Are maps of nitrate reduction in groundwater altered by climate and land use changes?

Abstract

Abstract. Nitrate reduction maps have been used routinely in northern Europe for calculating the efficiency of remediation measures and the impact of climate change on nitrate leaching. These maps are, therefore, valuable tools for policy analysis and mitigation targeting. Nitrate reduction maps are normally based on output from complex hydrological models and, once generated, are largely assumed constant in time. However, the distribution, magnitude, and efficiency of nitrate reduction cannot necessarily be considered stationary during changing climate and land use as flow paths, nitrate release timing, and their interaction may shift. This study investigates the potential improvement of using transient nitrate reduction maps, compared to a constant nitrate reduction map that is assumed during land use and climate change, both for nitrate loads and the spatial variation in reduction. For this purpose, a crop and soil model (DAISY) was set up to provide nitrate input to a distributed hydrological model (MIKE SHE) for an agricultural catchment in Funen, Denmark. Nitrate reduction maps based on an observed dataset of land use and climate were generated and compared to nitrate reduction maps generated for all combinations of four potential land use change scenarios and four future climate model projections. Nitrate reduction maps were found to be more sensitive to changes in climate, leading to a reduction map change of up to 10 %, while land use changes effects were minor. The study, however, also showed that the reduction maps are products of a range of complex interactions between water fluxes, nitrate use, and timing. What is also important to note is that the choices made for future scenarios, model setup, and assumptions may affect the resulting span in the reduction capability. To account for this uncertainty, multiple approaches, assumptions, and models could be applied for the same area. However, as these models are very time consuming, this is not always a feasible approach in practice. An uncertainty of the order of 10 % on the reduction map may have major impacts on practical water management. It is, therefore, important to acknowledge if such errors are deemed acceptable in relation to the purpose and context of specific water management situations.

A modified SWAT model for mechanistic simulation of soil water‐salt transport and the interactions with shallow groundwater

“…Note that θ‐formed Richards' equation is only restricted to unsaturated flow since the hydraulic gradient cannot be correctly represented in the form of soil moisture in saturated soils (Crevoisier et al, 2009; Zha et al, 2017). In addition, the more sophisticated MIKE SHE model can simulate the VZ processes by using pressure head‐formed (h‐formed) Richards' equation but explicitly solving advection‐dispersion equation (ADE); it considers the interaction between VZ and saturated aquifer but does not simulate the plant growth/development and salinity effects (Seidenfaden et al, 2022). Thus, it is still necessary to further enhance the capability of watershed models in accurately simulating the coupling processes of soil water‐salt transport, groundwater fluctuations, and crop growth under arid shallow groundwater environments.…”