A Geographic Information System-Based Framework for Discriminating Point and Non-Point Source Pollution in Data-Limited River Basins: A Case Study from Sri Lanka's Kelani River

Urban river systems in rapidly urbanizing regions of South Asia face increasing threats from wastewater discharges. This study develops a transferable GIS-based framework for discriminating point- and non-point-source pollution in data-limited tropical basins, using the Kelani River Basin of Sri Lanka as a case study. Seventy-one surface and groundwater samples were collected across headwater, transitional, and meandering zones during dry and wet seasons and analyzed for key physicochemical and biological parameters. The Canadian Council of Ministers of the Environment Water Quality Index (CCME-WQI) was calculated based on national and international water quality guidelines. Results classified overall water quality as "Marginal," revealing a classic spatial pattern of degradation from "Excellent" (WQI: 93) in pristine headwaters to "Poor" (WQI: 26) in the urban meandering zone, a trend observed in many urbanizing watersheds globally.​ Crucially, Spearman correlation analysis identified ammonium nitrogen (NH4+-N) (R < -0.7, p < 0.001) and total phosphorus (TP) as the primary drivers of degradation. The persistent, seasonally consistent NH4+-N signal is a possible robust indicator of chronic raw sewage input (point-source pollution), while wet-season TP spikes implicate agricultural and urban runoff (non-point source pollution). The study presents substantial inferential evidence between pollution sources, demonstrating that the river's assimilative capacity is overwhelmed by continuous point-source discharges with seasonal nutrient mobilization, a critical issue in regions where sanitation infrastructure lags behind urban growth.​ The framework provides a replicable approach for identifying hotspots and assessing river assimilative capacity. These findings highlight the global imperative​ for improved nutrient removal and the value of integrated, GIS-based monitoring systems to support adaptive governance​ under accelerating urban and climatic pressures.