antwerp), and in most
European cities, and meets the major criteria as mentioned above. Previous studies showed the
potential of this tree as biomonitoring species (Hofman et al., 2013; Delegido et al., 2014). Both
crown and leaves are large enough for remote sensing monitoring. Large leaves are more suitable for
analyses then small leaves because some measuring tools need to be fixed on the leaf surfaces. In
the process of species selection, the spatial distribution over the study area is considered.

For mapping purposes: around 150 plane trees are selected (See 3.2.1 15HAS and 3.2.2 15HAM).
For up-scaling purposes: two trees in contrasting areas are selected (See 3.2.3 15HAU).

3.2. Field campaign


During 2015 there were three different field campaigns in Antwerp. The first (15HAS) was organised
to obtain information about the spatial variation in urban air quality (the background pollution). The
second (15HAM) was organised to collect leaf samples to map the urban air pollution based on
physiological and reflectance based leaf characteristics. The third field campaign (15HAU) was
organised to scale up hyperspectral leaf reflectance data to canopy level.

3.2.1. 15HAS (2015 HYPERCITY Antwerp SIRM) campaign


The spatial variation in urban air quality is mapped as necessary background data for the selected
sample points and also to be able to interpret the measured physiological and reflectance data in
function of urban air quality. Saturation Isothermal Remanent Magnetization (SIRM), a magnetic leaf
biomonitoring approach (see also SIRM method), is used to map the background pollution. SIRM
gives the magnetization remanence of an object after applying a saturating magnetic field. It is a
good indicator for traffic-induced particulate matter. This methodology proved already relevant in
temperate (Kardel et al., 2012) and Mediterranean (Van Wittenberghe et al., 2013; Van
Wittenberghe et al., 2014b) climates.

Based on a set of selection criteria and a digital GIS map of Antwerp with the locations of all urban
trees for 2014, a selection of 150 plane trees was made. The used selection criteria were: (1) species
Platanus x. acerifolia, (2) trees that are located maximum six km from the city center and maximum
three km from the ring road (with an exception for the industry location south of Antwerp), (3)
mature trees with a trunk diameter between 30 and 50 cm, (4) the base of the trunk is bare soil, with

4
annual report April 2016 - HYPERCITY

grass or with plants (not paved or without concrete). Also the land use classes were taken into
account. 92 trees were selected at roads, 15 trees were selected in an industrial zone, 40 trees were
selected in a park area and three trees were selected in the Zoo of Antwerp.

Eventually 143 plane trees were sampled between 1/06/2015 and 8/06/2015, representing
contrasting areas in the city (Figure 1). Not all trees were accessible in the field. Especially some park
trees were too high for sampling. We sampled 95 trees at roads (from which 8 were at a tram track),
10 trees in industrial area, 29 in park areas and 9 trees at a parking or playground area. From each
tree at least seven fully developed and undamaged leaves were collected with a telescopic pruner.
Sampling was performed at the road-facing side of the lower canopy, at a height of about three
meter above the ground.

Between 11/06/2015 and 19/06/2015 the collected leaves were analysed for leaf area (on fresh
leaves), leaf dry weight and SIRM. Based on the results of the 15HAS dataset, a smaller sampling set
was selected for the 15HAM campaign (Figure 1).

Figure 1: City map of Antwerp with the 143 sampling points for the 15HAS campaign in blue and the subset of 44 sampling
points for the 15HAM campaign in green. Note that under every green dot is also a blue dot.

SIRM method
Saturation isothermal remanent magnetization (SIRM) is the magnetization remanence of an object
after applying a saturating magnetic field. Before the SIRM analysis, the fresh leaf area and the leaf
dry weight were measured.

Each leaf was packed in cling film and pressed into a 10 cm3 plastic container. The cling film
immobilizes the leaf sample in the container. Then the sample was magnetized with a pulsed
magnetic field of 1 Tesla using a Molspin pulse magnetizer (Molspin Ltd, Newcastle, UK, Figure 2). For
each magnetized sample, the saturation isothermal remanent magnetization (SIRM) was measured
twice (without taking the sample out) using a Molspin magnetometer (Molspin limited Pulsemag

5
annual report April 2016 - HYPERCITY

DS4, Newcastle, UK, Figure 2). The magnetometer was calibrated before and after every 15
measurements by means of a magnetically stable sample. The SIRM (10-6 A) was normalized for the
volume of the plastic container (10-8 m3) and for the leaf area (m2) using formula 1. Each SIRM value
is the average of two measurements on the same sample.

( ) (3 )

Formula 1: () = (2 )

Figure 2: SIRM equipment. Molspin pulse magnetizer left, Molspin magnetometer right.

3.2.2. 15HAM (2015 HYPERCITY Antwerp Mapping) campaign


Based on the SIRM results 44 trees were selected with a good spatial distribution over the city and
the land use classes, and a good distribution over the SIRM classes. 44 trees were sampled for the
15HAM campaign (Figure 1) between 5/08/2015 and 26/08/2015 (three trees per day). We sampled
31 trees at roads (from which 3 were at a tram track), 5 trees in industrial area, 6 in park areas and 2
trees at a parking or playground area.

From each tree one or several branches, with at least 20 fully developed and undamaged leaves in
total, were collected with a telescopic pruner. Sampling was performed at the road-facing side of the
lower canopy, at a height of about 3m above the ground. Immediately after cutting, the branches
were put in buckets with water to avoid dehydration. Under water a piece of the branch was cut off
to avoid blocking of the xylem water transport by cavitation.

The pruned branches were soon after sampling transported to the laboratory, where they were
placed close to the window. The buckets were filled with extra water and the light in the laboratory
(TL lamps) were turned on. In the laboratory the following measurements were performed on 15
fully developed and undamaged leaves:
- Chlorophyll fluorescence measurements with Plant Efficiency Analyzer (PEA);
- Hyperspectral reflectance measurements and sun-induced chlorophyll fluorescence;
- Fluorescence imaging with Fluorescence Imaging System (FIS);
- Chlorophyll content (relative for all, absolute for some leaves) (See Leaf characteristics);
and
- morphological leaf characteristics (leaf area, fresh weight, dry weight) (See Leaf
characteristics).

