MODIS-Based Investigation of Flood Areas in Southern Cambodia from 2002–2013

Abstract

In Cambodia and the Vietnamese Mekong Delta, floods commonly occur during the rainy season, and a better understanding of their spatio-temporal distribution is important for both disaster prevention and the improvement of agricultural production. This study investigated spatio-temporal flood inundation and land cover change from 2002 to 2013 in the southern part of Cambodia using Terra satellite on-board Moderate Resolution Imaging Spectroradiometer (MODIS) images. The algorithm for flood inundation detection, WFFI (Wavelet-based Filter for detecting spatio-temporal changes in Flood Inundation) was used, and the parameters were modified to fit the present study. The estimated inundation areas were validated using eight Landsat images. In a comparison between the original and modified WFFIs, the modified WFFI (70–96%) exhibited better accuracy than the original WFFI (30–70%). Overall, the temporal change in the flood inundation area presented a decreasing trend, and a link to the in-situ observed water level showed a decreasing trend during the rainy season. Furthermore, the estimated flood inundation exhibited a significant delay since 2008. Based on the yearly land cover MODIS product, the permanent water body and wetland areas decreased, whereas the cropland areas increased. This was as a result of increased agricultural productivity. However, water shortage was the major obstacle to increasing agricultural productivity, and it also had a negative impact on aquatic ecology, such as fish spawning grounds.