Dynamic factor analysis for estimating ground water arsenic trends

Drinking ground water containing high arsenic (As) concentrations has been associated with blackfoot disease and the occurrence of cancer along the southwestern coast of Taiwan. As a result, 28 ground water observation wells were installed to monitor the ground water quality in this area. Dynamic factor analysis (DFA) is used to identify common trends that represent unexplained variability in ground water As concentrations of decommissioned wells and to investigate whether explanatory variables (total organic carbon [TOC], As, alkalinity, ground water elevation, and rainfall) affect the temporal variation in ground water As concentration. The results of the DFA show that rainfall dilutes As concentration in areas under aquacultural and agricultural use. Different combinations of geochemical variables (As, alkalinity, and TOC) of nearby monitoring wells affected the As concentrations of the most decommissioned wells. Model performance was acceptable for 11 wells (coefficient of efficiency >0.50), which represents 52% (11/21) of the decommissioned wells. Based on DFA results, we infer that surface water recharge may be effective for diluting the As concentration, especially in the areas that are relatively far from the coastline. We demonstrate that DFA can effectively identify the important factors and common effects representing unexplained variability common to decommissioned wells on As variation in ground water and extrapolate information from existing monitoring wells to the nearby decommissioned wells.