Water Quality Assessment and Apportionment of Pollution Sources of Xilin River Using Multivariate Statistics, APCS-MLR and PMF Receptor Model
Pengfei Yu and Minquan Feng* Author for corresponding; e-mail address: mqfeng@xaut.edu.cn
Volume: Vol.52 No.6 (November 2025)
Research Article
DOI: https://doi.org/10.12982/CMJS.2025.093
Received: 12 November 2024, Revised: 28 September 2025, Accepted: 20 October 2025, Published: 12 November 2025
Citation: Yu P. and Feng M., Water quality assessment and apportionment of pollution sources of Xilin river using multivariate statistics, APCS-MLR and PMF receptor model. Chiang Mai Journal of Science, 2025; 52(6): e2025093. DOI 10.12982/CMJS.2025.093.
Graphical Abstract
Abstract
This study aims to investigate water quality and pollution source identification in the Xilin River Basin using Environmental Science and Engineering terminology. The study employs Water Quality Index (WQI), Principal Component Analysis (PCA), Factor Analysis (FA), Absolute Principal Component Scores Multiple Linear Regression (APCS-MLR), and Positive Matrix Factorization (PMF). WQI results indicate values of 96.55, 138.36, and 115.47 during normal season (NS), wet season (WS), and dry season (DS) respectively, with NS showing better water quality compared to WS and DS. PCA and FA identified four key factors explaining 80.08% to 83.55% of the total variance. APCS-MLR modeling results indicate that agricultural and livestock farming pollution sources (ALS), as well as industrial wastewater and urban domestic pollution sources (IUS), are the primary contributors to river water pollution. PMF simulations reveal slight variations in pollution sources across each season. Comparing the R2 values of APCS-MLR and PMF simulations, with averages of 0.80 and 0.67 respectively, indicates that APCS-MLR demonstrates higher stability and better simulation results.