An integrated statistical methods and modelling mineral-water interaction to identifying hydrogeochemical processes in groundwater in Southern Tunisia

Abstract

Groundwater is the most precious and valuable natural source of water in the southeast of Tunisia. The aim of this study is to find an adequate combination of methods for a qualitative description of geochemical processes into the Zeuss-Koutine, a Triassic and Miocene groundwater system, the unique source of water for this region of the Mediterranean (Medenine, Jerba, Zarzis and Jorf cities).Conventional classification techniques, statistical analyses and kriging methods were used to identify mineral processes distribution. This study finds that water chemistry is mainly dominated by dissolution/precipitation of minerals (calcite, dolomite, aragonite, anhydrite, gypsum and halite). Results obtained from principal component analyses (PCA) demonstrate that the variable responsible for water quality are largely related to soluble salts species (Na+, Cl-, Ca2+, Mg2+ and SO42- ). On the basis of the hierarchical cluster analysis (HCA), three groundwater clusters have been distinguished : cluster 1 is low TDS (<1000 mg L-1) Ca2+= Mg2+-Na+-SO4 2- waters which degrades into predominantly Na+-Ca2+-Cl--SO42- more saline groundwater (cluster 2) (TDS>3000 mg L-1) resulting from replacement of Ca by Na through cation exchange process on clay minerals. Cluster 3 is high TDS (>7000 mg L-1) NaCl waters from the aquifer system under confined conditions. Clusters 1 and 2 are located in preferential recharge zones whereas cluster 3 characterises downstream of the study area. Overall, a groundwater quality decrease has been observed similarly with a salinity increase from downstream to upstream towards the coast, where cation exchange processes and salinisation due to the long residence times appear to be the main processes responsible of more salty waters.