Five of these 20 leaves were subsequently used for SIRM analysis. For some trees extra fully
developed and undamaged leaves were sampled from the collected branches for absolute
chlorophyll content measurements.

6
annual report April 2016 - HYPERCITY

Chlorophyll fluorescence measurements with PEA (Plant Efficiency Analyzer)


PEA (Plant Efficiency Analyzer, Hansatech Instruments Ltd., Norfolk, UK) measures the fast
fluorescence induction in a dark adapted leaf after light exposure. The LED used provides a red light
with a peak at wavelength 650 nm, which is readily absorbed by the chloroplasts. The amount of light
re-emitted at longer wavelengths is measured and the fluorescence yield of the leaf is quantified.
The parameters measured are Fm (maximum fluorescence level) and F0 (minimum fluorescence level).
The dark-adaptation is performed with special leaf clips for at least 30 minutes. In the absence of
light there is no photosynthesis, thus all reaction centres will be completely open, waiting for light to
initiate the process of photosynthesis. A light pulse will induce the Kautsky curve (Figure 3).

Figure 3: Kautsky Curve, O indicates the minimum fluorescence level, J and I indicate intermediate levels and P indicates
the maximum fluorescence level.

In order to obtain the maximum fluorescence signal (Fm), the sample is dark adapted and the light is
saturating the plant (additional light does not cause additional fluorescence). Saturating intensities
vary from species to species, with the period of dark adaptation and the environmental conditions in
which the plant is growing. The minimal dark adaptation time (20 minutes + 10 minutes margin) and
the optimal light intensity (40%, 1333 mol m-2 s-1) for plane tree are experimentally defined.
Measurements are always performed at both adaxial and abaxial sides of the leaf, at the same spot
close to the mid vain (Figure 4). The results are analysed with Biolyzer 2.0 for calculating parameters,
and R/SPPS and Excel/Sigmaplot for making statistical analysis and graphs, respectively.

Figure 4: Position of PEA clip for dark adaptation and measurements. Also the position of
the FluoWat leaf clip is marked (see later).

The ideal dark adaption period was tested at 40% light intensity. The maximum quantum yield of PSII
(Fv/Fm) was measured at different time-intervals. The ideal time period for a healthy leaf to get dark
adapted is where Fv/Fm is closest to 0.83 (83%). In case of stress, the acquired value will be lower than

7
annual report April 2016 - HYPERCITY

0.83. In our experiment, the measured values approached 0.83 after 20 minutes dark adaptation
already. A margin is added and 30 minute dark adaptation period is defined for plane leaves.

The optimal light intensity, measured after 30 minutes dark adaptation time, was determined by
measuring at various light intensities. Again the ideal light intensity corresponds to Fv/Fm closest to
83%. In this experiment an optimal value 0.83 is already reached at a very low light intensity. We
decided to apply a light intensity of 40% during the HAM and HAU campaigns, which is high enough
to provide Fm peak values to distinguish stressed from non-stressed samples (Fv/Fm) and is not too
high to produce errors (overscale due to excessive light exposure) considering the variation of the
measured samples as we planned to investigate the impact of an air pollution gradient on the
physiology of urban plane trees.

The data measured with PEA (F0, Fm, metadata) are collected in Biolyzer 2.0. This software program
calculates all parameters like quantum efficiency, flux ratios and performance index. All data are then
copied into Microsoft Excel for easy data manipulation. In advance of plotting, the data are sorted.
The first pulse measurements and the second pulse (after 10 seconds) measurements are separated
as well as adaxial and abaxial measurements. This means there are four measurements per leaf being
first pulse adaxial, first pulse abaxial, second pulse adaxial and second pulse abaxial. Subsequently,
the false measurements are removed. False measurements are identified by plotting the parameter
and detecting data falling well out of the data cluster. In this way 30 to 40 outliers are deleted, which
accounts for about 5% of all data.

After data manipulation, the analysis is performed. The design of graphs and statistical analysis is
performed with SigmaPlot 12.5 and/or IBM SPSS Statistics 23 respectively. The data includes
biophysical data like SLA (Specific Leaf Area), LWC (Leaf Water Content), RCC (Relative Chlorophyll
Content), SIRM (Saturation Isotherm Remanent Magnetization), and the PEA data like F0 (minimal
fluorescence), Fm (maximal fluorescence), PHI(p0) (quantum yield efficiency), DI0/RC (effective
dissipation in an active reaction centre), PI(abs) (performance index on absorption basis). This makes
it possible to investigate all kind of interactions between degree of pollution, biophysical parameters
and physiological parameters measured by PEA. Comparison of ad- and abaxial measurements will
indicate whether or not leaf asymmetry has a significant influence on the physiological parameters
measured by PEA.

Hyperspectral reflectance measurements (and sun-induced chlorophyll fluorescence)


Hyperspectral measurements are performed by means of the FluoWat leaf clip (Alonso et al., 2007;
Van Wittenberghe et al., 2013) coupled with the ASD AgriSpec (Analytical Spectral devices (ASD) Inc.,
Boulder, USA) (Figure 5). The FluoWat leaf clip is developed in Valencia and experimentally tested for
the first time by the research groups involved in HYPERCITY (Alonso et al., 2007; Van Wittenberghe et
al., 2013). The FluoWat leaf clip is able to measure real leaf reflectance (R), transmittance (T)
(without fluorescence contribution) and fluorescence (F) emission under artificial and natural light
conditions when coupled to a spectroradiometer.

8
annual report April 2016 - HYPERCITY

Figure 5: The FluoWat leaf clip (white circle) and the ASD Agrispec.

