LANDSLIDE SUSCEPTIBILITY MODELING AND TREE GROWTH ECCENTRICITY ANALYSIS ON ACTIVE HILLSLOPES OF EAST TENNESSEE

Landslides are geohazards of concern around highly traversed corridors in East Tennessee. Roads through the steep slopes of valleys and ridges of the Southern Appalachian Mountains are susceptible to slope failure. State Route 116 in Anderson and Morgan counties, TN, has been heavily affected by unstable slopes, which has led to repairs by the Tennessee Department of Transportation (TDOT). The primary scope of this project is assessing the responsible geomorphological, geological, hydrological, and anthropogenic factors such as annual rainfall, soil, distance from streams, elevation, slope angle, slope curvature, bedrock geology, and land use in relation to the landslides in the study area. The secondary scope is to evaluate the effectiveness of tree growth rings in landslide prediction. One hundred ten landslides within the area have been identified with state LiDAR data, and about fifty percent were field verified. Geostatistical models, including Random Forest and Logistic Regression, were used to assess the contribution of the factors to the presence of landslides. Both models were used to produce landslide prediction maps that helped to visualize areas highly susceptible to landslides. The study indicated that the most critical factors were rainfall, elevation, and slope angles. In contrast, the soil, bedrock geology, and distance from streams did not show an association with the landslides in the area. This study found that numerous landslides have occurred in the study area and within the right of way of TDOT. Several landslides are active and have the potential to cause damage to major roadways. The study also found bent trees were ubiquitous in the active slopes. Therefore, representative bent trees were sampled by extracting tree rings with a borer. Tree-ring eccentricity analysis was performed based on the curvature of the tree along the unstable hillslope. The study found that tree ring analysis can effectively detect landslide movements and provide knowledge of landslide chronology. It is also possible to see past slope movement events caused by extreme rainfall events.