Vol. 17 No.

4 (398-404) ACTA SEISMOLOGICA SINICA July, 2004

Article ID: 1000-9116(2004)04-0398-07

The GIS and analysis of earthquake dam-

age distribution of the 1303 Hongtong M=8
earthquake*
GAO Meng-tan 1) ( ~ ~il ~ ) JIN Xue-shen 2) ( , ~ ~) AN Wei-ping 3) ( ~ : ~ - ~ - )
LLI Xiao-jian 4~ ( N ~ 1 [ )

1) Institute of Geophysics, China Earthquake Administration, Beijing 100081, China

2) Earthquake Administration of Hebei Province, Shijiazhuang 050021, China
3) Earthquake Administration of Shanxi Province, Taiyuan 030002, China
4) Center for Analysis and Prediction, China Earthquake Administration, Beijing 100036, China

Abstract

The geography information system of the 1303 Hongtong M=8 earthquake has been established. Using the spatial
analysis function of GIS, the spatial distribution characteristics of damage and isoseismal of the earthquake are
studied. By comparing with the standard earthquake intensity attenuation relationship, the abnormal damage dis-
tribution of the earthquake is found, so the relationship of the abnormal distribution with tectonics, site condition
and basin are analyzed. In this paper, the influence on the ground motion generated by earthquake source and the
underground structures near source also are studied. The influence on seismic zonation, anti-earthquake design,
earthquake prediction and earthquake emergency responding produced by the abnormal density distribution are
discussed.

Key words: the 1303 Hongtong M=8 earthquake; damage distribution; Geography Information System;
strong ground motion near source
CLC number: P315.9 Document code: A

Introduction
The 1303 Hongtong earthquake is an important earthquake in the eastern China. There is a lot
of information in historical documents about the earthquake and many traces destroyed by the
earthquake in the southern Shanxi Province. Many scholars have studied the earthquake from dif-
ferent aspects, but mainly limited the definitions of the isoseismal and three factors of the earth-
quake. Owing to being limited by the analysis technique, many useful damage information of the
earthquake are still not applied enough. Geography Information System (GIS) has been applied in
earth science study for many years and has generated very well effect. The historical seismic
information is characterized by apparently spatial distribution. The historical seismic damage has

* Received date: 2003-10-14; revised date: 2004-01-05; accepted date: 2004-03-12.

Foundation item: Nationalimportantfundamentalresearch"TheBasic Researchof ImportantProjectin DamageEnvironment"and
The importantproject"TheSeismicHazardAssessmentResearchand Anti-earthquakeStructureResearch"fromChinaEarthquake
Administrationduringthe 10thFive-yearPlan.
Contribution No. 04FE1008,Instituteof Geophysics,ChinaEarthquakeAdministration.
No.4 GAO Meng-tan,et al: GIS AND DAMAGEDISTRIBUTIONOF HONGTONGEARTHQUAKE 399

relation with many basic parameters, which have spatial distribution characters, such as ground
condition, source mechanism and basin. By introducing GIS into the historical seismic damage
analysis, the historical document and damage data in historical objects can be made full use and
will provide more useful information to the earthquake defense. Because the seismic damage can
reflect characters of the ground motion, so analyzing historical seismic damage data can search the
earthquake source character and the influences on strong ground motion generated by 3D under-
ground structures near the earthquake source (such as basin). We do a primary attempt by the
study on the 1303 Hongtong earthquake.