For the mapping campaign, 15 leaves per tree were analyzed with the FluoWat leaf clip coupled to
the ASD AgriSpec following the design in Table 1. Measurements were performed on the adaxial leaf
side (Figure 4). Per individual measurement, 15 files (replicas) are stored to get a more reliable mean
value. Both reflectance and transmittance was measured. Reflectance: fiber optic is attached to the
upside of the FluoWat leaf clip. Transmittance: fiber optic is attached to the downside of the FluoWat
leaf clip. The measuring cycle was ended by measuring the white reference once more.
Table 1: Measuring design for the mapping campaign 15HAM. Per tree 15 leaves are measured. R = reflectance
measurement; T = transmittance measurement; WR = white reference panel; L = leaf; ++ = with filter 650 nm; + = with filter
700 nm; - = without filter

File nr Measurement Explanation measurement code


0 14 R WR ++ Reflectance, white reference, filter 650 nm
15 29 R WR + Reflectance, white reference, filter 700 nm
30 44 R WR - Reflectance, white reference, no filter
45 59 RL- Reflectance, leaf, no filter
60 74 RL+ Reflectance, leaf, filter 700 nm
75 89 R L ++ Reflectance, leaf, filter 650 nm
90 104 T L ++ Transmittance, leaf, filter 650 nm
105 119 TL+ Transmittance, leaf, filter 700 nm
120 134 TL- Transmittance, leaf, no filter
R WR ++ Reflectance, white reference, filter 650 nm
R WR + Reflectance, white reference, filter 700 nm
R WR - Reflectance, white reference, no filter

The hyperspectral reflectance and transmittance measurements are performed in Raw DN. To
convert raw DN data to radiance, a cross calibration is used to radiometric calibrate the Agrispec.

For data treatment, following formulas are used (from Luis Alonso):

9
annual report April 2016 - HYPERCITY

Fluorescence Imaging with Fluorescence Imaging System FIS


The advantage of Fluorescence imaging over PEA lies in the fact that it produces an image instead of
a point (single pixel) measurement. As a consequence it provides an outcome indicating the
heterogeneous status of the leafs physiological status (health of leaf/chlorophyll abundance).

10
annual report April 2016 - HYPERCITY

Fluorescence imaging is performed with the FIS (Fluorescence Imaging System) developed by prof.
dr. Valcke at the University of Hasselt (Figure 6). Analogous to PEA, fluorescence images are taken
from both the adaxial and abaxial leaf side.

Figure 6: Fluorescence Imaging System (FIS) developed at UHasselt.

Roughly described, the set-up consists of a black and white CCD-camera (Charge-Coupled Device)
which is referred to as the imaging unit, LEDs (blue, red and green), a mirrored chamber for
homogenous light distribution and a clamp for insertion of samples (in this case leaves) which is
referred to as the excitation unit. The camera is connected to a computer and is controlled by
specialized software, this is the control unit. The software makes is possible to change the shutter
time and the gain. The camera takes 60 pictures in one second during each measurement. During the
first pass (first 30 pictures) the leaf is exposed to an ambient light intensity (300 mol), during the
second pass (last 30 pictures) the leaf is exposed to saturating light (with intensity around 1333
mol). After these 60 images, 3 more images are produced under red, blue and green illumination,
respectively. This way coloured images were obtained for analysis. By changing the shutter time or
gain the illumination of the image can be optimized (remark: an increasing gain negatively impacts
the sharpness of the image).

The images as acquired by the FIS, need to be digitally processed in order to allow a quantitative
interpretation. Image processing includes the following steps:

1. Background subtraction (correction for background light)


2. Intensity correction (correction for intensity loss towards edges of image)
3. Cropping and/or masking (intensity correction, removal of edges)
4. Geometric correction (a leaf is not flat, correction for the irregularities)

Image processing was performed using the (free) student version of KhorosPro2001 suite (Khoral
Research Inc., Albuquerque NM, USA).

Leaf characteristics
Leaf characteristics are measured in order to get detailed insight in the intra-canopy variation in leaf
anatomy and leaf dorsi-ventral asymmetry. These measurements are performed on 15 leaves per
location, to get reliable results.

The following characteristics were measured: leaf water content (LWC), specific leaf area (SLA),
chlorophyll content (CC) and nitrogen content (NC).

11
annual report April 2016 - HYPERCITY

Leaf water content affects the reflectance and fluorescence signal. It was calculated by subtracting
the leaf dry weight from the leaf fresh weight. The leaf dry weight was measured after drying the
leaves in a drying oven at 35 for at least 4 days.

LWC = fresh weight (g) dry weight (g) / mean leaf area (cm) (Formula 2)

Specific leaf area gives information about the leaf anatomy. The specific leaf area is the ratio of fresh
leaf area over dry leaf weight.

SLA = mean leaf area (cm) / dry weight (g) (Formula 3)

The relative chlorophyll content (RCC) was measured with the CCM-2000 (Opti-Sciences Inc. Hudson,
USA) on fresh leaves. Absolute CC was determined for 20 trees, on five leaves per tree taken at the
same location as for the 15HAM campaign. Measurements are performed to make a calibration
curve between the relative and absolute results and to convert the relative measurements to
absolute measurements (CC/fresh leaf weight). Therefore, the RCC was measured at five spots evenly
spread over the leaf. The same five spots were subsequently cut out with a 5 mm diameter puncher,
wrapped together in aluminium foil, immediately placed in liquid nitrogen and stored at -80C upon
analysis.

The procedure for the absolute CC measurements was as follows. The fresh weight of the punches
was determined and the sample (consisting of five punches per leaf) were put into a mortar together
with analytic sand (chemically neutral) to crush it. The sample was put into a 4 mL 80% acetone
solution and transferred to centrifuge tubes (Arnon, 1949; Lichtenthaler & Wellburn, 1983).
Centrifugation happened at 5000 rpm for 5 minutes. Afterwards the solution was transferred to a
clean centrifuge tube and the volume is measured for volumetric correction. The absorbance of the
samples was measured with a DW-2000 (Olis Inc., Georgia, USA). The absorbance was measured at
three wavelengths, namely at 663 nm for measuring chlorophyll a, at 646 nm for measuring
chlorophyll b and at 470 nm for measuring carotenoids. The measurements were calibrated with a
blank reference. The clarity of the extract can be verified at 750 nm. To calculate the concentrations,
following formulas were used based on Lichtenthaler and Wellburn (1983).

