POST-EARTHQUAKE DAMAGE ASSESSMENT USING SATELLITE AND
AIRBORNE DATA IN THE CASE OF THE 1999 KOCAELI EARTHQUAKE, TURKEY
D. Ozisik?, N. Kerle"
? General Directorate of Bank of Provinces, Ataturk Bulvari, No:21 Opera/Ankara/TURKEY (dozisik(gilbank.gov.tr)
? Dept. of Earth Observation Sciences, International Institute of Geo-information Science and Earth Observation,
Hengelosestraat 99. P.O.Box 6. 7500 AA Enschede/The Netherlands (kerle@itc.nl)
Commission VII, TS-PS: WG VII/5
KEY WORDS: Earthquakes, satellite, aerial video, change detection, integration, texture, user
ABSTRACT:
To date, prevention of n
property. Especially following earthquakes, ther
event hours. Remote sensing technology can pr
temporal resolution and synoptic coverage. This work aims at integrating
age assessment in the case of the 1999 Kocaeli earthquake, Turkey. Pre and
ange detection methodology to detect damaged areas at the regional level.
limitations in detecting structural damage, airborne oblique video imagery was used to improve
gery of Golcuk was analyzed by visual interpretation and multi level
satellite and oblique airborne imagery to improve dam
post-earthquake Spot4 HRVIR were analyzed using ch
Since vertical satellite imagery has
the damage information at the local level. Aerial video ima
thresholding of texture and color indices feature layers. Additionally,
ation requirements were investigated by interviewing emergency organizations in
s of moderate-resolution optical satellite imagery for post-earthquake damage
technology in emergency activities, user inform
Turkey. The results indicate significant limitation
assessment. Despite substantial processing, registration and integration challenges of
assessment at the local level. On the other hand, even using video imagery c
the information requires integration of GIS and RS data.
1. INTRODUCTION
1.1 Motivation
Natural disasters are described as rapid and extreme events
within the geophysical system that result in mortality and
property damage, which exceeds the response and recovery
capabilities of the affected area (Kerle and Oppenheimer,
2002). Prevention of disasters is only rarely achieved with
today's technology and knowledge. However, it is possible to
avoid or to diminish the impact of disasters with effective
disaster management strategies. The main objective of disaster
management is to increase preparedness, provide early warning,
monitor the disaster in real time, assess the damage and
organize relief activities (Ayanz ef al., 1997). For effective
disaster management there is a need for a variety of geo-spatial
information at different scales. Geo-information science,
including geographic information systems (GIS), remote
sensing (RS), and the Global Positioning System (GPS), can
provide concrete support for disaster management activities in
terms of efficiency, and speed up the data management,
manipulation, analysis, output and lead to more informed
decisions (Montoya, 2002). Within the disaster management
cycle, arguably the most challenging phase is the response
stage, since the situation after the event is usually not clear,
little is known about what exactly happened, where it happened
and how many people were affected (Steinle et al., 2001).
Following a disaster, potential high-speed acquisition and
dissemination of air and space-borne data with synoptic
coverage allow the event to be detected and monitored. The use
of remote sensing technology provides a quantitative base for
atural disasters is only rarely achieved, and such events continue to pose an increasing threat to life and
e is a need for rapid, accurate and reliable damage information in the critical post-
ovide valuable information for response activities due to potentially high spatial-
(not in a sense of image fusion, but of synergy) vertical
to determine the effectiveness of the use of remote sensing
aerial video imagery, it can improve damage
annot fulfil all user information requirements, as most of
information about damage and aftermath monitoring to assist
response operations (Van Westen, 2002). Earthquakes are a
type of disaster with a high mortality rate and widespread
destruction. Unpredictability and rapid affects are the major
characteristics of earthquakes. Urban areas are most vulnerable
with their concentration of buildings, infrastructure and
population. In the time of emergency, due to interruptions of
communication systems and confusing information coming
from different sources, getting timely, accurate and detailed
information about the disaster situation is usually a challenging
task. For quick mobilization of response and relief
communities, data need to be captured and analysis result made
available rapidly in the first 72 hours after the earthquake, as
people, injured or exposed by the disaster, will not normally
survive more than 72 hours. Remotely sensed data for post-
earthquake damage assessment can provide valuable
information for emergency activities.
1.2 Research Objectives
Substantial research on post-earthquake damage assessment
using remotely sensed data has already been carried out (Turker
and San, 2003; Yusuf ef al., 2003; EDM, 2000). Although a
multi-temporal approach using change detection is the most
commonly used methodology, it has substantial limitations,
such as a short time gap between pre and post-event imagery,
changes in brightness values due to external factors. Visual
interpretation and texture analysis (Chiroiu and Andre, 2001)
using only post-disaster imagery provide more flexible
solutions compared with the previous one. Remotely sensed
686
Inter
data
sens:
and
char:
main
satel
high
the r
of ur
the c
Ever
panc
imag
due
suffe
scan
infor
DSM
proce
the t
spati:
acqu
obsta
Aeri
lowe
imag
asses
majo
aerial
1995
earth
alone
by ir
Septe
senso
used
infori
calcu
infori
This
borne
The
infort
imag
local
sensil
wides
aware
techn
frequ
amon
situat
usefu
the us
to de
levels
resear
1.3 (
Kocat
withir
world
Indus