1 The GIS of the 1303 Hongtong earthquake

Using the Maplnfo, the GIS of the 1303 Hongtong earthquake has been built. The system has
important effect on studying and using the data of the 1303 Hongtong M=8 earthquake.
1.1 Basic data source
Basic data of the GIS include foundational geography information, fault, tectonic (such as
basin distribution), isoseismal and seismic damage.
Foundational geography information include the data of the administrative zonation of the
People's Republic of China, the data of administrative zonation of the Yuan Dynasty, village and
county (include the Ming Dynasty), fiver and lake. Historical maps are from The Atlas of Histori-
cal Map of China (Yuan and Ming Period) (TAN, 1982).
Data of fault and tectonic include fault data, seismic tectonic data, the Quaternary basin data
and distribution of earthquakes epicenter.
1.2 Main layers of GIS platform
Main layers include the foundational layer, the administrative zonation layer, the distribution
of seismic tectonics and faults layer, the basic earthquake parameter layer, the seismic damage
distribution layer, the actual isoseismic layer and theoretical isoseismic layer.
1.3 Main functions of GIS platform
The function of search and browse. The location of the village and county that were damaged
by the earthquake and its related information, such as the historical document, inscription on stone
tablet about the earthquakes can be displayed in the GIS platform. Using the search and browse
function, seismic intensity of every village and county, the theoretical intensity, the acceleration
value for anti-collapse design and its related geography information can be referred.
The damage spatial analysis function. Using the function, many studies can be done, such as
the statistical analysis of the seismic damage spatial distribution, the related analysis between
damage spatial distribution and geographic elements and the related analysis among seismic dam-
age, seismic tectonic and other data.

2 Damage spatial distribution and isoseismal

2.1 Seismic damage and its total distribution character
Data of seismic damage mainly come from The Compilation of Historical Materials of China
Earthquakes (XIE, CAI, 1983), The Compilation of Historical Earthquake Materials of Shanxi
Province (Earthquake Administration of Shanxi Province, 1991) and Historical Earthquake Cata-
logue of China (1831 BC~ AD 1969) (GU, 1983).
The 1303 Hongtong earthquake is the first M=8 earthquake sorted out from the China his-
torical data. The earthquake occurred in the region between Hongtong county and Zhaocheng
400 ACTA SEISMOLOGICA SINICA Vol. 17

county, Shanxi Province. In the earthquake, many villages and counties of the middle and southern
Shanxi Province and the border of Henan and Shanxi Province were destroyed. The villages and
counties seriously destroyed concentrated in the nearby county of Zhaocheng, which include
Huoxian, Lingshi, Jiexiu, Xiaoyi, Pingyao, Fenyang, Qixian, Xugou, Hongtong, Linfen, Fushan,
Xiangfen, Quwo. In the earthquake, 100 000 rooms and more than 1400 temples were destroyed,
the death and the wounded were heavy, more than 200 000 persons died (one saying 475 800 per-
sons died). The most seriously destroyed regions showed stick-bone-shape and located near the
seismogenic fault in Linfen basin.
The earthquake resulted in large-scale landslide, earthquake liquefaction and loess sinkage.
In the historical document, a village of Zhaocheng county is moved ten miles in the earth-
quake (Yuan Dynasty History, volume 50, Wuxing Journal) (XIE, CAI, 1983), which express the
earthquake was accompanied with the landslide.
In the historical document, there are many data about liquefaction of the earthquake. For
example, in some counties such as Xugou and Qixian of Taiyuan state, Pingyao, Jiexiu, Xihe and
Xiaoyi of Fenzhou state crack-channels were produced by the earthquake and black sand flowed
from spring in the earthquake (Yuan Dynasty History, volume 50, Wuxing Journal) (XIE, CAI,
1983).
More than 100 records about the earthquake in the historical document and the inscriptions
on stone tablets are included in the GIS. Using the GIS to analysis and study data is convenient.
2.2 Isoseismal
The isoseismal data come from Atlas of the Historical Earthquake in China (The Remote -
Yuan Dynasty Period), which are compiled by Institute of Geophysics of State Seismological Bu-
reau and Chinese Historical Geography, Fudan University (1983). By digital sampling the iso-
seismal data have been gained, then, the isoseismal data are revised with 1:250 000 electronics
geography maps. The result shows the digital isoseismal data are accurate for county and have
some inferences on village. Using the statistic function of GIS, the surface area of each density has
been gotten and is showed in Table 1.
2.3 Distribution of main village and county in each intensity area
Using the analysis function of GIS, the earthquake's influence intensity attributes are given
for every county and village in each intensity area as the foundation of further analysis. Table 2
gives the county and village numbers in each intensity area.