o Chlorophyll a = 12.21 * abs 663 (nm) 2.81 * abs 646 (nm) = chl a (g/ml)
o Chlorophyll b = 20.31 * abs 646 (nm) 5.03 * abs 663 (nm) = chl b (g/ml)
o Carotenoids = 1000 * abs 470 (nm) 5.27 * Ca 104 * Cb / 229 = tot. carotenoids
(g/ml)
Ca = chl a (g/ml)
Cb = chl b (g/ml)
o Surface area 5 mm diameter circle * 5 = 0.98125

For 20 trees of the 15HAM campaign, 5 leaves per tree were sampled for nitrogen content analysis.
For every set of five leaves, the total nitrogen content is measured using Kjeldahl method (Bchi 430
Kjeldahl Digester and Bchi 321 Kjeldahl Distiller, Bchi Labortechnik AG, Flawil, Switzerland).
Nitrogen represents protein content (and leaf structure) of a leaf and is often linked with chlorophyll
content. Like chlorophyll, nitrogen can be used to study the health status of leaves.

The collected samples were stored at -80C. To determine the nitrogen content, the samples were
dried in an 50C oven for three days. The temperature should not be higher than 100C to avoid

12
annual report April 2016 - HYPERCITY

destruction and loss of nitrogen due to evaporation. The dried samples were then first weighed and
crushed in mortars. We tested whether or not adding liquid nitrogen to easier crush the samples had
influence and if there was a different outcome between a coarse or fine crushed subsample. It turns
out that addition of liquid nitrogen has no influence but coarse crushed subsamples show much
more variation than fine crushed subsamples.

The organic matter was destructed in an 8 place Bchi-430 digester for heating. The samples were
put into tubes together with catalysers (K2SO4 and CuSO4) to increase the boiling point to approx. 373
C, 2 glass balls (to prevent sudden boil over) and H2SO4 to create an acid medium for destruction of
organic matter. The nitrogen complexed into ammonium sulphate (NH4SO4). The next step occured in
the distillation unit Bchi 315. Chemical reaction with catalyser (NaOH) set nitrogen free in the form
of ammonia (NH3). The ammonium was subsequently captured in a boric acid solution. Five drops of
indicator were added to the solution. The amount of nitrogen present was determined by back
titration with 0.1 M HCl. The nitrogen content was calculated by implementing the determined
parameters (x = dry weight, y = volume HCl) in Formula 4.
() 0.1 14 1000
Formula 4: () = /

3.2.3. 15HAU (2015 HYPERCITY Antwerp Upscaling) campaign


Based on the 15HAS campaign and additional sampling criteria (e.g. enough space for the boom lift),
two trees are selected for upscaling measurements. One tree was selected in an area with high air
pollution (tree nr 25; Thigh) and one in an area with urban background pollution level (tree nr 60, Tlow).
These trees are analysed on two levels, i.e. on leaf and canopy level (Figure 8).

Figure 7: Schematic overview of the leaf level and canopy level measurements of the upscaling campaign.

13
annual report April 2016 - HYPERCITY

Leaf level
For every tree, several branches with at least 15 fully grown leaves were sampled on six different
positions in the tree (Figure 7 and 8). Because of practical reasons, the samples were only taken at
two perpendicular wind directions (east and west orientation) and at 3 different heights (high H,
middle M and low L) (Table 2).

Table 2: Sample positions (code refers to Figure 8), height and date of sampling for tree 25 and tree 60 (numbers of tree as
in 15HAS campaign).

Sample position Tree 25 (Thigh) Tree 60 (Tlow)


1H 11.7 m 8/09/2015 12 m 15/09/2015
2M 7.6 m 8/09/2015 8 m 15/09/2015
3L 4.5 m 10/09/2015 4.0 m 14/09/2015
4H 12.0 m 9/09/2015 12 m 16/09/2015
5M 7.5 m 9/09/2015 8 m 16/09/2015
6L 4.5 m 14/09/2015 4.0 m 14/09/2015

High ( 12m)

Middle ( 8 m)

Low ( 4 m)

Figure 8: The six sampling positions in the tree crown for the 15HAU campaign (see also Table 2)

In laboratory conditions eight leaves per crown position are analysed for chlorophyll fluorescence
(with PEA), hyperspectral reflectance, imaging, chlorophyll content and other leaf characteristics (leaf
area, fresh weight, dry weight; see 3.2.2 15HAM and Leaf characteristics). The methods used are
the same as in 3.2.2 15HAM, except for the hyperspectral reflectance measurements. Hyperspectral
measurements were performed by means of the FluoWat leaf clip coupled with the ASD HandHeld 2
(Analytical Spectral devices (ASD) Inc., Boulder, USA) following the design in Table 3. Per individual
measurement, 15 files were stored to get a more reliable mean value. Only the leaves of Thigh (tree
25) position 2M had 10 files per measurement stored to save time on that day. Measurements were
performed on the adaxial and abaxial leaf side, at the same spot on the leaf (Figure 9). To know the
exact measuring spot on the leaf while changing the leaf clip, its position is marked with a small dot
on the leaf with a black pen, at the border of the leaf clip.
For five leaves of every canopy sample position the SIRM is analysed to know the crown
heterogeneity and to compare with previous SIRM results.

