Developing a straightforward precipitation-runoff model for monthly-based nutrient emission modeling in river systems

Nutrient emission modeling in river basins includes the estimation of nutrient fluxes and retention in streams and forms a vital part for the management of water resources as well as the exploration of ecological impacts of increased nutrient input in river systems. The nutrient emission model MONERIS (Venohr et al., 2011, Lemm et al., 2021) calculates landuse-specific nutrient fluxes for entire river basins on a monthly basis and a spatial resolution of 1km x 1km and requires an ensemble of input data such as land use, atmospheric deposition, tile drainage cover, connection to sewer systems, and many more. Hydrological flows majorly drive nutrient fluxes and emission pathway composition. Consequently, runoff is one of the key constituents that needs to fit to the emission model requirements in terms of represented pathways and environmental compartments (e.g. land-use types, groundwater-surface water boundaries, spatial-temporal resolution). Using available runoff data derived by third party models can introduce large uncertainties to the resulting nutrient fluxes and water quality. Our aim was to develop a novel runoff model that operates on a monthly basis, provides runoff components for all considered emission pathways and can be applied with commonly available input data. These include, beyond the typical hydrological components (snow storage/melt, surface runoff, natural/artificial interflow, lower interflow and groundwater), flow estimates from urban areas (separate sewers, combined sewer overflows, decentralized/large treatment plants, point sources). The overall goal is to set up a data base for a Europe wide water quantity and quality model. In the presented pilot study, the runoff model was applied to the Odra River Basin (119,000 km²), calibrated against observed runoff data from 11 independent upstream gauges, and validated by runoff data from 36 additional gauges for the years 2010-2020. We compared input data sensitivity and model performance of three available daily gridded precipitation and air temperature datasets (E-OBS, EURADKLIM and CHELSA). Results showed a good model accuracy (NSE: > 0.9, PBIAS: < 7 %) and suggest that, despite its simplicity, the runoff model complements the nutrient emission model MONERIS. The next step will be the modelling of water quantity and quality for Central Europe, and, ultimately, providing an open modelling platform that allows emission modelling of other parameters and substances (e.g. salinity, heavy metals or priority substances) or be extended by additional modules for phytoplankton growth or floodplain retention.