Applying the regression kriging method to explore the influence of agricultural activities on estimating spatial distributions of groundwater nitrate-N

Groundwater is one of the important water resources in western Taiwan and frequently used to meet water demands for irrigation, aquaculture, household, and public supply. In particular, groundwater nitrate-N pollution typically occurs in many agricultural regions owing to surface agricultural activities. Because numerous environmental factors can affect groundwater nitrate-N pollution, the delineation of extents of groundwater nitrate-N pollution is considerably critical according to auxiliary information of agricultural activities. The purpose of this study was to explore the influence of agricultural activities on estimating spatial distributions of groundwater nitrate-N by using regression kriging (RK) in the Choushui River alluvial fan, Taiwan. First, data on agricultural land use areas, such as crops, paddy fields, dry farmlands, orchards, livestock farming, and agricultural facility, were collected using geographical information system. Moreover, data on groundwater nitrate-N pollution surveyed by the Taiwan Water Resources Agency were determined according to medians of monitoring results between 2013 and 2020. Then, stepwise multiple linear regression (MLR) was used to explore the relationship between groundwater nitrate-N pollution and agricultural activities. Finally, RK was adopted to analyze the residuals between predicted nitrate-N obtained from MLR and observed nitrate-N. The study results indicated that groundwater nitrate-N pollution was positively related with orchard areas and negatively related with areas of agricultural attached facilities and livestock and poultry houses within a circle with a 1000-m radius centering a monitoring well. Moreover, RK estimates showed more spatial variability than ordinary kriging estimates for groundwater nitrate-N pollution because of orchards. To reduce groundwater nitrate-N pollution, feasible strategies of agricultural resources and environmental management are proposed based on the influence of surface agricultural activities on estimating spatial distributions of groundwater nitrate-N.