Evaluation of a spatialized agronomic model in predicting yield and N leaching at the scale of the Seine-Normandie Basin

The EU directive has addressed ambitious targets concerning the quality of water bodies. Predicting water quality as affected by land use and management requires using dynamic agro-hydrogeological models. In this study, an agronomic model (STICS) and a hydrogeological model (MODCOU) have been associated in order to simulate nitrogen fluxes in the Seine-Normandie Basin, which is affected by nitrate pollution of groundwater due to intensive farming systems. This modeling platform was used to predict and understand the spatial and temporal evolution of water quality over the 1971-2013 period. A quality assurance protocol (Refsgaard et al. Environ Model Softw 20: 1201-1215, 2005) was used to qualify the reliability of STICS outputs. Four iterative runs of the model were carried out with improved parameterization of soils and crop management without any change in the model. Improving model inputs changed much more the spatial distribution of simulated N losses than their mean values. STICS slightly underestimated the crop yields compared to the observed values at the administrative district scale. The platform also slightly underestimated the nitrate concentration at the outlet level with a mean difference ranging from -1.4 to -9.2 mg NO₃ L^-1 according to the aquifer during the last decade. This outcome should help the stakeholders in decision-making to prevent nitrate pollution and provide new specifications for STICS development.

Long term prospective of the Seine River system: confronting climatic and direct anthropogenic changes

To explore the evolution of a human impacted river, the Seine (France), over the 21st century, three driving factors were examined: climate, agriculture, and point source inputs of domestic and industrial origin. Three future scenarios were constructed, by modification of a baseline representative of recent conditions. A climate change scenario, based on simulations by a general circulation model driven by the SRES-A2 scenario of radiative forcing, accounts for an average warming of +3.3 degrees C over the watershed and marked winter increase and summer decrease in precipitation. To illustrate a possible reduction in nitrate pollution from agricultural origin, a scenario of good agricultural practices was considered, introducing catch crops and a 20% decrease in nitrogen fertilisation. Future point source pollution was estimated following the assumptions embedded in scenario SRES-A2 regarding demographic, economic and technologic changes, leading to reductions of 30 to 75% compared to 2000, depending on the pollutants. Four models, addressing separate components of the river system (agronomical model, hydrogeological model, land surface model and water quality model), were used to analyse the relative impact of these scenarios on water quality, in light of their impact on hydrology and crop production. The first-order driving factor of water quality over the 21st century is the projected reduction of point source pollution, inducing a noticeable decrease in eutrophication and oxygen deficits downstream from Paris. The impact of climate change on these terms is driven by the warming of the water column. It enhances algal growth in spring and the loss factors responsible for phytoplankton mortality in late summer (grazers and viruses). In contrast, increased seasonal contrasts in river discharge have a negligible impact on river water quality, as do the changes in riverine nitrate concentration, which never gets limiting. The latter changes have a similar magnitude under the three scenarios. Under climate change, riverine and groundwater nitrate concentrations increase and crop production is advantaged with reduced growing cycles and increased yields. In contrast, nitrate concentrations decrease under the good agricultural practices scenario, with a limited decrease in crop production. When these two scenarios are combined, the changes in nitrate concentrations balance each other and crop yields increase. The results of this numerical exercise indicate that the potential changes to the Seine River system during the 21st century will not lead to severely degraded water quality.

Agriculture and groundwater nitrate contamination in the Seine basin. The STICS-MODCOU modelling chain

A software package is presented here to predict the fate of nitrogen fertilizers and the transport of nitrate from the rooting zone of agricultural areas to surface water and groundwater in the Seine basin, taking into account the long residence times of water and nitrate in the unsaturated and aquifer systems. Information on pedological characteristics, land use and farming practices is used to determine the spatial units to be considered. These data are converted into input data for the crop model STICS which simulates the water and nitrogen balances in the soil-plant system with a daily time-step. A spatial application of STICS has been derived at the catchment scale which computes the water and nitrate fluxes at the bottom of the rooting zone. These fluxes are integrated into a surface and groundwater coupled model MODCOU which calculates the daily water balance in the hydrological system, the flow in the rivers and the piezometric variations in the aquifers, using standard climatic data (rainfall, PET). The transport of nitrate and the evolution of nitrate contamination in groundwater and to rivers is computed by the model NEWSAM. This modelling chain is a valuable tool to predict the evolution of crop productivity and nitrate contamination according to various scenarios modifying farming practices and/or climatic changes. Data for the period 1970-2000 are used to simulate the past evolution of nitrogen contamination. The method has been validated using available data bases of nitrate concentrations in the three main aquifers of the Paris basin (Oligocene, Eocene and chalk). The approach has then been used to predict the future evolution of nitrogen contamination up to 2015. A statistical approach allowed estimating the probability of transgression of different concentration thresholds in various areas in the basin. The model is also used to evaluate the cost of the damage resulting of the treatment of drinking water at the scale of a groundwater management unit in the Seine river basin.