Development and application of quantitative polluter-attribution relationship model for water quality variation: a case of water pollution-attribution caused by sand mining in the Poyang lake outflow channel, China.

Accurate identification of water pollution sources and quantification of their impact on water quality are essential for effective pollution control and precise management of the water environment. This article proposes a quantitative polluter-attribution relationship (QPAR) model based on recombination of samples and confidence interval estimation to evaluate contributions of sand mining activities to water pollution, illustrated through a case study of Poyang Lake. Specifically, the control variable method reorganizes the time series samples to construct submodels that quantify the impacts of three factors on water quality. Subsequently, a point estimation model for predicting the water quality variation is established based on submodels. By analyzing interval prediction results at varying confidence levels using residual distribution assumptions and kernel density estimation, the QPAR model is constructed to quantify the impact of sand mining activities on water quality variations. The results indicate that the QPAR model facilitates the analysis of the contributions of various influencing factors to the variations in total phosphorus levels in Poyang Lake, as well as the calculation of permissible sand mining volumes under different water quality scenarios. Meanwhile, the model can assist local governments in planning sand mining activities to minimize pollution impact on downstream water quality.