Analytical Evaluation of Uncertainty Propagation in Seismic Vulnerability Assessment of Tehran Using GIS

Disaster management processes are generally characterizing real world as composition of various criteria to overcome damage consequences. This situation usually entails development of hypotheses based on epistemic and multi-criteria treatments. These hypotheses have to be properly structured using logics which somehow handle uncertainties like fuzzy logic. One of the properties of geospatial information systems is their use in supporting spatial decision making under uncertainty. It is a complex process which is considered in different situations. The existence of uncertainty in geospatial data and various analyses has the potential to expose users to undesirable consequences in their decision making. Nowadays, natural disasters, particularly earthquake, are among the most important disturbances to sustainable development of countries and governments try to manage them in an optimum manner. They would typically try to decrease the amount of financial damages and loss of lives that would occur because of the events. In this research, geospatial information science/system has been implemented to estimate the seismic vulnerability and its probable damages for a particular scenario in Tehran. We applied fuzzy logic concepts to well known analytical hierarchical process for the damage assessment. Based on this modified approach we developed a hierarchy of effective factors in earthquake vulnerability due to definition of their priorities against a given earthquake scenario. The effect of uncertainty in geospatial data and analysis functions which are applied in estimating Tehran seismic vulnerability would affect the quality of decision making for estimating the damages. In this paper, we analyzed the implemented geospatial data of population statistics, building information, maps, digital terrain models and satellite images in the process of studying Tehran’s seismic vulnerability. In this research, we used Monte Carlo simulation approach for uncertainty modeling. We computed the statistical parameters for seismic vulnerability in various iterations of Monte Carlo process. After that we extracted the relationship between the number of Monte Carlo process and relative variation of seismic vulnerability’s layer variance. As a result of the analysis undertaken in this research, the number of iterations required to increase the precision of seismic vulnerability assessment is inversely related to the square root of the precision.