14
annual report April 2016 - HYPERCITY

Table 3: Measuring design for the upscaling campaign 15HAU at leaf level, eight leaves per tree are measured for
reflectance and transmittance both at adaxial (U for upside) and abaxial (D for downside) leaf side. R = reflectance
measurement; T = transmittance measurement; WR = white reference panel; L = leaf; U = adaxial leaf side up in leaf clip; D =
adaxial leaf side down in leaf clip; ++ = with filter 650 nm; + = with filter 700 nm; - = without filter

File nr Measurement Explanation measurement code


0 14 R WR ++ Reflectance, white reference, filter 650 nm
15 29 R WR + Reflectance, white reference, filter 700 nm
30 44 R WR - Reflectance, white reference, no filter
45 59 RLU- Reflectance, leaf, upside, no filter
60 74 RLU+ Reflectance, leaf, upside, filter 700 nm
75 89 R L U ++ Reflectance, leaf, upside, filter 650 nm
90 104 T L U ++ Transmittance, leaf, upside, filter 650 nm
105 119 TLU+ Transmittance, leaf, upside, filter 700 nm
120 134 TLU- Transmittance, leaf, upside, no filter
135 149 R WR ++ Reflectance, white reference, filter 650 nm
150 154 R WR + Reflectance, white reference, filter 700 nm
165 179 R WR - Reflectance, white reference, no filter
180 194 RLD- Reflectance, leaf, downside, no filter
195 209 RLD+ Reflectance, leaf, downside, filter 700 nm
210 224 R L D ++ Reflectance, leaf, downside, filter 650 nm
225 239 T L D ++ Transmittance, leaf, downside, filter 650 nm
240 254 TLD+ Transmittance, leaf, downside, filter 700 nm
255 269 TLD- Transmittance, leaf, downside, no filter
R WR ++ Reflectance, white reference, filter 650 nm
R WR + Reflectance, white reference, filter 700 nm
R WR - Reflectance, white reference, no filter

A B
Figure 9: Position of FluoWat leaf clip for hyperspectral measurements on the adaxial side or upside (A) and abaxial
side or downside (B).

Canopy level
Hyperspectral measurements are performed by means of the ASD HandHeld 2 (Analytical Spectral
devices (ASD) Inc., Boulder, USA) without the attachment of a fiber optic and following the design in
Figure 10 and Table 4. For every measurement, 20 files are stored to get a more reliable mean value.

15
annual report April 2016 - HYPERCITY

Measurements are performed on a day with clear blue sky (10/09/2015), on tree 25 between 10:50
and 14:20. At the start of the measurements, the light intensity of PAR (in mol m-2 s-1) is measured
with a quantum sensor (photometer model Li-189, Li-Cor, Nebraska, USA) while the sensor was
horizontally placed, and facing up.

Figure 10: Measuring design for 15HAU campaign on canopy level, for every position in the tree crown. Numbers represent
the order of measurements taken. A: vertical design with locations of the HandHeld with respect to the crown: spot 2 is at
180 cm from the tree crown; B: horizontal design with measuring positions of the HandHeld at three different distances from
the crown (the circle 2 = same position as 2 in A). C: Schematic view of HandHeld positions 1, 2, 3, 6 and 7 during
measurements. Positions 4 and 5 are respectively in front and behind position 2.
= measuring position in tree crown = location of HandHeld relative to the crown during canopy measurements

Table 4: Measuring design for 15HAU campaign at canopy level. Six locations (Figure 8) in the tree crown are measured. The
numbers in the table are conform the numbers in Figure 10. WR = white reference panel; h = horizontal position relative to
the ground; v = vertical position relative to the ground.

File nr Measurement Comment


0 19 WR h Measurement on a horizontal placed white reference panel
20 39 WR h Horizontal placed white reference panel
40 59 WR v Measurement on a vertical placed white reference panel
60 79 1 Measuring position of Handheld conform Figure 10
80 99 WR v Vertical placed white reference panel
100 119 2 Measuring position of Handheld conform Figure 10
120 139 WR v Vertical placed white reference panel
140 159 3 Measuring position of Handheld conform Figure 10
160 179 WR v Vertical placed white reference panel
180 199 4 Measuring position of Handheld conform Figure 10
200 219 WR v Vertical placed white reference panel
220 239 5 Measuring position of Handheld conform Figure 10
240 259 WR v Vertical placed white reference panel
260 279 6 Measuring position of Handheld conform Figure 10
280 299 WR v Vertical placed white reference panel
300 319 7 Measuring position of Handheld conform Figure 10
320 339 WR v Vertical placed white reference panel
340 359 WR h Horizontal placed white reference panel
360 379 WR h Horizontal placed white reference panel

16
annual report April 2016 - HYPERCITY

The vertical measurements were taken to know the difference in light intensity that reaches the
measurement locations in the crown. The horizontal references are taken to calculate the reflectance
and to know the differences in light intensity over time. Vertically placed white reference panel is
sensitive for shade effects when there is a change in the angle to the sun. A more fixed setup is
helpful to measure the white reference perpendicular (in horizontal position) with a fixed distance
between the HandHeld and the white reference. During hand held measurements the distance was
approximately the same, however it was not a permanent setup.

Leaf characteristics
Also during the 15HAU campaign, following characteristics were measured on 15 leaves per position
in the crown for both trees: leaf water content (LWC), specific leaf area (SLA), chlorophyll content
(CC) and nitrogen content (NC). Nitrogen content analysis was performed on 15 leaves but only for
four positions per crown (positions 1 and 4, high and positions 3 and 6, low in the crown).

Leaf cross sections


To study structural changes of palisade- and sponge parenchyma induced by traffic air pollution and
inter-canopy variation, leaf cross-section coupes are made. Therefore 15 leaves per crown positions,
from both trees of the upscaling campaign Thigh and Tlow, are sampled.

One punch per leaf is made, in the middle of the leaf with an overlap on the mid vain. The punches
are made with a square puncher (side length 8 mm). The punches are immediately put into a FAA-
solution (Formalin Acetic acid Alcohol). The FAA-mixture consists of 90 mL 70% ethanol, 5 mL
concentrated acetic acid and 5 mL 40% formalin. Each punch is stored in a small glass jar. Fixation has
to last at least 4 hours in a refrigerator at about 4C. The day after, the punches are washed with
subsequently 25%, 50% and 70% ethanol. The samples in the 70% ethanol solution can be stored for
months at 4C in the refrigerator.

The leaf tissue is imbedded in paraffin to be able to make coupes. The tissue is coloured as well for
better visibility. Therefore, all the water needs to be gone. Dehydration and imbedding in paraffin
happens automatically in the Autotechnicon Mono Tissue Processor. The tissues are then placed in
more paraffin which solidifies in a thick paraffin block. The paraffin block can easily be sliced into leaf
coupes (8 m) by a microtome. Next, the coupes are coloured with safranin and astra-blue to make
structures visible under the light microscope.

