ForClim

Reserach Project

Climate change significantly affects forest ecosystems, manifesting in ways such as droughts, wildfires, and diseases. The regional climate model MAR (Modèle Atmosphérique Régional), developed at ULiège, allows us to assess the current and future climate over Belgium at a 5km resolution. However, uncertainties persist in the model estimations due to the sensitivity of mechanistic models to error propagation and input data quality. Additionally, MAR lacks a comprehensive dynamic vegetation modelling, preventing accurate near-surface estimations.

This project seeks to explore the impacts of climate change on forest ecosystems by integrating a dynamic vegetation modelling, CARAIB, into MAR, and by assimilating remote sensing data such as MODIS Leaf Area Index (LAI) to refine the model inputs. The first results show a significant sensitivity of MAR results by enhancing the accuracy of LAI given as vegetation input into MAR.

New simulations over the period 2015-2022 indicate an increase in evaporation and transpiration across agricultural and forested areas. The increase in evapotranspiration correlates here with a decrease in soil water content. Comparisons between the previous model outputs and the new ones as well as with observations indicate a reduction in bias, but some discrepancies remain. This suggests that, while MAR represents the climate accurately on average, it struggles to replicate specific events, hampering the study of extreme events partially because of its sensitivity to vegetation.

The decrease of this sensitivity, and consequently a decrease in the uncertainties in the vegetation-climate interaction, can be reached by implementing a dynamical vegetation module into MAR.

The new implementation is achieved by coupling MAR with an existing dynamical vegetation model, the CARbon Assimilation in the Biosphere (CARAIB) model. By running concurrently, the two models can provide each other with the variables they require as inputs or replace the outputs of the other model. In this case, MAR would provide CARAIB with all the climatic data (precipitation, temperature, …) and CARAIB would provide MAR with the LAI required as input.

Nevertheless, this project aims to improve our understanding of the interaction between climate and forests. By refining modelling, this research strives to support more accurate predictions and strategies for forest management under changing climatic conditions.

This project sits at the intersection of climatology and geomatics, blending remote sensing with vegetation and climate modeling. By merging these disciplines, it encapsulates multiple branches of geography, offering a comprehensive approach to understanding environmental dynamics. Through this synthesis, it reveals the intricate connections between climate, vegetation, and spatial patterns, contributing to a deeper understanding of our changing world.

The Team

Thomas Dethinne (PhD Student)

Dr. Nicolas Ghilain

Prof. Xavier Fettweis (Co-Supervisor)

Prof. François Jonard (Supervisor)

Publications

• Predicting the response of forest ecosystems to climate extremes using MODIS data in the Regional Climate Model MAR during the 2018 drought in Belgium (https://hdl.handle.net/2268/312964)