BUILDING A LANDSLIDE INVENTORY FOR WEST VIRGINIA: STEP 1 IN STATEWIDE RISK ASSESSMENT

Recent funding from the FEMA Hazard Mitigation Grant Program and the WV Division of Homeland Security and Emergency Preparedness stimulated the initiation of a digital landslide inventory for the Mountain State of West Virginia. Over 75,000 previously documented landslides have been included. These landslide data come from a variety of sources, including the WV Division of Highways and student research projects, but most were mapped by the state and federal geological surveys between 1970 and 1990 using stereoscopic analysis of aerial photography. Over 6,000 additional landslides have been identified using 1 m DEMs based on QL2 LiDAR since January 2019. The overall landslide inventory is comprised of disparate types of point, line, and polygon data, collected from over a dozen sources with different purposes and perspectives, a fact that creates daunting uncertainty in meaningful interpretation as the project moves toward its ultimate goal of landslide risk assessment. Cautious analysis of these data indicates some preliminary generalizations can be drawn. More than 80% of newly mapped slope failures are translational or rotational slides, with blocky lateral spreads (associated with thick sandstone members) and long-runout debris flows (in high-relief areas with steeply dipping beds) constituting most of the remainder. Although rock fall is well represented in several historic data sets, there is little to no signature of individual rock falls on LiDAR-based DEMs. Most bedrock lithologies and topographic settings throughout the state are quite landslide prone, except broad low-relief valleys underlain by shale, siltstone, or carbonate bedrock in the Valley and Ridge physiographic province of eastern West Virginia. Most historic slides are associated with anthropogenic activities. Late Cenozoic drainage reorganization, valley incision, and periglaciation may be as significant as bedrock lithology and human activity in influencing the spatial distribution of landslides throughout most of the state.