Bridging Climate Science and Adaptation Plan: Operationalizing Landslide Risk Management under Climate Change Scenarios

Translating global climate projections into decision-relevant information for climate adaptation is a critical hurdle for applied geosciences. This study presents a climate-informed landslide risk mapping framework developed for Taiwan, designed to bridge climate science with operational landslide risk management under climate change. Statistically downscaled daily precipitation projections from CMIP6 are employed to characterize future rainfall extremes, integrating them with geological susceptibility, bare land ratio, and population density to represent hazard, vulnerability, and exposure, respectively. Relative landslide risk is assessed using a quantile-based classification approach under Global Warming Levels (GWLs) of 1.5 °C, 2 °C, and 4 °C. To support applications across multiple decision scales, landslide risk maps are generated at 5 km grid resolution for regional-scale screening, at the township level for administrative planning, and at minimum statistical areas for detailed exposure assessment. The results demonstrate a consistent intensification of landslide risk with increasing global warming levels. Significantly, mountainous regions in northern and eastern Taiwan exhibit a nonlinear expansion of high-risk clusters under the 4 °C warming scenario, indicating heightened sensitivity to extreme precipitation changes. To explicitly address uncertainty in climate model projections, the framework incorporates a risk credibility indicator based on inter-model agreement, enabling a transparent interpretation of model robustness and avoiding deterministic use of climate projections. The framework has been operationalized through the Climate Change Disaster Risk Adaptation Platform (Dr. A), a web-based geospatial decision-support system that allows users to visualize landslide risk patterns across warming scenarios and to perform spatial overlay analyses with infrastructure datasets such as transportation networks and settlements. By providing multi-scale and scenario-based risk information, this study contributes a transferable methodology for integrating climate projections into landslide risk assessment and adaptation planning within regions.