Pleistocene and Holocene groundwaters in the freshening Ledo-Paniselian aquifer in Flanders, Belgium

The Ledo-Paniselian aquifer presents a case study of evolution of fresh groundwater from sea water under the changing piezometric and climatic conditions of the Pleistocene and Holocene. Hydrogeochemical, isotopic, experimental and hydrodynamic results are used in the interpretation. The distribution of groundwater types in the Ledo-Paniselian aquifer is determined by two end members: fresh Ca-HCO3 recharge water and sea water-saturated sediments. Hydrogeochemical modelling supports the view that mixing of the end members and cation exchange are the main processes; calcite dissolution is also important. Cation exchange consists, in the first place, of desorption of the adsorbed marine cations (Na+, K+ and Mg2+) in exchange for the freshwater cation Ca2+. Groundwater δO is around the value of modern precipitation in the area (−6.5‰) for the samples with higher radiocarbon contents; it is <−7.0‰ for the groundwater containing the lowest radiocarbon levels. An overlapping transition zone exists between both groups. δ13C becomes heavier for the samples containing the lowest radiocarbon levels, indicating chemical dilution. Pore waters from the Bartonian clay show preferential flow paths. Faster flow paths are more strongly leached, leading to low total dissolved solids (TDS), low sulphate concentrations and low Mg2/Ca2 ratios; the slower pathways still contain gypsum, increasing the sulphate concentrations and TDS, and Mg2/Ca2 ratios are higher because they were less reduced by cation exchange resulting from freshening. Four methods for determining cation exchange capacity (CEC) and adsorbed cations are compared: the NH4OAc method, two BaCl2 methods (one in unbuffered and the other in buffered conditions) and a new NaCl/NH4Cl method. Reasonable CEC values are obtained with the NHOAc method. Comparing the measured equivalent fractions of the adsorbed cations with those calculated from the pore solutions, using the computer programme PHREEQC, it can be concluded that the NaCl/NH4Cl method produces the best results. The proton exchange capacity of decalcified sand from the Ledo-Paniselian aquifer was determined to be c. 1–1.5 meq/100 g in the pH range 5–8.5. A hydrodynamic model is developed to explain the evolution of groundwater and for evaluating the effects of pumping at both local and regional scales. Model calculations show that the observed freshwater-saltwater distribution is not the result of the present freshwater flow conditions but the result of different flow regimes during the ice ages when sea levels were much lower. Occurrence of a permafrost layer during cold periods could have had a dramatic impact on the groundwater flow system by, at least temporarily, decreasing the recharge of the aquifers. The existence of the Saalian ice sheet in The Netherlands could have influenced the flow in the deeper Eocene-Oligocene aquifers. The high pressures that existed under the ice sheet could have reversed the flow direction from north to south