Master Thesis Defense by Leonie Röntgen

Title: Exploring the Regional Climate Response to Vegetation Dynamics and Disturbance Events with the Coupled Climate-Vegetation Model RCA-GUESS

Abstract:
Forest disturbances driven by wildfires, insect outbreaks, wind-throws, or human influence alter the structure, species composition, and function of forests and can therefore affect the forest carbon budget and other ecosystem services. The frequency of European forest disturbances has steadily increased during recent decades, while the temporal and spatial dynamics of disturbances are under rapid change. Nevertheless, biophysical feedback mechanisms associated with realistic patch-destroying disturbances have never been investigated in a European coupled climate-vegetation model.
In the framework of this master’s project, a new regional modeling approach and analysis methods have been developed to examine the impacts of realistic forest disturbances on the European climate using a coupled climate-vegetation model for the first time. Realistic disturbances described by Senf and Seidl (2021) are included as the spatial distribution of the disturbance return period in the European version of RCA-GUESS and compared to a model simulation with the default, fixed disturbance return period of 300 years.
The newly developed analysis methods identify gridcells with significant vegetation differences when including realistic disturbances. Atmospheric differences are classified based on the respective disturbance probability for gridcells with significant vegetation shifts. Results distinguish between the magnitude of disturbance-induced changes on gridcells with predominately
evergreen or deciduous vegetation. The Pearson correlation coefficient between atmospheric differences and the disturbance return periods is calculated for surface temperature, radiative and turbulent fluxes, wind speeds, total cloud cover, and precipitation to analyse albedo and evapotranspiration feedback mechanisms. Both analysis parts distinguish between European subregions to assess varying magnitudes with geolocation.
The study reveals that the effects of forest disturbances depend on the relative abundance of evergreen and deciduous species and vary with geolocation. Parts of the analysis indicate a disturbance-induced winter and spring cooling in northern Europe, which is associated with an increase in SW upward radiation due to the albedo feedback in agreement with previous studies on more extensive forestation scenarios. Low correlations between the model run differences and the map of disturbance return periods and very small differences between the vegetation covers in the simulations indicate that mostly natural variability between the runs has been observed. Stronger effects of disturbances are expected in future studies applying the method with an adapted spin-up procedure that incorporates the disturbance map earlier and allows the vegetation to be substantially altered by the disturbance return period. The inclusion of the disturbance map in RCA-GUESS and the provided analysis methods, combined with an adapted spin-up, will be a robust tool to assess feedback mechanisms related to forest disturbances and could be further applied in studies on forest management scenarios.

Supervisor: Jens Hesselbjerg Christensen and Paul Miller, Thomas Pugh and Adrian Gustafsson (Lund University)
Censor: Peter Aakjær