Modeling and Analyzing Earthquake Induced Building Damages in a Gis

After a catastrophe like an earthquake a quick assessment of the damages has to be carried out in order to initiate repair actions. Especially building damages have to be recorded, in order to be able to give housing to the unsheltered as quickly as possible. To this end, the degree of damage of a building has to be captured, and based on these information decisions concerning the measures have to be taken (renovation, partial or total stub, etc.). Depending on the type of building, different factors influence the degree of damage and thus different actions are necessary. The damage indicators for a concrete building are for instance the number, size type and position of cracks, the position and status of damaged beams and supports, etc. A research project financed by the German Volkswagen Stiftung is dedicated to the investigation of a quick assessment and analysis of building damages. The acquisition is carried out with the help of close range photogrammetry, using digital cameras and photogrammetric analysis programs. The data is stored, managed and analyzed in a Geographic Information System (GIS). The paper concentrates on the GIS-part. Based on international coding schemes for earthquake damage qualification and a questionnaire developed at Istanbul Technical University, a detailed data model for a GIS has been established. The model includes both spatial and thematic components; furthermore it reflects the different building types with their respective peculiarities. An important property of the system is to make use of the analysis capabilities of the GIS to automatically derive values describing the degree of damage. The system was implemented in the GIS-product ArcView 3.0a by ESRI. The automation of the analysis processes was achieved using the object-oriented programming Language Avenue. In the paper first the detailed data model is presented. This model is then applied to several buildings of the city of Dinar (Turkey). The degree of damage of these buildings will be assessed automatically. Then the results will be compared to manually acquired results both in terms of quality and time efficiency. Furthermore, possible simplifications of the data model are discussed and their effects to the results are analyzed.