The making of leaf cross-section coupes is planned for spring 2016.

3.2.4. Additional measurements


Soil quality and magnetic susceptibility.
Soil samples are taken at all the locations of the 143 trees. These samples will be analysed as
reference for each urban class. Pollution monitoring with spectroscopy. (In progress)

Weather forecast and effect (temperature, precipitation, )


For every sampling day the weather forecast is collected prior to the measurements. Also the
weather data from VMM (Flemish Environment Agency) is consulted.

17
annual report April 2016 - HYPERCITY

4. First results and discussion


In this section, the first results obtained by the 2015 campaign of HYPERCITY are presented. The aim
of these preliminary results is to show what kind of analysis is possible. First the results of the 15HAS
campaign are presented, followed by those of the 15HAM and 15HAU campaigns.

4.1. 15HAS campaign


The 15HAS campaign was organised to obtain information about the spatial variation in urban air
quality (the background pollution). Figure 11 gives the overview of the sample locations with more
detail for the big park locations. Figure 12 shows the tree numbers on the map.

Rivierenhof

Stadspark and Zoo

Nachtegalenpark

Figure 11: Sample locations in Antwerp for the 15HAS campaign (big map) and more detailed maps for the three biggest
park areas and the Antwerp Zoo.

18
annual report April 2016 - HYPERCITY

Figure 12: Map of tree numbers of the 15HAS campaign.

Figure 13 maps the spatial variation in SIRM values measured during the 15HAS campaign. The
lowest SIRM value (average STDEV on tree level) is 10.9 6.5 A and is located in a park area
(Rivierenhof in the east of Antwerp), the highest value is 310.0 100.9 A and is located near a metal
processing company southwest of Antwerp. These maximum values are higher than the maximum
values reported in other studies for P. x acerifolia: 51 15 A in Valencia (Van Wittenberghe et al.,
2013), and 64.1 A in a street canyon in Ghent (Hofman et al., 2013), but are not exceptional for this
location in Antwerp (values measured up to 463.49 A on strawberry leaves during the AIRbezen
project 2014).

Figure 14 shows the SIRM data in a boxplot for every sampled land use class. The mean SIRM for the
different land use classes seems to follow a gradient from park (32.0 15.8 A; n = 29) over road
(44.7 29.4 A; n = 95) to industry (96.8 83.2 A; n = 10). The 7 trees near a tram track have a
mean of 84.7 27.3 A, the 9 trees on a playground or parking have a mean of 28.2 7.0 A.

19
annual report April 2016 - HYPERCITY

Figure 13: SIRM values (in A) per location for the 15HAS campaign. In total 143 trees were sampled and measured.

Figure 14: Boxplot of the SIRM values measured during 15HAS for every land use class. Sampling sizes: 29 trees in park areas
and 9 trees at a parking or playground area, 87 trees at roads, 8 at a tram track, 10 trees in industrial area.

20
annual report April 2016 - HYPERCITY

4.2. 15HAM campaign


The 15HAM campaign was organised to collect leaf samples to map the urban air pollution based on
physiological and reflectance based leaf characteristics. Figure 15 maps the spatial variation in SIRM
values measured during the 15HAM campaign. Figure 16 shows the range of the SIRM values over
the different land use classes.

SIRM data

Figure 15: SIRM values per location for the 15HAM campaign. Numbers on the map are the tree code as in the 15HAS
campaign (see Figure 13). In total 44 trees are sampled and measured.

21
annual report April 2016 - HYPERCITY

Figure 16: Boxplot of the SIRM values measured during 15HAM for every land use class. Sample sizes: 6 trees in park areas
and 2 trees at a parking or playground areas, 28 trees at roads, 3 at a tram track, 5 trees in industrial area.

We compared the SIRM values measured on the 15HAS campaign and on the 15HAM campaign
(Figure 17). In most of the cases the mean SIRM calculated on the second campaign (15HAM) is
higher than the SIRM calculated on the first campaign (15HAS), which can be explained by the longer
exposure time to the pollution. In 9 out of the 44 cases the second mean SIRM was lower than the
first mean SIRM calculated. The difference in mean SIRM values between the two campaigns is
statistically significant (paired samples t-test p = 0.012). Figure 18 shows that the selection of trees
for the 15HAM campaign covered the whole SIRM range measured during the 15HAS campaign.

Figure 17: Comparison SIRM measured on 15HAS campaign and SIRM measured on 15HAM campaign.

22
annual report April 2016 - HYPERCITY

Figure 18: Comparison coverage SIRM values for the 15HAS and 15HAM campaign.

Chlorophyll fluorescence measurements with PEA


The quantum yield efficiency PHI(P0) gives an indication of the probability that an absorbed photon is
taken up by the reaction centre. The lower the PHI(P0), the more photons are send out again as
fluorescence. The PHI(P0) ranges from 0.818 to 0.844. Most trees have a value above 0.83 which
means that there is no physiological stress. The quantum yield efficiency is spatially mapped in Figure
19. The range of PHI(P0) values over the land use classes is shown in Figure 20. There seems to be no
relation between the quantum yield efficiency measured by PEA and the SIRM values (Figure 21).

Figure 19: The factor PHI(P0) spatially mapped for the 44 trees of the 15HAM campaign. Data showed is the mean of 15
leaves per tree, of pulse 1 from the upside measurements.

23
annual report April 2016 - HYPERCITY

Figure 20: Boxplot of the quantum yield efficiency values measured during 15HAM for every land use class. Sample sizes: 8
trees in park areas (of which 2 trees at a parking or playground areas), 28 trees at roads, 3 at a tram track, 5 trees in
industrial area.

250

200
SIRM (A)

150

100

R = 0.0155
50

0
0.815 0.82 0.825 0.83 0.835 0.84 0.845 0.85
PHI(Po)

Figure 21: Linear regression between the quantum yield efficiency PHI(P0) and the average SIRM per tree.

