.png){kind=icon}
.png){kind=icon}
Return
Machine learning-based assessment of future debris flow susceptibility under CMIP6 precipitation scenarios in Qinghai Province, China
.png){kind=icon}
OA
AI
Q
K
W
L
L
DOI：10.1186/s40677-026-00393-5.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Climate change may affect debris flow occurrence by reshaping precipitation patterns, and the Qinghai–Tibet Plateau is particularly vulnerable, where debris flows often threaten local settlements and infrastructure. This study assesses current and future debris flow susceptibility in Qinghai Province by integrating environmental variables with precipitation projections from climate models. The objectives are to identify dominant conditioning factors, predict spatial and temporal changes under multiple emission pathways, and locate areas most sensitive to climate-driven disturbances. Extreme gradient boosting (XGboost) showed the strongest predictive performance in spatial cross-validation, with an area under the receiver operating characteristic curve of 0.950. High and very high susceptibility zones were mainly concentrated in eastern and central Qinghai, especially along river valleys and road corridors. Road distance and elevation were identified as the dominant predictors. Shapley Additive Explanations interaction analysis further indicated that the effects of precipitation variables can be amplified when they interact with topographic, drainage-related, and human-disturbance factors. Future changes were scenario-dependent, with the most pronounced expansion of high susceptibility areas projected under the high-emission scenario, where high and very high susceptibility zones together accounted for 8.5% of the study area. The transition matrix showed that susceptibility escalation mainly occurred in areas already classified as moderate or high under baseline conditions. Exposure overlay analysis revealed substantial overlap between future high susceptibility zones and densely populated and heavily built-up areas. The results suggest that climate change may intensify debris flow risk in Qinghai Province. Integrating machine learning with climate projections provides a reliable technical basis for identifying climate-sensitive areas and prioritizing risk prevention and control measures. Incorporating climate-adaptive susceptibility information into land use planning is essential for reducing future debris flow risks and protecting vulnerable communities and infrastructure.
Keywords:
Debris flow susceptibility
Climate change
CMIP6
Machine learning
Risk estimation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4
Papers:
264
Citations:
1.1K
Organization
Cited Papers