Can We Model Floodplain Inundation Patterns in Data-Scarce Areas?

Abstract

This paper proposes a methodology to model floodplain inundation patterns in data-scarce areas by using global remote sensing data. In particular, MODIS data are used for hydraulic model (HEC-RAS) calibration and validation purposes which is coupled with Geographic Information System (GIS) to map flood extent areas, while NASA's SRTM is used to describe floodplain topography. The Fogera floodplain (the upper Blue Nile in Ethiopia) is used as an example application to illustrate the methodology. To this end, parameter and input uncertainty is explicitly taken into account and visualized via probabilistic floodplain maps of the ensemble simulation. In view of that, model performance, reliability, and predictive uncertainties are critically discussed. This approach revealed that a
better characterization and visualization of the flood hazard. Also, the study investigates the impact of land-use changes on floodplain inundation patterns using a SWAT modeling system and the propagation of this land-use change in flood inundation patterns is seen again via probabilistic flood maps. This helps planners to use remote sensing data to model and monitor flooding.