Rapid loss assessment of population in seismic disasters

Earthquakes are one of the most threatening natural disasters, due to their suddenness and unpredictability. Population loss assessment is the sole basis for earthquake rescue. Compared with the existing methods based on the buildings, vulnerabilities often require accurate building data to evaluate building damage. This paper proposes a method for fast loss assessment of population without a detailed distribution data of building. In this paper, a casualty loss model of PAGER was used to estimate the fatality rate. The model is expressed in the form of a lognormal cumulative distribution functions of seismic intensity with two free parameters. Lacking the township population distribution data, this research first convert the building regions of Google Earth into polygons, and then take the polygon area ratios as population proportion to obtain the amount of population in each shaking rank. Based on the fatality rate and population distribution data, the number of deaths and homeless can be calculated, and then an estimate of how much relief water, food, and tents the government should prepare. The proposed method was exemplified in the Ludian earthquake. Compared with the actual casualties of Ludian, the estimated result shows that the error range is less than 30%. Therefore, it can be used to provide the support of disaster data for rescue.