LANDSLIDE HAZARD ASSESSMENT FOR NORTHERN FAYZ ABAD DISTRICT, BADAKHSHAN PROVINCE, AFGHANISTAN

Natural disasters in Afghanistan, including avalanches, earthquakes, floods, and landslides, affect more than 200,000 people each year. Landslides are a primary risk because they often initiate through other hazards, such as earthquake shaking, rainfall and flood related infiltration, undercutting, over steepening, and runoff. This research is developing a landslide hazard package for Fayz Abad District, Afghanistan, a region that experiences intense seismic activity, monsoons, and landslides. Products includes a landslide inventory, landslide susceptibility map, and hazard map focusing on at-risk settlements.

Landslides are mapped via heads-up-digitizing using high-resolution satellite imagery and 5m Digital Elevation Model (DEM) derivatives, including slope, cosine of slope, and hillshades with different solar angles. Delineation parameters include sharp slope and elevation changes that reflect head scarps, slide scars visible by escarpments and stripped vegetation.

A landslide susceptibility model is constructed using a lithology, tectonic, and hydrologic data following established methodologies. Drainage systems are delineated using hydrology tools in ArcGIS with 5m DEMs. Incremental buffers around rivers and faults represent proximal influence of features on landslide susceptibility. Variables divided into nine susceptibility levels, low to high, are assigned standard weights to ensure like-to-like comparison for model implementation.

Developing a hazard map for Afghanistan is difficult because of displaced populations and incomplete census data. For district population, density estimates modeled from old census data and land usage remote sensing were used. Settlement boundaries were digitized from satellite imagery to represent population locations. Local modeled landslide susceptibility values and landslide density from inventory around populations determine risk.

Results suggest landslides in Fayz Abad district are highly controlled by local hydrology. Debris flows are visible in headwaters, and slumps, flows, and apparent creep widely seen in association with basal incision. Agricultural populations residing in valley bottoms downstream from headwater regions and in proximity to incising slopes seem to be at greatest risk to landslide events.