Comparative application of four methods of groundwater vulnerability mapping in a Slovene karst catchment

Four methods of groundwater vulnerability mapping have been applied in a Slovene karst catchment and validated by tracer tests. The test site is characterised by high water table fluctuations, manifested in intermittent lakes and variable drainage divides. A first multi-tracer test (two injections) allowed subdivision of the catchment into zones of different degrees of contribution ('inner zone' and 'outer zone'). For vulnerability mapping, only methods that consider the specific nature of karst aquifers such as heterogeneity and duality of infiltration processes, were selected: EPIK, PI, the 'Simplified Method' and the 'Slovene Approach'. For validation, a second multi-tracer test (four injections) was carried out. The time of first detection and the normalised recovery were used as validation criteria. The results suggest that EPIK and the Simplified Method sometimes overestimate vulnerability, while PI and the Slovene Approach tend to deliver more realistic results, at least during low-flow conditions. The Slovene Approach gives the clearest guidance on how to deal with hydrologic variability, for example by assigning lower vulnerability to occasionally active sinking surface waters than to permanent ones. As a conclusion, commonly accepted validation techniques are needed and should be applied by default to evaluate different vulnerability mapping methods and the resulting maps.