Table 1 T h e s u r f a c e area o f e a c h intensity Table 2 T h e n u m b e r s o f county and village in each intensity

area area
Intensity Surface area/km 2 Intensity County number Village number
X 3 402 X 7 47
IX 5 948 IX 7 56
VIII 26 249 VIII 20 216
VII 65 530 VII 51 520
VI 372 100 VI 286 3 351

It should be pointed out that the village data are that in the year 1991, which may have some
error with the data of Yuan Dynasty. But considering that the history of Shanxi Province is the
longest in China, the error should be small. Using the data to describe the damage spatial distribu-
tion should be enough.
No.4 G A O M e n g - t a n , et al: G I S A N D D A M A G E D I S T R I B U T I O N O F H O N G T O N G EARTHQUAKE 401

3 Main factors influencing the damage distribution characteristic

3.1 Theoretical isoseismai model
The attenuation relationship of earthquake intensity given by WANG Su-yun in Earthquake
Motion Parameter Zoning Map of China are adopted°:
/minor = 2.240 + 1.446M - 3.07 lg(R + 9) o"1 = 0.517 4 (1)
/major = 5.019+ 1.446M -4.1361g(R+ 24) 0-1 = 0.5174 (2)
The direction of ellipse major axis is taken as the direction of fault strike of the 1303 Hong-
tong M=8 earthquake, which is NE20 °. According to the attenuation relationship, the spatial loca-
tion data of each isoseismal are obtained, then the data are converted to the TAB file of the GIS.
The lengths of the major and minor axes of different intensity are achieved according to the
data of eastern China provided by the Compiling Group of China Seismic Intensity Zonation Map
(State Seismological Bureau, 1981). The isoseismal axes's length of the earthquakes magnitude
between 8 and 8 ~4 are showed in Table 3.

Table 3 The isoseismic axis's length of each intensity in China east (8- 8
Intensity X IX VIII VII VI
Half major axis / km 37 64 109 182 296
Half minor axis / km 14 28 57 118 253

The theoretical isoseismal model and the II0°E 112"

corresponding TAB file are established accord-

380N
ing to Table 3. Comparing two group data com-
ing from Table 3 and the attenuation relationship,
the result shows the intensity axes's lengths
coming from Table 3 are larger. In the view of
conservative assessment, the data coming from
Table 3 are thought to be more corresponding to
36°
actual historical giant earthquake, so the theo-
retical isoseismal and related TAB files are es-
tablished according to Table 3. According to the
TAB file, theoretical influence intensity value Ic
are given for every village and county. The dis-
tribution of actual isoseismal and theoretical
isoseismal are showed in Figure 1. --
Actual
isoseismal
_ _ Theoretical ~///~ Basin
isoseismal
3.2 Spatial distribution of intensity differ- Actual intensity VII Theoreticalintensity

ence
Figure 1 Distribution of actual isoseismal and
Comparing the actual influence intensity
theoretical isoseismal
value le with theoretical influence density value
Io the intensity difference value AJ of every village and county are achieved, then A/is given for
every town as a attribute and a single TAB file is established for conveniently comparing with
other factors.

WANG Su-yun, GAO A-jia, YU Yan-xiang. 1999. The Definition of Regional Ground Motion Attenuation Relationship of China. The
Sub-project Report of the ninth Five-year Plan Key Project (95-05-03-04-04).
402 ACTA SEISMOLOGICA SINICA Vol. 17

3.3 Comprehensive analysis of damage distribution

The study result of the near source strong earthquake observation and numerical simulation
show the distribution of strong ground motion is closely related to the propagation direction of the
earthquake source rupture. The south and north of the isoseismal of the 1303 Hongtong M = 8
earthquake are dissymmetry in the area damaged most seriously. The north half axis is far larger
than the south, which may be related with the rupture process• Comparing with the Northridge
earthquake of America on 17 January 1994, the Izmit earthquake of Turkey on 19 August 1999
and the Gujarat earthquake of India on 26 January 2001, the 1303 Hongtong M = 8 earthquake
maybe propagate NNE-ward and the distribution of damage may be influenced by the propagation
direction of the rupture.
In recent years, the data of seismic
Ill°E 112 ° 113 °
damage and strong ground motion record
in the world show basin has close relation

