Developing Building and Population Inventories Using Dasymetry and Remote Sensing Techniques for Urban Loss Estimation – Case Study: City of Sari

Residential buildings and associated population distribution modeling, is an integral part for assessing and estimating the direct and indirect risks/losses related to natural disasters, including earthquakes in the built environments such as in urban and rural settings. In this study, the geospatial database for the buildings were derived by compiling existing census data sets and the geospatial map that was derived by the comprehensive urban planning and development of the study area (Sari metropolitan area in Mazandaran, north of Iran) as obtained by the related consultant engineering firm. Combining the information from different data sets with different scales or spatial resolution involves some techniques that are mainly based on statistical inferencing methods and/or utilizing some spatial distribution patterns that exhibit the built environment topology. Once the distribution of the buildings or the residential units are derived and finalized, the population distribution pattern can be extracted based on similar aforementioned concepts. The second step in this research involved the development of the population inventory. In this study, as mentioned earlier, the statistical datasets from different sources (related to urban zones and city blocks) were combined first. For the purpose of comparison and overall verifications, some digital imageries such as aerial and satellite images were collected in order to extract urban features such as buildings. In this fashion, detailed geometric information concerning each individual building footprint are extracted that can be summarized according to their related city block polygons. This scheme was applied for some selected parts of the city that can be evenly utilized and extended for any desired geographic extents. In conducting the present research, a simplified method for building height (or city parcel height) assignment was also delivered that indicates the average number of stories within each city block. After the urban database has been developed (building and population inventories), the earthquake damage to the elements at risks (i.e. building destruction and human casualty) can be assessed in the study area. As a case study, the metropolitan area of Sari was focused, and some probable hypothetical scenario earthquakes were considered. For each scenario earthquake, damages to the building stock and the associated human losses were estimated according to some predefined domestic vulnerability functions and a global human casualty model as reported. For example, as a result, considering the Khazar Fault earthquake scenario, more than 5000 housing units are expected to experience very heavy damage to destruction. This study has presented a comprehensive model according to the existing data sets and the local conditions for the study area that can be modified, upgraded and updated with newer and more accurate data whenever accessible. Also the same method can be applied for other parts of the country. Moreover, the results indicate that the expected damages, losses and human casualties are high and alarming suggesting more detailed study on the region and also devising and implementing suitable disaster management measures for the region.