Leaf characteristics
The relative chlorophyll content ranges between 14.8 and 34.9. The specific leaf area ranges
between 88.4 cm2g-1 and 176.3 cm2g-1. These values are in line with the values found by
Khavaninzadeh et al. (2015) (mean SLA of 97.2 cm2g-1 at the bottom of the canopy for Platanus x
acerifolia) and Van Wittenberghe et al. (2014a) (SLA ranged from 35.1 to 160.0 cm2 g-1 for a dataset
with P x acerifolia, Celtis australis, Morus alba and Phoenix canariensis). The leaf water content
ranges between 0.0107 g/cm2 and 0.0154 g/cm2. These values lie in the range from 0.005 to 0.034 g

24
annual report April 2016 - HYPERCITY

cm-2 measured by Van Wittenberghe et al. (2014a) for a dataset with P x acerifolia, Celtis australis,
Morus alba and Phoenix canariensis. The leaf characteristics relative chlorophyll content, specific leaf
area and leaf water content are spatially mapped in respectively Figure 22 (RCC), 23 (SLA) and 24
(LWC).

The range of the values for relative chlorophyll content, specific leaf area and leaf water content are
shown in a boxplot in respectively Figure 25 (RCC), 26 (SLA) and 27 (LWC).

Figure 22: The mean relative chlorophyll content (RCC) per tree spatially mapped for the 15HAM campaign.

25
annual report April 2016 - HYPERCITY

Figure 23: The mean specific leaf area (cm2/g) SLA per tree spatially mapped for the 15HAM campaign.

Figure 24: The leaf water content (g/cm2) LWC per tree spatially mapped for the 15HAM campaign.

26
annual report April 2016 - HYPERCITY

Figure 25: Boxplot of the relative chlorophyll content values measured during 15HAM for every land use class. Sample sizes:
8 trees in park areas (of which 2 trees at a parking or playground areas), 28 trees at roads, 3 at a tram track, 5 trees in
industrial area.

Figure 26: Boxplot of the specific leaf area values (cm2 g-1) measured during 15HAM for every land use class. Sample sizes: 8
trees in park areas (of which 2 trees at a parking or playground areas), 28 trees at roads, 3 at a tram track, 5 trees in
industrial area.

27
annual report April 2016 - HYPERCITY

Figure 27: Boxplot of the leaf water content values (g cm-2) measured during 15HAM for every land use class. Sample sizes: 8
trees in park areas (of which 2 trees at a parking or playground areas), 28 trees at roads, 3 at a tram track, 5 trees in
industrial area.

28
annual report April 2016 - HYPERCITY

There seems to be no relation between the SIRM and the measured leaf characteristics (Figure 28).

Figure 28: Linear regression between leaf characteristics (relative chlorophyll content, specific leaf area and leaf water
content) and SIRM values.

29
annual report April 2016 - HYPERCITY

4.3. 15HAU campaign


The 15HAU campaign was organised to scale up hyperspectral leaf reflectance data to canopy level.
Results from both leaf and canopy level are presented here.

SIRM analysis

Figure 29: Colour code for SIRM values in the 15HAU campaign for tree 25 (high pollution area) and tree 60 (urban
background pollution area). Numbers give the position code in the tree (as in table X). The colour code is comparable with
the maps showed above (see Figure 14 and 15).

car on ring road

Tree 25 Tree 60
(on roundabout)

When the SIRM values between 15HAS, 15HAM and 15HAU are compared, the last measurements
show little difference between tree25 and tree60 (Table 5). The supposed differences in air quality
between the two areas Thigh and Tlow assumed from the first SIRM campaign 15HAS, do not match the
measurements.
Table 5: Comparison average tree SIRM values over the three different campaigns.

SIRM (A) 15HAS 15HAM 15HAU


Thigh (Tree 25) 45.05 23.10 55.02
Tlow (Tree 60) 27.52 38.77 53.32

30
annual report April 2016 - HYPERCITY

4.3.1. Leaf level


Hyperspectral leaf reflectance measurements and sun-induced chlorophyll fluorescence

Figure 26 is an example of possible reflectance and transmittance results retrieved from the spectral
data. Further analysis is in progress.

T25_3L_all leaves
0.60 0.00

0.50 0.10

Transmittance
0.40 0.20
Reflectance

0.30 0.30

0.20 0.40

0.10 0.50

0.00 0.60
300 400 500 600 700 800 900 1000
Wavelength (nm)

aR_up aR_down aT_up aT_down

Figure 30: Average reflectance and 1-transmittance in the visible, far-red and part of the near-infrared for tree 25 at location
3L (east low side). aR_up = average apparent reflectance at adaxial leaf side, aR_down = average apparent reflectance at
abaxial leaf side, aT_up = average apparent transmittance at adaxial leaf side, aT_down = average apparent reflectance at
abaxial leaf side.

First results obtained with the ARTMO toolbox on the relation between the spectra and the leaf
characteristics, will be presented by Jochem Verrelst.

4.3.2. Canopy level


Hyperspectral canopy measurements

In progress

In table 6 the light intensity of PAR (in mol m-2 s-1) measured at the start of the canopy level
measurements is shown. The quantum sensor (photometer model Li-189, Li-Cor, Nebraska, USA) was
horizontally placed, and facing up.
Table 6: Timing of canopy measurements, light intensity and height above ground of measuring location in crown tree 25.

1H 2M 3L 4H 5M 6L
Time start measure 11:00 12:20 12:00 13:15 13:40 14:00
Light intensity 1768 2502 2381 2581 2623 2578
mol mol mol mol mol mol
Height in crown 12m 7.5m 4m 12m 7.5m 4.5m

31
annual report April 2016 - HYPERCITY

5. Future steps
5.1. Valencia 2016
The campaign in Valencia would be organised between the 15th of June and 15th of August 2016.
Before this time period, we assume the leaves are not exposed long enough to significantly detect
differences in morphological and physiological leaf characteristics. After the time period, soil drought
and air temperature stress might be overruling air pollution stress. Plane tree would be under soil
drought stress which could affect the measurements. The accents of the campaign will be on the
spatial distribution (mapping). This was not the focus during the BIOHYPE project.

