Drought and late-frost risk in a changing climate

Project Description

Climate change is increasing both drought and spring late-frost events—two extremes that can severely damage forest ecosystems when they occur together. Yet, their statistical interdependence remains poorly understood. In this project, we developed a novel statistical framework to quantify the joint probability of drought and late-frost risk, focusing on Bavaria as a case study. Using high-resolution historical climate data and future climate projections, we modeled how the dependency between these extremes evolves over time. Our newly developed Y-vine copula regression model enables flexible modeling of complex dependencies and supports spatial risk prediction under different emission scenarios. The results provide a data-driven foundation for improved forest risk assessment and long-term ecosystem management under climate change.

Results

• Developed a novel Y-vine copula regression model for two-response regression settings
• Successfully captured heavy tails and asymmetric dependence structures between drought and frost
• Quantified the joint probability of extreme events in the historical record of Bavaria (1951–2020)
• Developed a new prediction method for spatial and temporal risk forecasting
• Demonstrated increasing joint mild-risk probability of drought and late-frost over time
• Enabled direct comparison between historic and projected climate conditions
• Provided a statistical basis for improved forest ecosystem risk assessment

Follow-up

An extension of this project involves expanding the spatial scope from Bavaria to the European domain. This could involve combining the development of high-resolution daily climate datasets with further refinement of the Y-Vine copula framework. The goal is to identify regional differences in the interplay between weather conditions, late-frost, and drought across Europe. This will support more comprehensive forest risk assessments and inform adaptive management strategies under accelerating climate change.