Assessment of Natural Hazard Damage and Reconstruction: A Case Study from Band Aceh, Indonesia

Abstract

There is an increasing interest in assessing the strengths and weaknesses of remote sensing imagery and geographic information system products as they relate to estimating populations at risk before, during, and after natural hazards. This research examines the spatial and temporal effectiveness of satellites and extent of damage products that were created for Banda Aceh, Indonesia after the 26 December 2004 tsunami. SPOT, FORMOSAT, MODIS and Landsat ETM+ imagery provided high temporal resolution data within three days of the tsunami. However, high-resolution commercial satellites (Quickbird, IKONOS) provide the most accurate data that can be used to assess infrastructure damage in cities like Banda Aceh before and after natural disasters. Of the six extent of damage products (USAID, USGS, Dartmouth Flood Observatory, DLR, SERTIT, DPRC) created after the tsunami, DLR provided the most accurate data on the extent of damage in Banda Aceh (94% agreement with Quickbird imagery). When these products are combined with population data estimates of populations at risk can be created to identify the areas most severely affected by a natural disaster. Pre-tsunami and post-tsunami imagery combined with imagery collected two and a half years after the tsunami show an extensive reconstruction of structures. Satellite imagery and geographic information system methods used in this research are most effective in regions where field data is sparse or difficult to collect within the first week following a natural disaster.