Pukyong National University

Pukyong National University Research Team Publishes Consecutive Studies in Landslides

- Research achievements by Dr. Chang-Ho Song, Ph.D. candidate Ho-Hong-Duy Nguyen, and Professor Yoon-Tae Kim

Research conducted by a team from Pukyong National University (President Sang-Hoon Bae)―comprising Dr. Chang-Ho Song of the Smart Infrastructure Technology Research Institute, Ph.D. candidate Ho-Hong-Duy Nguyen of the Department of Ocean Engineering, and Professor Yoon-Tae Kim of the Department of Ocean Engineering―has been consecutively published in the May issue of 'Landslides', one of the world’s leading international journals in the fields of landslide science and geotechnical engineering.

The research was carried out as part of the Marine Urban Disaster Mitigation Technology Education and Research Team, which is supported through the BK21 FOUR Program funded by the Ministry of Education. The team published two papers: “Physically Based Data-Driven Analysis for Large-Scale Investigation of the July 2025 Rainfall-Induced Landslide in Sancheong, South Korea” and “Deep Neural Network Framework for Predicting Debris Flow Entrainment Growth Rate in Diverse Terrain Conditions.”

The first paper was authored by Ph.D. candidate Ho-Hong-Duy Nguyen as the first author, with Dr. Chang-Ho Song serving as a co-author, while the second paper was led by Dr. Chang-Ho Song as the first author, with Ho-Hong-Duy Nguyen participating as a co-author. Professor Yoon-Tae Kim served as the corresponding author for both studies.

is a globally recognized international journal representing the fields of landslide science, debris flow research, slope failure analysis, and geohazard engineering. It is widely regarded as one of the top-ranked journals in geotechnical engineering according to the Journal Citation Reports (JCR) published by Clarivate Analytics.

The first study presents a large-scale landslide investigation methodology that integrates physics-based analysis with data-driven analytical techniques to examine the rainfall-induced landslide that occurred in Sancheong County, Gyeongsangnam-do, South Korea, in July 2025. By comprehensively considering slope instability mechanisms caused by rainfall infiltration together with topographic and geotechnical characteristics, the research team was able to effectively analyze the actual behavior and triggering mechanisms of the landslide event.

The second paper developed a deep neural network-based framework for predicting the entrainment growth rate of debris flows under a wide range of terrain conditions. By integrating artificial intelligence-driven analytical techniques with information on topographic, hydrological, and geotechnical characteristics, the research team proposed a methodology capable of quantitatively predicting how debris flows increase in size and volume as they travel downslope.

The researchers expect that these studies will contribute to the development of foundational technologies for more accurate prediction and response to landslides and debris-flow hazards in an era characterized by increasingly frequent extreme rainfall events and compound disasters driven by climate change.

Professor Yoon-Tae Kim stated, “As climate change continues to increase the risk of geotechnical disasters such as landslides and debris flows, we plan to further advance the development of high-precision disaster prediction and response technologies through research that combines physics-based analytical approaches with artificial intelligence technologies.”