Optimising landslide trigger analysis through integration of diverse dendrogeomorphic approaches and multi-source climate and documentary data

Abstract

Landslides on forested slopes, which often lack extensive documentary data and monitoring, can be evaluated using tree-ring records. This method is commonly employed for assessing hydrometeorological triggers similar to sites with documentary-based landslide catalogues. However, uncertainties arise in setting dendrogeomorphic thresholds and using diverse climate data sources for trigger identification. This paper reconstructs the activity of six landslides using tree-ring records in a region with well-documented landsliding and long-term climate data from rain gauges. The objective was to test different dendrogeomorphic approaches for identifying landslide event year and to analyse various rain, snow, and soil moisture parameters from meteorological stations and climate reanalyses to identify triggers. This involved examining how variations in each research step affect results and potentially enhancing the interpretation of landslide triggers. Documentary data indicated frequent landsliding during spring and summer, while tree-ring chronologies showed the strongest connection to increased winter and spring precipitation measured by rain gauges, suggesting an enhancement of winter precipitation as a potential climate preparatory factor. Furthermore, the results showed that combining tree-ring and documentary data yielded more statistically informative links than using tree-ring data alone. Additionally, integrating gridded data from ERA5-Land climate reanalyses enhanced the interpretation of explanatory factors. While increased winter and spring precipitation were most strongly linked to landslide initiation using precipitation data, volumetric soil water content derived from reanalysis data was crucial for identifying summer landslide occurrences. Incorporating soil water content into dendrogeomorphic research improved the interpretation of landslide initiation timing. Therefore, regional tree-ring-based chronologies with improved identification of landslide triggers can significantly reduce uncertainties in decision-making when it comes to landslide hazard management. Future research should explore the effectiveness of freely available satellite-based precipitation data with tree-ring chronologies, particularly in regions with limited information on precipitation intensity.