Contamination characteristics and source apportionment of a desert freshwater lake based on APCS-MLR and PMF models

This work investigated the evolution of water quality and pollution sources in Hongjiannao Lake, the largest freshwater lake in Chinese deserts, by analyzing the temporal variations of key water quality indicators from 2013 to 2024 and their correlations with potential influencing factors such as tourist numbers and lake surface area. Absolute Principal Component Scores–Multiple Linear Regression (APCS-MLR) and Positive Matrix Factorization (PMF) models were applied to identify and quantify the major pollution sources. The results indicated that the main pollutants in Hongjiannao Lake were the permanganate index (CODMn), fluoride (F-), total nitrogen (TN), chemical oxygen demand (COD), and volatile phenols (VP). The sharp decline in tourist numbers in 2020 due to the COVID-19 pandemic resulted in the best water quality during the study period. Concentrations of TN and ammonia nitrogen (NH3-N) were higher in the wet season compared to normal-flow season, whereas F- concentrations exhibited the opposite trend. COD exhibited a positive correlation with tourist numbers, indicating strong influence from tourism intensity. Volatile phenols and NH3-N showed a significant positive correlation, both primarily originating from industrial wastewater discharged by coal chemical enterprises. Source apportionment results revealed that both APCS-MLR and PMF models identified the same four pollution sources: tourism and domestic source, industrial source, natural source and agricultural source. The average contribution rates of the APCS-MLR model were 38.79%, 34.19%, 13.00%, and 14.02%, respectively, while those of the PMF model were 51.54%, 15.20%, 26.04%, and 7.24%. Both models indicated that COD, TN, and the permanganate index were mainly influenced by tourism and domestic source, NH3-N and volatile phenols by industrial source, sulfides (SO42-) by natural source, and total phosphorus (TP) by agricultural source. However, the PMF model suggested that NH3-N was influenced by tourism and domestic source, while F- was impacted by natural source. The combined application of APCS-MLR and PMF models enabled more reasonable and accurate identification of pollution sources. This research provides a scientific basis for pollution control in desert freshwater lakes and contributes to safeguarding the habitat of the Relict Gull, a rare and endangered bird species.