I
i"'•~''" i . i f i ~ ° ~
with the distribution of damage. In this
research field, a lot of studies have been
37"N
done by American scientists in Los Ange-
les and Japanese scientists in some basins
in Japan (Wald, Graves, 1998; Satoh, et al,
• "." "" .~ "..v~: "..." 2001). Author and Doctor YU Yan-xiang
simulated the influence in Beijing basin
caused by the Sanhe-Pinggu earthquake
36"
using the finite difference method and
found seismic waves had been obviously
-. magnified by the Beijing basin (GAO, YU,
2002). In the distribution map of intensity
• Town Town w i t h A l = l ~/////A Basin
a Town with ~ I>~2 - - Actual isoseismat difference of the 1303 Hongtong M=8
earthquake (Figure 2), the areas of inten-
Figure 2 The spatial distribution of intensity
sity X, IX and VIII in south and north are
difference
obviously larger about several tens kilo-
meters than the average areas of the eastem China. After putting the layer of intensity difference
on the basin layer, the high intensity difference regions are discovered all in the north of Linfen
basin and the south of Taiyuan basin. In the south of Taiyuan basin, the X region extends toward
north about 60 kin, the IX and VIII regions extend both toward north about 50 km, and the towns,
which intensity difference value equal to 2, all concentrate here. It is obvious that the basin magni-
fies the damage.
Some abnormal intensity towns located in Linfen basin and Changzhi, Shanxi Province. A lot
of particular studies about Linfen basin have been done ~. Linfen rift valley is the main tectonic
cell of Linfen basin. The amplitude of rift valley is large and along the axis direction the Linfen
rift valley is deep trough trending NNE. In Linfen rift valley the most thickness of the Cenozoic is
2.2 km and the Quaternary is 0.8 km. This kind of tectonics has amplificatory effect on ground
motion.

(DThe ScientificandTechnologyGroupof SeismicZonation,EarthquakeDisasterMitigationof Linfen,Governmentof ShanxiProvince

and StateSeismologicalBureau.1993.SeismicZonationand EarthquakeDisasterMitigationof LinfenDistrict,342~344.
No.4 GAO Meng-tan,et al: GIS AND DAMAGEDISTRIBUTIONOF HONGTONGEARTHQUAKE 403

The abnormal intensity areas near Changzhi, Shanxi Province locate in Lucheng basin, which
may have relation with basin responding. While considering earthquake damage of the southwest-
ern Shanxi Province, the basin effect is a very important factor.
The historical seismic damages show that the difference of earth's surface soil layer may pro-
duce the difference of seismic damages. Figure 2 expresses the high abnormal intensity areas
mainly distribute along the Fenhe river valley and other river valleys. The foundation failures,
such as sand liquefaction, aggravate the whole damages.

4 Some revelations to the earthquake disaster prevention from the

1303 Hongtong earthquake
Main principle of defense seismic damage is preserve in China. The earthquake-resistant de-
sign mainly accord to the National Seismic Zoning Map and the industry earthquake-resistant de-
sign criterion. The basic principle is "three level" defense, namely a small earthquake does not
result in damage, buildings can be fixed after a middle earthquake and buildings cannot collapse
during a great earthquake. The important measure of controlling the number of the wounded and
the died is that buildings cannot collapse during a great earthquake (The Ministry of Construction,
2001). In southern part of Shanxi Province, the basic defense acceleration values are 0.15g and
0.20g. According to the Building Code (GB50011-2001), the defense acceleration value for a great
earthquake is 2 times of the basic defense acceleration values. In many areas, the earth's surface
acceleration generated by the Hongtong earthquake exceeds the anti-collapse level. In addition,
the national anti-seismic design criteria are
109°E 111 ° 113 °
not carried out very well in rural region. If
the 1303 Hongtong earthquake reoccur, will __ ~Xinzhou
still have a large number of buildings col- 38°N
~ q
lapsed. /~("i" : l/S Tai'
4.1 Analysis of the building perform-
ances of anti-earthquake collapse
The anti-earthquake collapse accelera-
tion value (2 times of the value in the zona-
tion map) within isoseismal VIII of the 36 °
Hongtong earthquake are determined by w ~ .¢.-
using the results of the National Standard
GB 18306-2001, Earthquake Ground Motion
Parameters Zonation Map of China (HU, et
al, 2001), and assign it to all towns as an
attribute. We compared the a and le (accel-
eration related to le is calculated according -- Isoseismal ~ Basin ° P o i n t f o r a ( I ~ ) > ~ a
to the building code) for every town in this
region. The towns with a(Ie)>a are showed Figure 3 The distribution of towns that anti-
collapse capable lower than the
in Figure 3. 1303 Hongtongearthquake
If an earthquake with magnitude equal
to that of the 1303 Hongtong earthquake occurs, many towns will not satisfy the requirement of
anti-earthquake collapse in the region in Figure 3, even all the buildings meet the building code.
The number of such towns is 493, 10% percentage of all the towns.
404 ACTA S E I S M O L O G I C A SINICA Vol. 17

