KASIMIR – Effective and Realistic KArst Hydrological SIMulation and InfeRence

• Applicants: Professor Dr. Andreas Hartmann; Dr.-Ing. Thomas Reimann
• Subject Area: Hydrogeology, Hydrology, Limnology, Urban Water Management, Water Chemistry, Integrated Water Resources Management
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 560588049

Providing water to around 25% of the world’s population, karst aquifers are important natural water resources. However, the realistic and process-based hydrological modelling of karst system spring discharge still poses a challenge due to system complexity and data scarcity. Therefore, lumped models (LMs) are commonly applied in practice, which represent the system in a spatially aggregated manner using simplified process representations. LM parameters, however, might not be directly physically interpretable or measurable and inferences made about karst system properties with LMs are often compromised, among other aspects, by parameter uncertainty and an unknown degree of model realism. Due to the global importance of LMs for karst hydrological modelling, there is a need to tackle and resolve these problems to enable more efficient and realistic karst hydrological modelling in the future. We hypothesize that, given methods of uncertainty quantification, sensitivity analysis, and model evaluation, it is possible to identify the most realistic LM for a given karst system, unveiling system characteristics and enabling accurate simulations. We will focus on establishing relationships between LM parameters and physical system properties for a large number of synthetic karst systems, generated using recently developed stochastic karst conduit network generators in combination with a state-of-the-art spatially distributed process-based model (PBM). We will simulate corresponding synthetic spring discharge time series, which serve as a known ground truth. Ultimately, this allows for the discovery of patterns in model realism, process representation, and parameter relationships for a large number of karst systems in an unprecedented and generalized manner.

• DFG Programme: Research Grants