Seasonal redistribution of Forest Evapotranspiration Under Climate Change in Central Germany

Forest ecosystems are increasingly exposed to climate-driven shifts in water availability, yet the underlying changes in evapotranspiration (ET) processes remain poorly quantified at landscape scale. Building on a recently established operational soil-moisture monitoring framework for Central German forests (https://life.hydro.tu-dresden.de/BoFeAm/dist_bfa/index.html), we combine long-term observations and climate projections with process-based modeling to examine how climate change affects both the magnitude and the component structure of actual ET. Using the LWF-BROOK90 model, we quantify transpiration, rain and snow interception evaporation, and soil and snowpack evaporation across more than 3,000 forest sites in Central Germany. The model was driven by homogenized historical climate data (1961–2020) and an ensemble of 21 CMIP5 regional climate projections (2021–2100), covering major Central European tree species: Norway spruce (Picea abies), Scots pine (Pinus sylvestris), European beech (Fagus sylvatica), and pedunculate oak (Quercus robur).

Our results show a clear seasonal redistribution and shift of the forest water balance. Springs are projected to become wetter and more evaporative, supporting increased early-season transpiration. In contrast, summers exhibit strong declines in precipitation and reduced ET, accompanied by a substantial rise in soil-moisture stress days (REW < 0.4), particularly in upland conifer forests. Annual ET increases slightly due to higher winter rain interception, driven by a shift from snowfall to rainfall, most notably in Norway spruce. However, these annual increases hide growing summer water limitations, as higher evaporative demand exceeds declining soil-water supply.

By resolving individual ET components, this study highlights how changing climatic conditions propagate through canopy processes, soil moisture dynamics, and species-specific water use. The findings support assessments of forest resilience, help identify drought-tolerant species, and inform expectations of ecohydrological feedbacks in temperate agroforestry landscapes. Looking ahead, our combined monitoring–modeling framework is transferable to other regions where long-term climate data and stand-level information on vegetation and soils are available, supporting improved characterization of water and carbon cycle responses under future climate conditions.