Simulation and prediction
Terrestrial systems and the processes taking place in them can be subdivided into different subsystems and sub-processes which all have their specific behavior but which are influencing each other strongly. Therefore, one of our aims is to develop multi-compartment multi-process simulation models that integrate different subsystems of the terrestrial systems (e.g.: geo-, bio-, hydro-, atmosphere). We use model simulations as tools to gain knowledge about the functioning of terrestrial systems and to design experiments and interpret observations so as to verify or falsify hypotheses. We use data assimilation techniques to inform simulation models with observations and provide real-time forecasts of different variables of interest in terrestrial systems, including the possibility of real-time control for water resources management. Because of the high computational cost of simulations, we make use of the Jülich supercomputing infrastructure in cooperation with the Centre of High-Performance Scientific Computing in Terrestrial Systems, Geoverbund AB C/J. Last but not least, our simulation models are used for policy advice, e.g. to predict the impact of terrestrial management options on soil and water resources.
The following methods represent our core expertise:
Simulation models:
- High resolution soil-root-plant systems: R-SWMS, CRootBox, PlaNet-Maize, DuMux
- Soil-carbon models: AGRO-C
- River basin scale water and nutrient flux models: GROWA, mGROWA, DENUZ-WEKU-MEPhos
- Integrated terrestrial systems modeling: TerrSysMP
- Data-assimilation methods: TerrSysMP-PDAF
- Seamless forecasting of hydrological states (e.g. soil water content, plant available water content groundwater table depth, soil temperature) and fluxes (e.g. precipitation, evapotranspiration, groundwater recharge, river discharge) for catchments, regions, Europe.
- High-performance scientific computing
More information can be found here: