Evaluating root strength index as an indicator of landslide-prone slopes in eastern kentucky

Considering stabilizing effects of vegetation may improve landslide susceptibility analysis, but deriving metrics such as root strength is slow and difficult to accomplish at regional scales. A lidar-derived root strength index (RSTI) may serve as a proxy for regional root strength estimates but has not been widely leveraged in statistics-based landslide susceptibility modeling, nor has it been implemented without time-intensive ground-truthing. Landslides (n=1086 documented) triggered by record precipitation in July 2022 in eastern Kentucky provided a unique opportunity to incorporate a purely GIS-derived estimate of RSTI due to the availability of high-resolution, pre-failure lidar data. During this event, landslides predominantly originated on or immediately downslope of areas where RSTI was relatively low. Most (83.2%) landslides occurred where mean RSTI is <9.3 and mean slopes are >12°. Slopes <12°, not typically considered susceptible to failure, still experienced landslides where mean RSTI was <9.3. Approximately 24% of landslides >1200 m² (n=15 of 62) occurred where RSTI was above a threshold of 9.3 and slope exceeded 32°, which indicates that a moderate to high RSTI is unlikely to offset the destabilizing effects of rapid precipitation on steep slopes with thin soil. Combining RSTI thresholds and regional slope cutoffs successfully accounted for 75.5% of landslides triggered by the July 2022 storm event, suggesting that slope and RSTI could serve as a reasonable model on their own. Furthermore, statistics-based models demonstrated an 8% improvement when RSTI was incorporated in addition to typical topographic parameters. This RSTI workflow can be easily incorporated into refined susceptibility workflows in other landslide-prone regions where high-resolution lidar data is available.