04 
International Archives of the Photogrammetry, Remote Sensing 
1 
6. ACTUAL CONDITION OF BUILDING DAMAGE 
6.1 Classification of Building Damage 
The evaluation type of the three damage data sets is 
based on the evaluating criteria based on the central 
government. The criteria can be explained as follows. 
1) Small damage: There is no damage to the supporting 
system (foundation and pillar). For example, the surface of 
the partition wall has cracks. 
2) Medium damage: Supporting system is damaged. Beam 
is damaged and pillar has small cracks. Wall is partly 
damaged. Supporting power becomes weak but concrete is 
not crushed and steel rod is not cut down. 
3) Heavy damage and collapse: Supporting system is 
severely damaged. Pillar and concrete wall are heavily 
cracked and collapsed. Buildings are partly and completely 
collapsed and lean. In the second evaluation by the central 
government in Düzce city, the evaluation criteria "collapse" 
is used. 
6.2 Statistical Trend in Buildings Damage by First and 
Second Earthquakes 
Fig.9 is a map of Düzce city's total damage. The heavily 
damaged area is the central part of the city and there are 
relatively high RC buildings. The damage types classified 
by number of stories. The ratio of medium and heavily 
damaged buildings over 3 stories becomes abruptly 
significant. It is known that there are many high RC 
buildings in the central part of the city and the damage 
degree in this area thus becomes high. 
Building damage status on a more detailed spatial scale. 
This is the damage situation for each building in 
Uzunmustafar, which is one of the highly damaged areas in 
  
US 0%) 
Damage Ratio 
d 
[^ Small(<30%) 
  
  
  
  
  
  
  
| Heavy/Collapse wv 
  
  
and Spatial Information Sciences, Vol XXX V, Part B4. Istanbul 2004 
Düzce city. A star indicates the damage type, classified by 
the star types. It is painted over for the heavily damaged or 
collapsed buildings and hatched for the medium damaged 
buildings. Quite a few buildings are classified as heavily 
damages or medium damaged. In DiMSIS, the attributes 
are stored as these types of marks, and in general, the map 
objects with attributes are called connectors. Basically, we 
collect, sort, analyze and merge the connectors when we 
need to summarize and analyze the data in an arbitrary 
spatial scale. Base on the built GIS database, we can 
conduct a damage analysis for the first and second 
earthquakes. 
7. INTEGRATING MAP DATABASES 
QUICKBIRD IMAGE DATA 
USING 
Quickbird image data (QB data) showed in fig.10 was 
captured January 2003. Individual map database set of city 
area, new area developed by government and World Bank 
area by vector data are gathered. These three database set 
are measured independently with absolute coordinate. 
Matching has done by following procedure. 
I) Give factor to QB data to match to map database of city 
area. 
2) Add database of new area and check offset of image and 
vector data comparing certain object such as buildings or 
edge of roads. 
3) Move vector data by the offset to match to the QB data. 
This procedure can be justified because by the first step 
all the vector data and QB data has a good match all over 
the city. QB data of all area captured and corrected as 
uniform data can be treated as same accuracy. 
    
    
  
  
   
  
| Wicavy/Colbpse EM ddle OSmalVNe Dam age 
  
4006 
3500 Er = 
3000 fre 
2500 EE 
2000 
  
1500. 
Num ber of Bulls 
1000. F— 
sou foe 
  
  
Num bep ef Steri: 
Number of Buildings Classified by 
Damage Degrees and Number of Stories 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
   
  
  
  
  
  
  
  
  
  
  
  
  
  
  
    
  
N © y pe 
Fig. 9 Building Damage Status after the Second Earthquake