transport equation is run, with a 10-day time step, to compute the evolution of
pesticide concentrations in each aquifer.
In this case, a regional application of the MODCOU-NEWSAM model [71] was
developed, describing most of the Champagne area and covering the Vesle basin in
its entirety (Fig. 1). The surface domain of the basin is discretised by over 5,000 cells
within a progressive multi-scale grid of embedded square meshes. The groundwater
domain is modelled by a multilayer structure of 5,900 cells. Both surface and aquifer
grid cells vary from 250 m to 2 km in size. Two aquifer layers describe the main
aquifers of the basin (Cenomanian Chalk aquifer and tertiary complex multilayered
ensemble).
Prior to any pesticide-related simulations, hydro-dynamism was calibrated using
measurements of seven hydrometric stations located on the river and four piezometers distributed along the watershed; nitrate concentration measurements from
14 boreholes were also used to calibrate solute transport.
For each cell of STICS-Pest, the spatialised model provides both water drainage
and leached fluxes of atrazine and associated metabolites. Under agricultural lands
only, leaching fluxes were diluted by the water drainage computed by STICS-Pest.
Then computed concentrations were transmitted to the aquifer system through the
unsaturated zone. Under urban and forest areas, the MODCOU water balance was
used with associated pesticide concentrations set to 0. This value is justified by there
being no treatment with atrazine in forests and by urban uses on impervious surfaces
that directly transfer to rivers.
Figure 7 shows the location of each monitoring point used for the model
calibration on the basin as well as a few sample results regarding piezometric
heads (Fig. 7a), river flow (Fig. 7b) and nitrate concentration (Fig. 7c, d) at different
stations and boreholes.
Good agreement was found for piezometric heads (Fig. 7a), river flow (Fig. 7b)
and nitrate concentration (Fig. 7c, d) at different stations and boreholes.
The changes in atrazine at a borehole are accurately reproduced by the simulations during this period of available data (Fig. 7e). Maps of simulated concentrations
for the Cenomanian aquifer layer at the end of December 2001 (deadline of
authorised sale of atrazine) and in 2013 (final time step of the simulation), shown
in Fig. 7, are also in agreement with observed concentrations.
A mass balance, at the scale of the entire area and over the simulation period
(1971–2013), shows a 0.39% ratio between total simulated leached and applied
masses of atrazine, which is close to the value previously determined, also using
STICS-Pest, by [35] on the Orgeval catchment. More generally, this order of
magnitude is similar in many studies reported in the literature [72–76].
6 Pesticides at the Seine River Basin Scale
At this stage of the modelling procedure, the results obtained on the Vesle basin
show that it is possible to simulate the transfer of atrazine in a soil-groundwater-river
system. However, considering the Seine River basin, various residence times in
How Should Agricultural Practices Be Integrated to Understand and. . .
153
pesticide concentrations in each aquifer.
In this case, a regional application of the MODCOU-NEWSAM model [71] was
developed, describing most of the Champagne area and covering the Vesle basin in
its entirety (Fig. 1). The surface domain of the basin is discretised by over 5,000 cells
within a progressive multi-scale grid of embedded square meshes. The groundwater
domain is modelled by a multilayer structure of 5,900 cells. Both surface and aquifer
grid cells vary from 250 m to 2 km in size. Two aquifer layers describe the main
aquifers of the basin (Cenomanian Chalk aquifer and tertiary complex multilayered
ensemble).
Prior to any pesticide-related simulations, hydro-dynamism was calibrated using
measurements of seven hydrometric stations located on the river and four piezometers distributed along the watershed; nitrate concentration measurements from
14 boreholes were also used to calibrate solute transport.
For each cell of STICS-Pest, the spatialised model provides both water drainage
and leached fluxes of atrazine and associated metabolites. Under agricultural lands
only, leaching fluxes were diluted by the water drainage computed by STICS-Pest.
Then computed concentrations were transmitted to the aquifer system through the
unsaturated zone. Under urban and forest areas, the MODCOU water balance was
used with associated pesticide concentrations set to 0. This value is justified by there
being no treatment with atrazine in forests and by urban uses on impervious surfaces
that directly transfer to rivers.
Figure 7 shows the location of each monitoring point used for the model
calibration on the basin as well as a few sample results regarding piezometric
heads (Fig. 7a), river flow (Fig. 7b) and nitrate concentration (Fig. 7c, d) at different
stations and boreholes.
Good agreement was found for piezometric heads (Fig. 7a), river flow (Fig. 7b)
and nitrate concentration (Fig. 7c, d) at different stations and boreholes.
The changes in atrazine at a borehole are accurately reproduced by the simulations during this period of available data (Fig. 7e). Maps of simulated concentrations
for the Cenomanian aquifer layer at the end of December 2001 (deadline of
authorised sale of atrazine) and in 2013 (final time step of the simulation), shown
in Fig. 7, are also in agreement with observed concentrations.
A mass balance, at the scale of the entire area and over the simulation period
(1971–2013), shows a 0.39% ratio between total simulated leached and applied
masses of atrazine, which is close to the value previously determined, also using
STICS-Pest, by [35] on the Orgeval catchment. More generally, this order of
magnitude is similar in many studies reported in the literature [72–76].
6 Pesticides at the Seine River Basin Scale
At this stage of the modelling procedure, the results obtained on the Vesle basin
show that it is possible to simulate the transfer of atrazine in a soil-groundwater-river
system. However, considering the Seine River basin, various residence times in
How Should Agricultural Practices Be Integrated to Understand and. . .
153