Data Conditioning Modes for the Study of Groundwater Resource Quality Using a Large Physico-Chemical and Bacteriological Database, Occitanie Region, France

When studying large multiparametric databases with very heterogeneous parameters (microbiological, chemical, and physicochemical), covering a wide area, the probability of observing extreme values (Z-score > 2.5) is high. The information carried by these few samples monopolizes part conveyed entire database. study spatial structure data identification mechanisms responsible for water quality are then strongly degraded. Data transformation can be proposed to overcome problems. This deals database 8110 groundwater analyses (Occitanie region, France), on which bacteriological load was measured in Escherichia coli Enterococci, addition electrical conductivity, major ions, Mn, Fe, As pH. Three modes conditioning were tested compared treatment raw data. results show that log best option, revealing relationship between E. content all other parameters. By reducing impact without eliminating them, it allowed concentration first factorial axes PCA, consequently better definition associated processes. principal components their cartographic representation improved. square root function led an intermediate improvement logarithmic absence conditioning. application should allow targeted, more efficient, therefore, less expensive monitoring Regional Health Agencies.