Joint 52nd Northeastern Annual Section / 51st North-Central Annual Section Meeting - 2017

Paper No. 25-12
Presentation Time: 1:30 PM-5:30 PM

DETERMINING LANDSLIDE SUSCEPTIBILITY OF CLARION COUNTY THROUGH THE USE OF A FREQUENCY-RATIO LANDSLIDE SUSCEPTIBILITY MODEL IN ARCGIS


HICKMAN, Neil Patrick1, STORER, Nathan Dale2 and KARIMI, Bobak1, (1)Biology and Geosciences, Clarion University of Pennsylvania, 840 wood st., Clarion, PA 16214, (2)Biology and Geosciences, Clarion University of Pennsylvania, 840 Wood St., Clarion, PA 16214, n.p.hickman@eagle.clarion.edu

A landslide is the downward and outward movement of slope-forming material, such as natural rock, soil, artificial fill, or a combination of these materials reacting to the force of gravity. The term landslide is a general term that includes rock falls, debris slides, earth-flows, and slumps among others. Clarion County is in the high to moderate susceptibility area of a Pennsylvania state-wide landslide susceptibility map. However, landslides are not considered a serious risk in Clarion County, and only one major landslide has been recorded occurring on May 22nd, 1983 causing minor traffic issues. While major susceptibility studies have been conducted for the state, this leaves too coarse of a model to accurately indicate landslide susceptibility within Clarion County. To develop a more refined and high-resolution landslide susceptibility map for Clarion, we use a frequency-ratio landslide susceptibility model in ArcGIS. In this raster based model, landslide susceptibility (S) for a pixel can be found using GIS techniques in which several contributing factors for the occurrence of landslide are explored and evaluated in order to find the Landslide Susceptibility Index (LSI). The landslide susceptibility for each pixel is determined by averaging the LSI from the input contributing factors. In pursuing this methodology, we required a digital point shapefile of landslides within Clarion County, which we created from a preexisting landslide map by the USGS. We explored and evaluated 5 m pixel resolution raster images of land cover, lithology, mean annual precipitation (MAP), groundwater depth, vegetation type, soil coverage, elevation, slope, and aspect for our susceptibility model. Many of these factors had to be converted to raster images based on vector datasets. We evaluated the LSI for each factor explored by performing a Receiver Operating Characteristic (ROC) curve analysis. This analysis is also used to evaluate the final susceptibility maps based on all, or some combination of contributing factors, and select the best susceptibility map for Clarion County. We present our refined, high-resolution landslide susceptibility map for Clarion County and a discussion of the largest contributing factors for landslides within the region.