Entropy-weighted model for assessing the environmental safety of surface waters in the Southern Bug river basin

Abstract

Introduction. Ensuring the environmental safety of river basins is a critical challenge for Ukraine, particularly for strategically important waterways like the Southern Bug, which is subject to significant technogenic and agricultural pressure. Traditional assess-ment methods often rely on fixed weighting coefficients that fail to capture the local specifics of pollution. The aim of this study is to quantitatively assess the environmental safety of surface waters in the Southern Bug River basin using an entropy-weighted water quality index (EWQI), which accounts for the spatiotemporal variability and informational significance of hydrochemical indicators. Methods. The study is based on a database of hydrochemical observations for the period 2020–2024, collected from 36 moni-toring stations across the upper, middle, and lower reaches of the river. The analysis included 12 key water quality parameters. The methodology involved data normalization and the calculation of entropy weights using Shannon’s information theory to determine the contribution of each parameter to the overall pollution level. Analytical tools included the calculation of seasonal EWQI values, spatial visualization using OpenStreetMap (OSM), Principal Component Analysis (PCA) for factor identification, and k-means clus-tering for zoning the basin. Results. Spatial analysis revealed a distinct downstream gradient of water quality deterioration: from clean waters (Classes II–III, EWQI ≤ 1.0) in the upper basin to polluted and highly polluted waters (Classes V–VII, EWQI > 3.0) in the estuarine zone near Mykolaiv. A significant seasonal trend was established, with the mean EWQI increasing from 1.85 in the cold period to 2.46 in the warm period, indicating a 33% degradation in water quality due to intensified eutrophication processes. Entropy weight analysis identified ammonium (22%), phosphates (18%), and BOD₅ (15%) as the dominant contributors to the index, confirming the preva-lence of biogenic and organic pollution. PCA results indicated that three factors – organic load, nutrient enrichment, and mineraliza-tion – explain more than 80% of the variance in the data. Conclusions. The study confirmed that the entropy-weighted model provides an objective and sensitive tool for assessing aquatic ecosystems, effectively revealing spatial heterogeneity and seasonal risks. The research highlights that the warm season rep-resents a period of critical ecological stress for the Southern Bug. The practical value of the model lies in its applicability for auto-mated assessment and spatial mapping within the state environmental monitoring system, providing a scientific basis for optimizing monitoring networks and management decisions.