Spatiotemporal Dynamics of Flood Exposure in Bangladesh: A GIS and Remote Sensing Based Approach

The need for flood risk evaluation in Bangladesh is underscored by its unique geographic location, dense population, and regular encounters with severe hydro-meteorological phenomena. Previous literature has brought broad attention to flood extent mapping and simulation modeling; however, a gap remains in understanding the spatial and temporal dynamics of flood exposure in Bangladesh. This study attempts to fill that gap and introduces a state-of-the-art flood risk assessment methodology, focusing particularly on the population being exposed. The methodology leverages the processing and visualization capabilities of Google Earth Engine and ArcGIS Pro to conduct a weighted linear modeling (WLM) approach using Sentinel 1 SAR and Gridded Population of the World (GPW) as input features. This method identifies the recurrent zone of flood impact and quantifies the population directly exposed to those recurrent flood events. The results of this study deduced that around 15,649,130 individuals in the country were directly susceptible to flood risk in 2023 during monsoon season, with the northeastern zone being especially hard-hit. Approximately 33.46% of the population in Sylhet division was directly affected by the flood. Our findings, derived from a seamless integration of high spatial resolution multispectral remote sensing and a linear overlay model, provide a granular perspective on flood exposure dynamics in Bangladesh that is necessary to fortify the flood preparedness mechanism. Overall, this paper creates a space for future flood exposure and risk assessments in Bangladesh by replicating the geospatial modeling approach initiated in this research.