Del tardofranquismo a la transición (1970-1975)

Twenty years after the first manure acts were implemented in the Netherlands, the improvement of groundwater quality has been demonstrated. This achievement was possible thanks to accurate groundwater dating allowing the existing groundwater quality time-series to be interpreted in a new way. The question is whether these time series are still necessary, if groundwater ages are available. This approach of trend detection relies on groundwater data to be related to the time of recharge. The range of groundwater ages at the screens of the monitoring network provides measurements of pollutant concentrations that represent the entire period of agricultural history. With this approach, existing poor quality time-series of groundwater quality data may no longer be required to show trends, but rather provide sufficient data to overcome the natural variations that obscure the anthropogenic trends. Instead of using time-series, trend detection is now feasible on a limited number of high quality measurements, combined with groundwater ages. This approach is most useful for the detection of existing trends in groundwater quality and to appoint areas at risk of failing to reach good chemical status. To demonstrate trend reversal in response to new measures or action programs, groundwater quality monitoring is still needed; preferable in shallow screens where young groundwater is found.

Present day concentrations of nitrate in groundwater in Noord-Brabant are still above the EU Quality Standard of 50 mg NO3 per liter. The downward trends demonstrated in this thesis decrease the concentrations of nitrate towards the Quality Standard. These trends, linked to fertilizing practices through regression analysis, are particularly useful to predict the moment of compliance with the threshold values by extrapolation. The next challenge is to demonstrate

that good chemical status will be achieved by 2015. This also requires continued monitoring the shallow screens. If present day limitations on the application of fertilizers appear to be insufficient, stricter rules must be enforced. Legislation is often evaluated within a few years after its implementation, but the question is when it will be possible to reliably detect the effect of such measures. In principle, the ages of groundwater in the shallow screens represent the shortest response time of the monitoring network to detect changes water quality resulting from changing agricultural practices.

Groundwater ages provide a direct link between groundwater quality and past land use changes. This direct link between the input of pollutants and the response of groundwater quality forms a simple empirical model for predicting future groundwater quality. In contrast to simple linear regression through groundwater quality time-series, this approach is capable of predicting future trends – as well as trend reversal – in groundwater quality due to land use changes. Directly linking groundwater quality and historical land use changes implicitly assumes that the leaching processes remain the same. The question is whether changing agricultural practices, such as the direct injection of manure into the soil, have an effect on the net leaching of pollutants to groundwater.

Reactive pollutants and transport modeling

Trends in reactive pollutants in relation to groundwater age are better understood with the aid of a stream-tube reactive transport model. In this study the stream tube model was only used for the analysis of trends in reactive solutes, and intentionally not calibrated. However, such a calibration is in principle possible with the available data. Additional measurements of different geochemical variables will provide even more insight in the geochemical reactions. For example, the complete composition of all dissolved gases, including N2 produced by denitrification, would be extremely helpful. Isotope ratios of NO3-N and N2-N have been used to distinguish between atmospheric and nitrate-derived N2 (e.g. Wassenaar et al. 2006; e.g. Singleton et al. 2007), while the different sources of SO4 (i.e. pollution and pyrite) have been distinguished by S isotopes (e.g. Moncaster et al. 1992; e.g. Schwientek et al. 2008). Additionally, O-isotopes of NO3 and SO4 provide even more insight into the sources of these solutes.

The non-linear behavior of reactive pollutants requires the use of reactive transport models for the extrapolation of trends in groundwater quality. Calibrating a reactive stream tube model against a large data set of concentration measurements makes it possible to estimate an effective geochemical parameterization of subsurface and build a representative 1D stream tube model. The dimension of this effective parameterization of the subsurface (1D) would match the dimension of point data available to calibrate the model. In contrast, the parameterization of a 3D reactive transport model strongly relies on generalizations of the subsurface and its properties. A different stream tube model is needed for each land use type and hydrological setting. Each of the stream tube models requires a separate data set for calibration.

The contribution of reactive solutes in groundwater to the surface water quality depends on the concentrations of reactive solutes along the groundwater flow path, and the distribution of groundwater travel times of discharging groundwater. The profiles of concentrations of pollutants along representative stream tubes in relation to travel time combined with particle tracking travel time distributions provide quick and reliable predictions of surface water quality for base flow conditions. Because of the computational demand of fully reactive 3D flow and transport models, such reactive stream tube models are very practical.