Effectiveness of different spatial interpolators in estimating heavy metal contamination in shallow groundwater: a case study of arsenic contamination in Hanoi, Vietnam

Abstract

Starting from emerging issues of arsenic contamination of groundwater, this study was conducted in Thanh Tri, a sub-urban district in southeastern side of Hanoi, in order to: (1) archive comprehensive understandings of arsenic contamination in shallow groundwater, (2) generate risk map and compare effectiveness of different spatial interpolation approaches including Kriging, IDW and Radial Basis Function of Geographic Information System (GIS) in estimating arsenic concentration.
The study found that the concentration of arsenic varied from 4.71μg/L to 394.93μg/L, with an average of 81.92μg/L, far higher than the standard of 10μg/L set by the World Health Organization. Distinct distributions of arsenic concentrations were observed for different geographical and geohydrological aspects. High arsenic concentrations were found mostly in the southwest of the study area and concentrations decreased gradually in other directions. Regarding geohydrological distribution, the upper aquifer (Holocene) yielded levels of arsenic ranging from 4.71μg/L to 316.67μg/L; while concentrations in the lower aquifer (Pleistocene) varied between 5.43μg/L and 394.93μg/L. The lowest arsenic concentration was found in the upper aquifer and the highest concentration was found in lower aquifer.
IDW gave the best prediction with an average error (MAPE) of 32.82%. An arsenic risk map was created from the IDW method. The map showed groundwater arsenic contamination and high contaminated areas in up to 99.8% of the district. The findings from this study will be valuable for households and policymakers to initiate early mitigation efforts and protect the population from chronic arsenic poisoning