MODELING AND VERIFYING AFFECTED POPULATION AND MORTALITY ON NORTHERN SUMATRA AFTERMATH THE INDIAN OCEAN TSUNAMI, 2004

Many existing analytical methods use Geographic Information Systems (GIS) and remote sensing to model tsunami extent and affected population. Data for analysis come from field surveys, digital data, maps and satellite imagery. However, the results of modeling are rarely verified with field data, especially if the modeling output is human loss. Field verification is expensive and sometimes impossible due to the logistics reasons. While verification of models is a necessary component it is not always present.

Conducted study is an example where GIS based model of tsunami affected population was verified with actual collected data on population mortality rates. GIS model consisted of spatially distributed data on elevations, shoreline orientation, administrative units and population. Survey on mortality was done aftermath tsunami; data were collected in camps for displaced population. Modeled and surveyed data were then compared and statistically analyzed. In addition, GIS model of the spatial extent of disaster was verified using similar model produced by remote sensing technique.

The results showed high correlation between modeled affected population and surveyed mortality data; high correlation was also found between spatial extent of the disaster produced by GIS and by remote sensing method. While GIS modeling was done within several days, the actual estimation of mortality by survey took almost a year. The use of surveyed data improve spatial model and can be used to analyze factors contributing to mortality rates thus creating a potential for GIS based mortality model. Developed spatial GIS model can be used during disaster management to allocate efficiently emergency resources for affected population.