Date of Graduation
8-2024
Document Type
Thesis
Degree Name
Master of Science in Geology (MS)
Degree Level
Graduate
Department
Geosciences
Advisor/Mentor
Aly, Mohamed H.
Committee Member
Lamb, Andrew P.
Second Committee Member
Boss, Stephen K.
Keywords
Geohazards; GIS; Landslides; Machine Learning; Natural Disasters
Abstract
The purpose of this study is twofold: the first involves examining the effectiveness of the Streamlined Landslide Identification Protocol (SLIP) as a method for creating an accurate landslide inventory as well as examining the densities of the landslides within our study area, the second purpose of this study is examining a machine learning methodology known as Random Forests for predicting the occurrence of landslide based on likelihood within our study area. The area of interest for this study is Greer’s Ferry Lake Arkansas. In the first paper I used the Streamlined Landslide Identification Protocol (SLIP) to create a detailed landslide inventory for Greer’s Ferry Lake, Arkansas, and analysis landslide densities for the surrounding area. The landslide inventory (SLIP) was created using high-resolution (1-m) lidar where we created a Digital Elevation Model (DEM) and Digital Slope Model (DSM) for the purposes of identifying landslide occurrences. The SLIP approach to landslide inventories involves using the high resolution (1m) lidar and individually classifying landslides within a given study area based on observable landslide features (head scarp, internal scarp, flanks, internal features, and the toe). The 2,610 landslides that were identified through the (SLIP) landslide inventory was then used as a basis for a density analysis of landslides found in Greer’s Ferry Lake, Arkansas. The second paper involved a machine learning technique known as Random Forest, which I applied to the landslide inventory created from (SLIP) to create a weighted landslide susceptibility map for Greer’s Ferry Lake, AR. To create our Random Forest predictive model, we used 9 potential Landslide Triggering Factors (LTFs): slope, aspect, and flow direction, distance from roads, distance from faults, precipitation, soil type, lithology, and land use and land cover. The 9 LTFs were weighed in the importance for landslide susceptibility which is determined through the Random Forest model. Through the creation of the landslide susceptibility map we aim to determine the relationships between the landslide triggering factors and landslide occurrences. Ultimately the aim of this study is to provide a resource for officials and residents on areas within our study area which have a high susceptibility to landslide occurrence and other forms of mass wasting events, so they may plan accordingly.
Citation
Rameden-Tipton, T. (2024). Predicting and Analyzing Landslides Dynamics for Greer’s Ferry Lake, Arkansas. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/5532
