EVADIFF : a 3-year project in close collaboration with Jozef Stefan Institute (Slovenia)

An innovative methodological framework applied in water protection issues led to a new DSS with a web-based interface.

The EVADIFF project was launched in 2012 to assess existing models and develop a Decision Support System to help reduce diffuse pollution from phytochemicals.

The project involved several research teams from IRSTEA (Lyon, and Antony), INRA and ARVALIS. Of the various complementary topics taken over the project, the collaboration with the Department of Knowledge Technologies at Joseph Stefan Institute (Slovenia) has led to a PhD to be defended on 25th November 2016 by Vladimir Kuzmanovski (Integrating decision support and data mining for risk evaluation and management : a framework and a case study in agriculture), and a new DSS on the risk management of phytochemicals transfer towards water resources. This DSS includes expertise of ARVALIS from databases on cropping practices and water quality as well as outputs from Aquaplaine® and Aquavallée® diagnosis tools. Decision rules issues from those tools have been formalized in DEXi® model in connection with ARVALIS soil databases, climatic stations, and the water balance model of CHN®. The rules determine the periods of water saturation in the soil layers.

The DSS aims to identify water flow modes and periods of water leakage according to soil types. A risk of phytochemical transfer is assessed depending on the date of sowing and the date of active substance application as well as for various water flows : rapid infiltration, Hortonian runoff, or by saturation, drainage network, and the specific water system of clay soils.

The decision support system also delivers a set of agricultural practices to be performed in order to avoid or mitigate surface and ground water pollution.