Natural arsenic contamination of groundwater represents one of the most impacting environmental and health issues in several regions of the world, particularly in South and Southeast Asia, where millions of people may be drinking water with unsafe arsenic levels. In the densely populated region of the Red River Delta in Vietnam, water-quality data have shown the occurrence of arsenic concentrations above the WHO guideline value of 10 μg/L in both the shallow Holocene and deep Pleistocene aquifers.
Although previous studies considered the chemical conditions as a main proxy for arsenic contamination, this scenario has led to hypothesize a human-influenced contamination mechanism in deep aquifer.
It was demonstrated that the excessive groundwater pumping, due to the high water demand in intensive urban areas, can contribute to the quality deterioration of the Pleistocene aquifer triggering the compaction of interbedded clays and the subsequent expulsion of water containing dissolved arsenic or arsenic-mobilizing solutes.
To get a better understanding of the presence of arsenic in the deep aquifer and of the clay-compaction mechanism, the Weights of Evidence (WofE) Bayesian-based model was considered a reliable methodology, allowing to quantitatively evaluate the spatial correlation between arsenic pollution and factors potentially influencing the occurrence of elevated arsenic concentrations. These factors include: i) geochemical parameters (redox potential and iron concentration) in the Pleistocene aquifer, ii) arsenic concentration in the Holocene aquifer and iii) urban change patterns. The latter factor was expressed by using satellite QuikSCAT (QSCAT) data processed with the innovative Dense Sampling Method (DSM) to quantify urbanization in terms of building volume patterns and their rates of change in a decadal time span.
A direct correlation was found between arsenic contamination in the Pleistocene aquifer and: i) negative values of redox potential and ii) high iron concentration in the deep aquifer, iii) high arsenic concentration in the Holocene aquifer and iv) highly urbanized areas.
All these factors were combined to generate a predictive probability output, expressed as susceptibility map of the Pleistocene aquifer to arsenic contamination. The performance of this map was evaluated through the application of the same procedures adopted in previous groundwater vulnerability studies.
The relationship between high arsenic concentration and chemical parameters detected in the deep aquifer has proved to be consistent with the outcomes derived from the analysis performed in other works. The analysis on DSM data indicates that the excessive groundwater pumping can induce a vertical downward migration of chemical components and/or transport of arsenic contaminated waters from the shallow to the deep aquifer, determining the release of dissolved arsenic stored in deep interbedded clays into adjacent aquifers due to clay-compaction. These results provide preliminary information on the nature of deep groundwater arsenic and on the sources responsible for its mobilization.