Exploitable results
Within the framework of this research project, two models MIKE SHE/DAISY and SMART2 have been used to predict the risk of nitrates and aluminium contamination of groundwater linked to agricultural activities and/or atmospheric deposition. Readily available data from standard European data bases have been used as the basis for modelling. Due to the need for providing model outputs at a large scale, with the idea of policy making at European scale, and the shift from experimental data to geo-referenced databases as model inputs, models have been adapted to a larger scale than usually implemented.
With regard to simulation of nitrogen leaching to aquifers in relation to agricultural activities on a large scale, using the MIKE SHE/DAISY model, it appears that the magnitude of uncertainty is related to the type of simulation results in terms of temporal and spatial scales. Large uncertainties have been detected on flux concentrations leaving the root zone at grid level whereas uncertainties in simulated concentrations at aquifer level on catchment scale were much smaller. In relation to the consequences of atmospheric deposition of acidifying substances on concentration of aluminium in groundwater, large prediction intervals due to uncertainty in input data were noted. However, despite this large uncertainty, SMART2 was able to predict a notable decrease in aluminium concentration when testing a reducing deposition scenario on a large scale over a long time period. It was also highlighted that the width of the prediction interval highly depends on whether block median concentrations or block areal exceedances are considered.
Turning to the determination of the different error sources, continuous soil parameters contribute more to the uncertainty in aluminium concentration, whereas soil and vegetation maps are the main contributors to uncertainty in nitrates concentration caused by atmospheric deposition of acidifying substances.
From the groundwater management perspective, economic risk analyses have demonstrated that uncertainties of the outputs of hydrological models run at European scale may result in large prediction intervals for the societal economic value of measures to reduce groundwater degradation. Scenarios for contamination reduction, tested in the framework of the project, have resulted in prediction intervals that vary between 6% and 63% of the expected economic outcome of the scenarios. It should be stressed that the economic value assigned to the groundwater resource plays a major role in the uncertainty of the prediction intervals.
Turning to strategic implications following from the technical and economic findings of the project, emphasis has been put on data quality and availability of European geo-referenced databases, on the development of relevant geo-referenced databases related to soil characteristics, on the use of subjective probability when expert knowledge is requested as data source, and finally on the need to take cumulative uncertainty into account when performing risk analysis.
More information is available at http://projects.gim.lu.uncersdss
Buscando datos de OpenAIRE...
Se ha producido un error en la búsqueda de datos de OpenAIRE
No hay resultados disponibles