Estimation of flood damage functions for river basin planning: a case study in Bangladesh

Located at the low-lying deltaic floodplain of Ganges-Brahmaputra-Meghna river basin, Bangladesh suffers damages from flooding with regularity. From the perspective of long-term planning and management, a reliable flood damage function is a critical component in the estimation of flood-induced economic loss. Such functions are, however, notoriously difficult to develop. This study utilizes in-stream water level and flood-affected area (FAA) data from Flood Forecasting and Warning Center and Bangladesh Water Development Board to evaluate the best form and data input characteristics of flood damage functions for Bangladesh. The performance of various function configurations (geographic data, water level data, and function form) was tested. The Nash-Sutcliffe efficiency and residual error analysis results suggest that, in general, the logistic function performs better than the other two function forms, and the maximum of daily-maximal water level is the best suited to estimate (FAA). As expected, when information is available from all basins (the Ganges, the Brahmaputra, and the Meghna), the resulting flood damage functions provide the most accurate estimations of FAA. Furthermore, the comparison between single- and multivariable flood damage functions does not demonstrate a clear advantage of using multivariate function in our study area. When flood damage functions with finer spatial and temporal resolution can be constructed using remote sensing technology or hydrodynamic modeling, the intra-year and district-level changes to FAA can be evaluated. These findings provide a better flood management plan for Bangladesh and have potential to be generalized to other similarly flood-affected nations.