4.2 Some revelations to the earthquake disaster prevention

According to the above data and analysis, the isoseismal of the Hongtong earthquake is larger
than the isoseismals from the theoretical model. This fact is mainly from the spatial distribution of
the basin structure and soil. It may not be conservative to use the intensity attenuation relationship,
which is suitable to the eastern China in this region. So the prevention level may be underesti-
mated. If it is possible, the more detailed studies of seismology and seismic zonation should be
developed for Taiyuan basin and Linfen basin. The seismic zonation map of this region should
focus on the building performance of earthquake collapse, especially on seismic design parameters
related to the collapse.
The buildings in many towns in southern part of Shanxi Province do not meet the
anti-collapse requirement. Once the large event reoccurs, the local people may suffer a huge casu-
alty. The earthquake disaster prediction should be performed as soon as possible. On the basis of
this study, the comprehensive plan of earthquake disaster prevention should be prepared for the
southern part of Shanxi Province.
On the consideration of the return period of large earthquake in Shanxi rift valley (about three
hundreds years), it is necessary to make well preparation for the large earthquake response in this
region. At the same time, the joint prevention work of cities in this region is also needed.
The results introduced in this paper suggest that adopting GIS system to study the damage
distribution of the historical earthquake is benefit to the analysis on the characteristics of strong
ground motion in the near source region, the analysis on the influences of source and
three-dimensional structure on the ground motion. It is significant for the pertinence prevention.
So this method is worthy to be recommended.
References
State Seismological Bureau. 1981. China Seismic Zoning Research Report [M]. Beijing: Seismological Press, 79-83 (in Chinese).
Earthquake Administration of Shanxi Province. 1991. The Compilation of Historical Earthquake Materials of Shanxi Province [M].
Beijing: Seismological Press, 55-85 (in Chinese).
GAO Meng-tan, YU Yan-xiang. 2002. Three-dimensional finite-difference of ground motion in Beijing [J]. Earthquake Research in China,
18(4): 356-364 (in Chinese).
GU Gong-xu. 1983. Historical Earthquake Catalog of China (1831 BC~AD 1969) [M]. Beijing: Science Press, 16-18 (in Chinese).
HU Yu-xian, GAO Meng-tan, DU Wei, et al. 2001. GB18306-2001: Teaching Material of Earthquake Ground Motion Parameters Zona-
tion Map of China [M]. Beijing: China Standard Press, 22~60 (in Chinese).
Institute of Geophysics of State Seismological Bureau, Chinese Historical Geography, Fudan University. 1983. Atlas of the Historical
Earthquake in China (The Remote - Yuan Dynasty Period) [M]. Beijing: China Cartographic Publishing House, 152-156 (in Chi-
nese).
Satoh T, Kawase H, Sato T, et al. 2001. Three-dimensional finite-difference waveform modeling of strong ground motion observed in the
Sendal basin, Japan [J]. Bull Seism SocAmer, 91(4): 812~825.
TAN Qi-xiang. 1982. China Historical Map (Yuan-Ming Dynasty Period) [M]. Shanghai: Cartographic Publishing House, 9-10 (in Chi-
nese).
The Ministry of Construction. 2001. Building Code [M]. Beijing: China Construction Industry Press, 1~2 (in Chinese).
Wald D J, Graves R W. 1998. The seismic response of the Los Angeles basin, California [J]. Bull Seism Soc Amer, 88(2): 337-356.
XIE Yu-shou, CAI Mei-biao. 1983. The Compilation of Historical Materials of China Earthquakes [M]. Beijing: Science Press, 165-185
(in Chinese).