An integrated approach combining LISA, BI-LISA, and the modified COPK method to improve groundwater management in large-scale karst areas

Abstract

Groundwater protection and sustainable utilization of wells in large-scale karst areas face many challenges due to the complexity of numerous groundwater basins and aquifers. Therefore, scientific evidence is needed for groundwater management in these areas. A method combining source vulnerability, spatial autocorrelation, and bivariate spatial autocorrelation is proposed for groundwater management research. Based on the COP method (composed of flow Concentration, Overlying layers, and Precipitation regime), the COPK method is modified by replacing the sinkhole distance with sinkhole density, and incorporating land use and karst network development. The modified COPK method is used to evaluate the vulnerability of nitrate in groundwater in large-scale karst areas. Guizhou Province in China, which is a typical karst region, was evaluated as a study area. Comparatively to the classical COP method, the revised vulnerability map shows more areas of high vulnerability, mainly in the middle of the study area. According to the receiver operating characteristic curve (ROC) analysis, the modified method had a higher accuracy than the classical COP algorithm. Based on the accurate vulnerability index, vulnerability was classified using the natural break method, and the spatial autocorrelation of vulnerability clustering characteristics was studied. The source vulnerability was divided into High-High (HH), High-Low (HL), Low-High (LH), Low-Low (LL), and Not Significant (NS) clusters. This result can clarify the vulnerability clustering of groundwater in specific areas, thereby helping to improve groundwater management accuracy and efficiency. In addition, bivariate spatial autocorrelation was used to study the nitrate content of wells and source vulnerability. The bivariate local Moran's index showed a spatial negative correlation between the two. Similarly, the BI-LISA map can be used to characterize the spatial aggregation of nitrate content and well vulnerability. This result can help groundwater managers develop reasonable measures to protect water wells.