CONTROLS ON LANDSLIDE SUSCEPTIBILITY IN SOUTHEASTERN OHIO

The simplest landslide hazard maps delineate areas of negligible landslide occurrence and areas where landslides occur with a frequency or magnitude that may adversely affect people or property. An approach to create such a map over broad regions compares mapped locations of landslides with topographic and geologic characteristics to define a negligible-hazard zone. To this end, we used digital landslide inventory maps of a 23272.6 km² area of southeastern Ohio produced by the Ohio Dept. of Natural Resources, Division of Geological Survey; digital elevation data at 10 m pixel resolution; and digital geologic and hydrogeographic maps obtained from the U.S. Dept. of Agriculture. The landslide inventory contains polygons delineating landslides originally mapped on USGS 1:24,000 topographic maps and classifies landslides into three categories: active/recently active, past, and susceptible slopes. Past landslides and susceptible slopes are not incorporated into this study; we focused on active/recently active slides as they are defined uniformly throughout the quadrangle maps. We used a Geographic Information System (GIS) to analyze landslide area, slope angle, rock type, and local relief. Local relief was computed from a 10-m digital elevation model (DEM) by aggregating and differencing the maximum and minimum elevations within 10,000, 62,500, and 250,000-m² grid cells. Slope was computed for each 100-m² grid cell. Within each landslide polygon, we determined that 5-20° slopes have the highest occurrence of landslides in each rock unit. Mudstone and shale had the largest landslide area per total surface area. Calculation of local relief showed similar landslide distribution at different aggregate sizes. Slope angle and rock type had the greatest effect on slope stability, whereas topographic relief and proximity to channels had a small effect on landslide susceptibility. Of the 140,983 landslide points for slopes < 10°, only 3342 were located within 10-m of stream channel centerlines suggesting that only 2% of landslides resulted from river bank failure. Landslide density on low angle slopes (<7°, 10%) is 0.06 slides/km²; thus, the estimated likelihood of landslides occurring in these areas assuming historical evidence of landslides dates back 200 years, is very low (3x10^-4 slides/km²/year).