SAR-NDWI Integration in Ephemeral Water Detection Mapping for Hydrological Analysis in ASALs

Abstract. Arid and semi-arid lands (ASALs) like Marsabit and Isiolo Counties in Kenya face significant water scarcity, necessitating accurate mapping of surface water resources. Even with mostly dry and arid conditions in these regions, the areas also experience flash floods and downpours that quickly flow through wadis and rivers flowing away from these regions. This study leverages a multi-sensor remote sensing approach, combining Synthetic Aperture Radar (SAR) data from ALOS PALSAR and optical data from Landsat 8 (NDWI), to map water availability in the years 2020 and 2019. The research being the first step series to leveraging multi-sensor analysis and Digital Twins to understand the dynamic nature of ASALs. The results highlight the superiority of SAR in detecting ephemeral and obscured water features, revealing a vast “SAR-only” water area compared to a minimal “common water” area identified by both sensors. A novel experimental Multi-Sensor Water Index (MSWI) was developed to enhance detection accuracy, demonstrating its utility in capturing dynamic hydrological processes. These findings underscore the critical role of SAR in water resource management for ASALs, providing insights for drought resilience, pastoralist livelihoods, and sustainable water planning.