Capability and limitations of a simple grid-based model for simulating land use influences on stream nitrate concentrations

Catchment scale nitrogen models are valuable tools for exploring how stream water quality is likely to be influenced by future changes in climate and land use. Many nitrogen modelling systems are highly complex and include too many calibration parameters to be confident in how these would be affected by changes in the catchment structure or meteorological drivers. This paper examines how successfully a simplified model of nitrogen budgets, linked to a hydrological model, can be used to represent the variability in stream nitrate responses resulting from different land uses. The model has been tested in a small (71 km2) transitional agricultural catchment in north east Scotland using a suite of time-series data from 15 different sites across the catchment in addition to the key mainstream monitoring points. Results from the model application indicate that the model can successfully simulate the difference between the key land use types of arable crops, improved grassland, rough grassland and forestry, at spatial scales of 6 km2 or greater. Larger errors in simulation of stream nitrate for some smaller sub-catchments are believed to be linked to the limitations of the model in representing some aspects of agricultural management, especially related to livestock inputs to the system.