Identification of High Risk Areas Through Integrated Building Inventories

The assessment of the risk to urban areas is of primary importance for decision makers, insurance and investment portfolio mangers, and government emergency and disaster mitigation planners. In this report we present a method for estimating the expected direct dollar loss from building damage in an urban area subjected to earthquakes. The method considers the probabilities of occurrence of earthquake events on any fault in the region, the distribution of buildings, their physical characteristics and earthquake resistance, and the damage potential of different class buildings.

One of the key issues in damage and loss estimation is the consolidation of a building inventory that contains sufficient information on building characteristics. This report presents a method for combining building information from existing inventories using relational data base management systems (RDBMS). In particular the RDBMS ORACLETM is used for this purpose. Damageability of different classes of structures is represented by the damage-motion relationships developed by the Applied Technology Council (ATC,1985). A theoretical formulation for computing expected losses is presented in this report. The formulation utilizes the information on seismic hazard exposure, building inventories and damage-motion relationships.

Integration of the information on seismic hazard, building inventories and damage potential of building classes is performed within a geographic information system (GIS). These systems are particularly useful when data are spatially distributed or results are to be displayed over a region. Building inventories for the city of Palo Alto were obtained in order to test the model. The classification of buildings was based on the information available in the inventories. When such information was insufficient for direct classification, a probability distribution of likely building classes was developed. The seismic hazard for this area was obtained using the program SHA (Lamarre, 1988). The hazard is estimated in terms of annual probabilities of exceeding peak ground acceleration at various sites. In this study, the amplification of the soil was not included although the importance of this parameter is recognized. Given the seismic exposure and building classifications, expected damage distribution was obtained and displayed using the GIS ARC/INFOTM. Damage distribution is developed for a future time period of 50 years. The dollar damage ratios obtained in this example are only for purposes of demonstrating the model rather than forecasting absolute values of damage and loss. Since building practices are relatively uniform in this location, the distribution of damage appears to be governed by the hazard exposure.