In Valencia plane tree, palm tree and orange tree are possible test species.

5.1.1. 16HVS (2016 HYPERCITY Valencia SIRM) campaign


During two weeks in May 2016 we will organise a preparatory campaign. During this campaign we
aim to select around 100 sampling points and map the spatial variation in urban air quality as
necessary background data using SIRM. 15 leaves per tree will be sampled and send to the University
of Antwerp for analysis. By June 10th we plan to have the SIRM results. Based on these results we
choose a subset of trees for the mapping campaign.

5.1.2. 16HVM (2016 HYPERCITY Valencia Mapping) campaign


Between 15th of June and 15th of August 2016, a three week period will be scheduled for the
mapping campaign. During mornings, branches with at least 20 leaves will be sampled. During the
afternoons the measurements will take place. We aim to measure 3 to 4 trees a day. This way we can
measure a total number between 45 and 60 trees in three weeks.

In contrast to the measurements in Antwerp, the measurements in Valencia will be performed


outside under the natural sun light conditions. The focus will be on dorsiventral sun-induced
fluorescence and reflectance measurements. Also, there will be a special focus on chlorophyll
content. The aim is to further develop and optimise the possibilities and applications of chlorophyll
content mapping.

5.1.3. 16HVU (2016 HYPERCITY Valencia Upscaling) campaign


If the time schedule, weather and practical arrangements (boom lift) are good, we plan an upscaling
campaign for two to three weeks between 15th of June and 15th of August 2016. The canopy of two
trees in areas with contrasting air quality will be measured with the ASD HandHeld comparable to
what was done in Antwerp .

5.2. Antwerp 2016


If possible we would schedule a Flex EUFAR flight campaign in September 2016 in Antwerp.

Two sensors: a fluorescence (LIF) and multispectral sensor

Time window: 14 days

5.3. Valencia 2017


If possible we would schedule a Flex flight campaign in the summer of 2017 in Valencia.

32
annual report April 2016 - HYPERCITY

6. References

Alonso, L., Gomez-Chova, L., Vila-Frances, J., Amoros-Lopez, J., Guanter, L., Calpe, J. & Moreno, J.
(2007) Sensitivity analysis of the Fraunhofer Line Discrimination method for the
measurement of chlorophyll fluorescence using a field spectroradiometer. Igarss: 2007 Ieee
International Geoscience and Remote Sensing Symposium, Vols 1-12, 3756-3759.
Arnon, D.I. (1949) Copper Enzymes in Isolated Chloroplasts. Polyphenoloxidase in Beta Vulgaris. Plant
Physiol, 24, 1-15.
Balasooriya, B.L.W.K., Samson, R., Mbikwa, F., Vitharana, U.W.A., Boeckx, P. & Van Meirvenne, M.
(2009) Biomonitoring of urban habitat quality by anatomical and chemical leaf
characteristics. Environmental and Experimental Botany, 65, 386-394.
Delegido, J., Van Wittenberghe, S., Verrelst, J., Ortiz, V., Veroustraete, F., Valcke, R., Samson, R.,
Rivera, J.P., Tenjo, C. & Moreno, J. (2014) Chlorophyll content mapping of urban vegetation
in the city of Valencia based on the hyperspectral NAOC index. Ecological Indicators, 40, 34-
42.
Doganlar, Z.B. & Atmaca, M. (2011) Influence of Airborne Pollution on Cd, Zn, Pb, Cu, and Al
Accumulation and Physiological Parameters of Plant Leaves in Antakya (Turkey). Water Air
and Soil Pollution, 214
Hofman, J., Stokkaer, I., Snauwaert, L. & Samson, R. (2013) Spatial distribution assessment of
particulate matter in an urban street canyon using biomagnetic leaf monitoring of tree crown
deposited particles. Environmental Pollution, 183, 123-132.
Joshi, P.C. & Swami, A. (2009) Air pollution induced changes in the photosynthetic pigments of
selected plant species. Journal of Environmental Biology, 30
Kardel, F., Wuyts, K., Maher, B.A. & Samson, R. (2012) Intra-urban spatial variation of magnetic
particles: Monitoring via leaf saturation isothermal remanent magnetisation (SIRM).
Atmospheric Environment, 55, 111-120.
Khavanin Zadeh, A.R., Veroustraete, F., Wuyts, K., Kardel, F. & Samson, R. (2012) Dorsi-ventral leaf
reflectance properties of Carpinus betulus L.: An indicator of urban habitat quality.
Environmental Pollution, 162, 332-337.
Khavaninzadeh, A.R., Veroustraete, F., Van Wittenberghe, S., Verrelst, J. & Samson, R. (2015) Leaf
reflectance variation along a vertical crown gradient of two deciduous tree species in a
Belgian industrial habitat. Environmental Pollution, 204, 324-332.
Lichtenthaler, H.K. & Wellburn, A.R. (1983) Determinations of total carotenoids and chlorophylls a
and b of leaf extracts in different solvents. Biochemical Society Transactions, 11, 591 - 592.
Van Wittenberghe, S., Verrelst, J., Rivera, J.P., Alonso, L., Moreno, J. & Samson, R. (2014a) Gaussian
processes retrieval of leaf parameters from a multi-species reflectance, absorbance and
fluorescence dataset. Journal of Photochemistry and Photobiology B-Biology, 134, 37-48.
Van Wittenberghe, S., Alonso, L., Verrelst, J., Hermans, I., Valcke, R., Veroustraete, F., Moreno, J. &
Samson, R. (2014b) A field study on solar-induced chlorophyll fluorescence and pigment
parameters along a vertical canopy gradient of four tree species in an urban environment.
Science of the Total Environment, 466, 185-194.
Van Wittenberghe, S., Alonso, L., Verrelst, J., Hermans, I., Delegido, J., Veroustraete, F., Valcke, R.,
Moreno, J. & Samson, R. (2013) Upward and downward solar-induced chlorophyll
fluorescence yield indices of four tree species as indicators of traffic pollution in Valencia.
Environmental Pollution, 173, 29-37.

33

View publication stats