Multivariate Assessment of Surface Water Quality and Pollution Dynamics Using Physicochemical and Nutrient Indicators

Abstract

Surface-water systems integrate natural geochemical processes, seasonal hydrology, catchment inputs, and anthropogenic pressures, making multivariate evaluation essential for identifying dominant water-quality patterns. This study assessed physicochemical, nutrient, organic-pollution, and microbiological indicators across three monitoring locations and three seasonal classes. Twelve primary indicators were evaluated through descriptive statistics, Kruskal-Wallis tests with Benjamini-Hochberg adjustment, epsilon-squared effect sizes, Spearman rank correlation, and standardized principal component analysis (PCA). Spatial differences were significant for all twelve indicators after multiple-testing correction, with particularly large effects for dissolved oxygen, phosphate, ammonia nitrogen, nitrite nitrogen, temperature, and nitrate nitrogen. Seasonal variation was comparatively weak; only specific conductivity remained significant after false-discovery-rate correction. Strong positive correlations linked ammonia with nitrite, nitrite with phosphate, conductivity with total coliform, and biochemical oxygen demand with total coliform. PCA showed that the first three components explained 73.03% of total standardized variance, with the first two accounting for 61.53%. The results demonstrate that spatial heterogeneity was the dominant organizing feature of water quality, while nutrient and organic-pollution variables formed coherent multivariate gradients. The approach provides a reproducible framework for prioritizing monitoring indicators and interpreting pollution dynamics in surface-water systems.