Machine learning-based rainfall-induced landslide susceptibility model and short-term early warning assessment in South Korea

20613 6 109899 2 333128 document 109899 s10346-025-02513-y.pdf application/pdf 679526bfa7014a43f895a4856e58a7c2 MD5 8085069 2025-05-23 06:08:34 https://pure.iiasa.ac.at/id/eprint/20613/1/s10346-025-02513-y.pdf 20613 1 1 application/pdf text en public cc_by_4

s10346-025-02513-y.pdf

published 109910 1 333144 document 109910 indexcodes.txt text/plain 45bfa340793f9e3c29a637e85b042a5b MD5 17498 2025-05-23 06:18:51 https://pure.iiasa.ac.at/id/eprint/20613/2/indexcodes.txt 20613 2 2 text/plain other Generate index codes conversion from text to indexcodes en public

indexcodes.txt

lightbox.jpg

preview.jpg

medium.jpg

small.jpg

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/109899 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/109899 http://eprints.org/relation/issmallThumbnailVersionOf https://pure.iiasa.ac.at/id/document/109899 archive 353 disk0/00/02/06/13 2025-05-23 06:09:57 2025-05-23 06:09:57 2025-05-23 06:09:57 article show Lee S. Roh M. Jo H.-W. 9770 0000-0001-6127-883X Joon K. Lee W.-K. Machine learning-based rainfall-induced landslide susceptibility model and short-term early warning assessment in South Korea pub BNR AFE The extreme rainfall events associated with climate change trigger landslides. Approximately 60% of South Korea comprises mountainous terrain, with steep slopes rendering it particularly prone to landslides. Despite the implementation of early warning systems by the Korea Forest Service (KFS), landslide damage remains substantial, with approximately 2345 hectares affected over the past 5 years, resulting in severe human and economic costs. The current 24-h early warning system, based on Tier 3 administrative division (Town), faces challenges in accurately identifying high-susceptibility landslide areas. Thus, a daily landslide susceptibility model that integrates landslide-associated conditioning factors with meteorological, topographic, and environmental data was designed to assess landslide susceptibility with a spatial resolution of 100 m. Using AutoML, we identified Random Forest as the optimal model for predicting landslide susceptibility. Training the model with landslide data from 2016 to 2022 resulted in an accuracy of 0.93, AUC of 0.98, and F- 1 score of 0.98. A kappa value of 0.85 indicated the effective classification of past landslides using testing data. Location-based validation using 2023 occurrences revealed highly susceptible classifications for 88% of 43 landslides, while spatial scale-based hazard assessment using observed data indicated high hazard for 96% of 607 landslides in Tiers 3 and 4 (Township). Weather forecasting was also found to affect accuracy, with 76% accuracy for forecasts made at 5:00 PM and 41% for forecasts made at 8:00 AM. It was confirmed that further calibration of forecasting data can enhance the performance of the susceptibility model. The designed process thus enhances landslide prevention and preparedness on both local and regional scales, offering a crucial tool for mitigating the impact of landslides in South Korea. 2025-04-16 published Springer doi:10.1007/s10346-025-02513-y 3478 public Landslides TRUE 1612-510X info:eu-repo/semantics/article