Development of an Operational Satellite-Based Flood Monitoring Model for Tanzania

Andrew Kimati, Subira Munishi, Victor Kongo


Timely information during water related disasters is of utmost importance for flood preparedness and risk reduction. Real time observation and monitoring of flooded areas is an expensive and time-consuming exercise. Satellite remote sensing is a quick and affordable approach that can be used for concurrent floods detection at different scales. This is important as it facilitates timely information for emergency response to disaster management departments, even in scarcely instrumented catchments. This study presents a novel approach for flood tracking using satellite technology to map flood affected areas. An open-source water detection algorithm is developed that employs readily available satellite images and the Google Earth Engine (GEE) platform. Dar es Salaam and Singida regions in Tanzania were used as the case study for validation of the proposed approach. Use is made of Sentinel-1 satellite images and GEE coding. The after-flood tracking GEE code was validated with the physical flood extent markers and after-event flood extent survey points of the regions provided by the Ministry of Water (MoW). The findings reveal that the approach supports mapping flood extent areas by giving promising results after the satisfaction from validated data. Relevant parameters were then coded in order to develop the flood map of Tanzania. The findings of this study demonstrate the usefulness of open-source GEE in rapid flood inundation mapping. 

Keywords: Floods, Remote sensing, Flood tracking, Satellite images, Google earth engine


Full Text:



  • There are currently no refbacks.