INTEGRATED ESTIMATION OF SEISMIC PHYSICAL VULNERABILITY OF TEHRAN USING RULE BASED GRANULAR COMPUTING
- 1GIS Dept., School of Surveying and Geospatial Eng., College of Eng., University of Tehran, Tehran, Iran
- 2Center of Excellence in Geomatic Eng. in Disaster Management, School of Surveying and Geospatial Eng., College of Eng., University of Tehran, Tehran, Iran
- 3Department of Earth Observation Science, University of Twente, the Netherlands
Keywords: Granular Computing (GrC) algorithm, Geospatial Information System (GIS), Earthquake Physical Vulnerability Assessment, Multi-Criteria Decision Making
Abstract. Tehran, the capital of Iran, is surrounded by the North Tehran fault, the Mosha fault and the Rey fault. This exposes the city to possibly huge earthquakes followed by dramatic human loss and physical damage, in particular as it contains a large number of non-standard constructions and aged buildings. Estimation of the likely consequences of an earthquake facilitates mitigation of these losses. Mitigation of the earthquake fatalities may be achieved by promoting awareness of earthquake vulnerability and implementation of seismic vulnerability reduction measures. In this research, granular computing using generality and absolute support for rule extraction is applied. It uses coverage and entropy for rule prioritization. These rules are combined to form a granule tree that shows the order and relation of the extracted rules. In this way the seismic physical vulnerability is assessed, integrating the effects of the three major known faults. Effective parameters considered in the physical seismic vulnerability assessment are slope, seismic intensity, height and age of the buildings. Experts were asked to predict seismic vulnerability for 100 randomly selected samples among more than 3000 statistical units in Tehran. The integrated experts’ point of views serve as input into granular computing. Non-redundant covering rules preserve the consistency in the model, which resulted in 84% accuracy in the seismic vulnerability assessment based on the validation of the predicted test data against expected vulnerability degree. The study concluded that granular computing is a useful method to assess the effects of earthquakes in an earthquake prone area.