Distribution of iodine concentration in drinking water in China mainland and influence factors of its variation

Objective: To identify the current spatial distribution of iodine concentration in drinking water (dWIC) at the township-level across China and its influencing factors through visualization and spatial statistical analysis by the geographic information system. Methods: The dWIC for each township was used to describe the distribution by ArcGIS 10.7. The spatial aggregation characteristicswere analyzed by spatial auto-correlation analysis. The inverse distanceweightmethodwas used to predict the dWIC at nonsampling sites. The correlation between the dWIC and the distance from each township to the Yellow River as well as the depth of tube wells were analyzed by ordinary least squares and geographically weighted regression, respectively. Results: A total of 37,541 townships were included in this study. dWIC ranged from 0 to 1113.7 mu g/L, and the median was 3.3 mu g/L. There were 35,606 townships < 40 mu g/L (94.85 % of surveyed townships), 40 mu g/L <= 1015 townships <= 100 mu g/L (2.70 % of surveyed townships), and 920 townships > 100 mu g/L (2.45 % of surveyed townships). The results were statistically significant of global autocorrelation analysis (Moran's I = 0.43, Z = 922.15, P < 0.01). Local Moran's I showed that 3128 townships (8.33% of surveyed townships) belong to H-H cluster areas. The dWIC was partially negatively correlated with the distance from each township to the Yellow River, as well as positively correlated with the depth of tube wells in partial areas. Conclusions: The dWIC varied widely acrossmainland China (from 0 mu g/L to 1113.7 mu g/L). 94.85% of surveyed townships were below 40 mu g/L and 2.45 % of surveyed townships were exceeding 100 mu g/L. Moreover, the distance from each township to the Yellow River may be one of the geneses of iodine-excess areas. Finally, this study has provided a visible reference of dWIC for the precise control strategy and focused monitoring in China.