HARNESSING ASTER-DEM-DERIVED WATERSHED FEATURES FOR LANDSLIDE SUSCEPTIBILITY MAPPING

Acquiring the exact and unrestricted watershed delineation produced using Digital Elevation Model (DEM) data is challenging for landslide susceptibility modelling. High-resolution DEMs are crucial for accurate landslide susceptibility mapping, capturing detailed terrain features, slope gradients, and watershed boundaries. Accurate watershed delineation needs high-resolution DEMs that find the data, manage the complex processing, and manage the high costs involved; thus, ASTER open-source DEMs, with their moderate resolution, offer a more accessible option when high-resolution data is not available. This study explores open-source capabilities for ASTER 30m satellite images from the EarthData website to delineate Hulu Langat, Selangor watersheds. The DEM and landslide datasets (relief and slope, land use, proximity to rivers and roads, slope, lithology, and aspect) were transformed to the same projection and coordinate system. The ASTER 30m was merged and clipped based on the Hulu Langat district boundary lines. The watershed delineation for the entire Hulu Langat Selangor was then determined for all datasets using the Arc Hydro tools in Geographical Information System (GIS) software. Maps of the Hulu Langat region's landslide susceptibility that show the incredibly high, high, moderate, and low-risk zones were produced using DEM watershed datasets from the National Geospatial Centre (PGN) and the open sources platform of ASTER and ArcGIS Pro's overlay tool. A comparative study of classification techniques (Equal Interval and the Natural Breaks) and past landslide events was also conducted to explore the map's reliability and practicality. The regions vulnerable to landslides were then depicted on a map created by spatially analysing these characteristics to apply in real situations of the study area. ASTER open-source DEMs offer a practical solution with moderate-resolution elevation data, making them useful for broad watershed delineation and improving landslide susceptibility mapping, especially when high-resolution data is hard to find.