Modelling the groundwater nitrate legacy

This report details the findings of a project jointly funded by the British Geological Survey (BGS)and Defra through the Environment Agency. The overall aim of the work was to investigate the use of new models to inform decision-making on nitrate pollution in groundwater and the potential for incorporating unsaturated zone processes into the model currently used by the Environment Agency to delineate Nitrate Vulnerable Zones (NVZs). The potential application as supporting evidence for the Water Framework Directive has also been considered as nitrate pollution of groundwater remains the most significant reason for failure of WFD environmental objectives across England. The background to the nitrate legacy in groundwater and to the approaches to NVZ designation is described in Stuart et al. (2016). A series of developments to the BGS Nitrate Time Bomb (NTB) model have been made to improve a number of areas and approaches used in the first version of the model. The improvements included a spatially and temporally distributed nitrate input function, improved unsaturated zone thickness estimation, travel time attribution using a 1:250,000 geological map, estimating nitrate velocity in the unsaturated zone using groundwater recharge and aquifer properties, and introducing nitrate transport processes in low permeability superficial deposits and the saturated zones. These now allow the model to be applied at sub national scale. Using the improved model we have also made the first estimate of the mass of nitrate stored within the unsaturated zone and how this will change over time to improve UK nitrate budget estimates. The new version of the BGS NTB approach was applied in three case studies at different scales which compared its outcomes to the results from other modelling to demonstrate that the model can be benchmarked against the other nitrate modelling approaches: • For a basin-scale model of the Thames Chalk (Howden et al., 2010 & 2011). The NTB model gave comparable results to the original study back to 1925 provided that the same nitrate input function was used. Both models failed to predict nitrate concentrations in the Thames after the mid-1980s. • At the multi-borehole scale in the Permo-Triassic. A similar approach was used to the BGS model in the Eden Valley. This replicated the existing model for the area used by the Environment Agency both in terms of trend assessment and in the lack of dilution available within the aquifer block for blending purposes. • At the single borehole scale in the Chalk of the South Downs. The existing Environment and National Park model constructed by AMEC treated the unsaturated zone very similarly to the NTB model. This model provided a good fit to observed concentrations and confirmed the importance of estimating unsaturated zone delays. The assessment of modelled travel time from different areas of the catchment clearly illustrated the arable areas that would give a relatively rapid respond to changes in nitrate management. To illustrate the potential application of the BGS NTB model to support the Environment Agency’s NVZ designation methodology, areas of England were identified where unsaturated zone lags may be significant and where there is uncertainty in the NVZ designation. A major advantage of the BGS NTB model is that it covers the whole of England (and Wales) in a consistent way. A national overview of areas of designation uncertainty identified large areas of England, in particular the chalk outcrop of southern and eastern England. These were compared to areas with significant unsaturated zone travel time indicating where travel time may be contributing to designation uncertainty. The results suggest that the model may be useful both for identifying currently impacted groundwater which reflects legacy fertilizer application and also where additional designation could be needed as impacts have not yet emerged. Application of the model to support implementation of the WFD has also been considered and whilst no quantitative analysis has yet been carried out there are a number of ways that the model could be of significant benefit. For example, the model could be used to estimate when trend reversal would be expected to occur as a result of measures (at a specific location or across a groundwater body) and the time required to achieve good chemical status (alternative objective setting). A further application could be for scenario testing such as evaluating the effects of different land use/management measures as part of cost benefit analysis or considering the long term impacts of climate change through changing fertiliser use and/or recharge