EcoExtreML

Climate extremes like droughts and heatwaves impact how water, energy, and carbon move through ecosystems. To capture the exchange between the land surface and the atmosphere, we aim to apply the “STEMMUS-SCOPE” model across ecosystems at a global scale. The model STEMMUS simulates soil water and heat, and the model SCOPE the vegetation photosynthesis and land surface energy balance.

Like other land surface models, STEMMUS-SCOPE relies on a large number of numerical calculations using various meteorological data and surface parameters as inputs. To overcome challenges such as computational costs and optimization issues in calibration, we developed open-source tools for efficient computing and data handling following FAIR guidelines within the context of EcoExtreML project. In this project, we improved the representation of soil-water-plant-energy interaction in the model, developed a machine learning approach to emulate the model, and enhanced the re-usability of the model using the Basic Model Interface (BMI).

The technologies developed in EcoExtreML contributed to developing a digital twin of soil-plant systems and the results were used in other projects like the GEWEX PLUMBER2 Land Surface Model Benchmarking Evaluation Project, and NWO-KIC WUNDER project.