What is the impact, on groundwater and rivers in the Sèvre Niortaise and Marais poitevin catchment areas, of the extraction, of the natural environment, of 12 million cubic meters of water per year? almost half of which will be stored in the open air? According to the Société anonyme de l’eau des Deux-Sèvres, Coop 79, which carries the project of sixteen “substitution reservoirs” intended to accommodate 6.2 million cubic meters of water for farmers, these mega-basins would be beneficial for local water resources.
Coop 79 relies on a study commissioned from the Bureau of Geological and Mining Research (BRGM) according to which filling these immense infrastructures by pumping in winter would result in “significant gains” the height of the slicks – up to several meters – and “great support” river flows. Even if farmers want to continue to take an additional 5.8 million cubic meters in the hot season.
However, these reassuring conclusions of the BRGM are called into question by a counter-expertise led by the hydrogeologist Anne-Morwenn Pastier. Doctor in Earth sciences, specialized in geomorphology, the scientist dissected the infographics presented in the report. She drew questions from it which must be presented on Friday February 3, during a public meeting in Saint-Léger-de-la-Martinière (Deux-Sèvres), at the invitation of organizations opposed to the storage project. large-scale water.
“It’s not precise enough”
The numerical hydrological model – called “Jurassic” – that the BRGM has run is not relevant, according to Anne-Morwenn Pastier. It is a regional model which integrates data on rainfall, temperatures, the nature of the geological layers, their permeability, and which covers more than 20,000 square kilometers with a grid of 1 kilometer squares. “It does not make it possible to deduce local incidences, the effects on small watercourses, wetlands… Because the hydrological network behaves differently in one cell and in its neighbour, depending on the porosity of the geological layers, if it there is an underground river or not, she summarizes. It’s not precise enough. I noticed that the river Sudden, for example, is not included in the study. »
The BRGM itself notes in the “model limits” that this one can only answer “questions on a regional scale”. The data is indeed calculated for the center of each of the squares, they cannot accurately represent the entire mesh, especially when it includes “several abstraction structures, a nearby watercourse or canal”, acknowledge the authors. For that, a finer mesh would be needed.
You have 43.64% of this article left to read. The following is for subscribers only.