Automated catchment scale modelling of hydrological phenomena and water quality

An automated approach to hydrologic and water quality modeling on a catchment scale is presented, one that automatically produces suitable models from domain modeling knowledge and measured data. An essential component of the methodology is the domain knowledge library, comprising knowledge on hydrological and nutrient loading processes and containing alternative formulations for some of them. The library is written in a formalism compatible with the equation discovery tool ProBMoT. Given a user specification of a modeling task, ProBMoT searches the space of alternative candidate models encoded in the library. The generated models are optimized against the provided measured data, and the best-fitted model is proposed as the most suitable for modeling the observed system. The methodology was applied to the Quarteira River catchment (Algarve, Portugal). Analysis of the average annual water balance, sediment yields, and nutrient loadings in the Quarteira River support the findings of the previous research work, while the ProBMoT results are comparable or better than the results obtained with the SWAT model.