Enhancing Disaster Response and Resilience Through Near-Time GIS for Flood Monitoring and Analysis in Niger River Basin, Nigeria
Keywords: Flood Monitoring, Disaster Response, Geospatial Analysis, Flood Resilience, Flood Mapping, NDWI, Google Earth Engine
Abstract. This study develops an integrated framework leveraging Google Earth Engine for near real-time flood mapping and impact analysis in the flood-prone Niger River Basin of Nigeria. Multi-temporal optical, radar and terrain data quantified changing flood hazards and exposure across 150,000 km2 between peak floods in 2022. Results indicate over 50% rise in inundation, with 15,000 hectares of vegetation and 143,000 residents enduring impacts. Attributing factors include elevated antecedent rainfall versus historical medians coupled with accelerating catchment modifications expanding runoff. Floodplain zones face recurrent impacts, necessitating adaptation. Accurate flood delineation was achieved by applying water indices like NDWI on Landsat and Sentinel-2 data using land cover and land use. Exposure analytics overlay flood extents on land use, infrastructure and demographic layers to estimate affected populations and livelihoods. Google Earth Engine enabled rapid data processing using cloud parallelization, while random forest integration powered machine learning semantic segmentation for robust feature extraction. Going forward, assimilating real-time data from radar and hydrological sensors would enable predictive flood risk models using machine learning algorithms on this cloud GIS platform tailored for resilience applications globally. In a changing climate, such scalable geospatial technologies provide evidence-based decision support capabilities to emergency planners targeting proactive adaptation investments for vulnerable communities based on quantified flood risk analytics.