Quantitative evaluation of landslide susceptibility FOR four counties of Northeast Ohio

Arpita Nandi¹ & Abdul Shakoor² 

¹ Michael Baker Corporation (e-mail: anandi@mbakercorp.com)

² Kent State University (e-mail: ashakoor@kent.edu)

      Abstract: This study quantitatively evaluates the factors responsible for the spatial distribution of landslides in northeastern Ohio.  The area is well known for creep, rotational slides, and earth flows occurring in silts and silty clays along the stream valleys.  Landslide activities in the study area depends on multiple factors such as slope angle, slope aspect, soil type, soil erodibility, engineering properties of soil, precipitation, proximity to stream and flood-prone area, and land cover pattern.  A comprehensive, multi-source database was prepared which contained the factor layer information responsible for spatial distribution of landslides.  In order to examine the effect of various factors on landslide susceptibility, Susceptibility Priority Number (SPN), Landslide Susceptibility Index (LSI), and logistic regression models were used.  Landslide susceptibility maps were prepared in ArcGIS using these three models which classified susceptibility into four categories: low, moderate, high, and very high.  Logistic regression model yielded the most reliable results and corresponded with the actual physical conditions of the area.  In this model more than 81% of the existing landslides were plotted in the high and very high susceptibility zone.  The results also indicated that factors such as slope angle, proximity to stream and flood-prone area, and soil erodibility were statistically significant in controlling the slope movement in northeast Ohio.