Predictions of physical, chemical, and biological processes in soils, aquifers, rivers and across compartments are strongly affected by uncertainties and errors in model structure, parameters and forcing data. Thus, finding new model formulations (data-driven, physics-based, knowledge-guided or hybrid), estimating parameter sets, quantifying uncertainty, correcting for errors and evaluating competing models are crucial for reliable predictions at any scale (lab/field/catchment). We invite contributions on improved concepts, approaches and computational algorithms in these areas, especially on:
- new methods for inference of model equations or their parameters, inverse modelling and data assimilation in spatially distributed systems,
- data worth and optimal experimental design strategies that maximize information from (computational) experiments and minimize uncertainty,
- novel theories and concepts for spatial and temporal analysis of model - data mismatches,
- formal and informal frameworks that diagnose, detect and resolve modelling errors and capture conceptual model uncertainty,
- approaches how to bring (and extract) sound scientific reasoning into machine learning, be it for inferring new models, model error correction or data analysis,
- constraint learning techniques and novel likelihood formulations as methods to incorporate expert knowledge and soft information into the model/parameter inference process,
- new measurement technologies (point-scale to remote sensing and “soft information”) that aid developing, learning or parameterizing models, or can help diagnose and detect structural model deficiencies, as well applications and benchmarking efforts in any of these fields.
- Bayesian approaches that address any of the topics above