Assessment of Soil Erosion by Quantile Estimates for European Regions[Évaluation de l'érosion des sols par des estimations quantiles pour les régions européennes]

The decision to adopt one or another of the sustainable land management alternatives should not be based solely on their respective benefits in terms of climate change mitigation but also based on the performances of the productive systems used by farm holdings, assessing their environmental impacts through the cost of specific resources used. This communication uses the symbolic data analysis tools in order to analyse the conditional quantile estimates of the fertilizer costs of specific productions in agriculture, as a replacement proxy for internal soil erosion costs. After recalling the conceptual framework of the estimation of agricultural production costs, we present the empirical data model, the quantile regression approach and the interval data techniques used as symbolic data analysis tools, mainly symbolic principal component analysis and symbolic clustering of the estimation intervals. The comparative analysis of econometric results for main products between European regions illustrates the relevance of the displays obtained for inter-regional comparisons based on specific productivity.