Validation of the Global NASA Satellite-based Flood Detection System in Bangladesh

Floods are one of the most destructive natural forces on earth, affecting millions of people annually. Nations lying in the downstream end of an international river basin often suffer the most damage during flooding and could benefit from the real-time communication of rainfall and stream flow data from countries upstream. This is less likely to happen among developing nations due to a lack of freshwater treaties (Balthrop and Hossain, Water Policy, 2009). A more viable option is for flood-prone developing nations to utilize the global satellite rainfall and modeled runoff data that is independently and freely available from the NASA Satellite-based Global Flood Detection System. Although the NASA Global Flood Detection System has been in operation in real-time since 2006, the ‘detection’ capability of flooding has only been validated against qualitative reports in news papers and other types of media. In this study, a more quantitative validation against in-situ measurements of the flood detection system over Bangladesh is presented. Using ground-measured stream flow data as well as satellite-based flood potential and rainfall data, the study looks into the relationship between rainfall and flood potential, rainfall and stream flow, and stream flow and flood potential for three very distinct river systems in Bangladesh - 1) Ganges- a snow-fed river regulated by upstream India 2) Brahmaputra - a snow-fed river that is also braided 3) Meghna - a rain-fed river. The quantitative assessment will show the effectiveness of the NASA Global Flood Detection System for a very humid and flood prone region like Bangladesh that is also faced with tremendous transboundary hurdles that can only be resolved from the vantage of space.