Multi-source Earth Observation Derived Data for Delineating the 2011 High Flood Line in the Okavango Delta (Botswana) for Flood Risk Mitigation and Management

Globally, climate-related extremes, particularly floods and droughts, affect freshwater ecosystems and related human systems. The seasonal, annual, and decadal inundated extent of the Okavango Delta in north-west Botswana is highly variable. In 2011, an extreme regional flooding event occurred in the Cubango-Okavango River Basin because of a relatively high rainfall event at its upstream area in the highlands of Angola. This resulted in flood hazards to natural resources in the Okavango Delta wetland system and flood risks to communities living in the Basin. Substantial damages to property and infrastructure occurred in Namibia along the Kavango River and within the Okavango Delta and its proximity in Botswana. The effect of the high flood pulse was felt in downstream areas, particularly in the town of Maun. Therefore, we aim to develop an integrated wetland flood mapping method for populated areas in the Cubango-Okavango River Basin. The objectives of our study were to integrate high spatial resolution Earth observation data to delineate high flood waterline in the Okavango Delta, identify floodplain areas and vegetation cover types that are susceptible to flooding, and develop flood risk mitigation and management strategies. RapidEye data were used to calculate normalized difference water index (NDWI), while Sentinel-2 imagery and shuttle radar topographic mission (SRTM) digital elevation model (DEM) were used to derive recent baseline vegetation cover types. Expert knowledge on the area, existing topographic maps, review of documentation by local land management authorities on the flood event, and ground surveys were used for validation of the results. Preliminary results reveal that flood recession farms on the eastern part of the Okavango Delta Panhandle are susceptible to flooding. Additionally, Shashe River floodplains in the town of Maun are susceptible to flooding; hence, the allocation of land parcels along the area should be circumvented. The results of our study contribute to the development of large-scale flood risk map products for populated areas around the Delta. Also, the study provides recent baseline maps on vegetation cover types in the Okavango Delta system at a landscape scale for wetland monitoring. These are essential elements of integrated wetland flood mapping.