Model parameter sensitivity and identifiability for urban river water quality: impact of domestic wastewater discharges from on-site treatment facilities

Abstract

River water quality degradation is a risk to human health. Hence, many water quality models have been developed to predict the future states of water bodies and understand how the current water treatment systems will respond to future pollution loads and climatic drivers. A Japanese river was evaluated with the River Water Quality Model No.1 (RWQM1), and parameter sensitivity and identifiability analyses were executed on the model output using parameter sensitivity ranking, collinearity index, and Fisher Information Matrix-derived criterion. Among RWQM1 kinetic parameters, those related to hydrolysis, growth of aerobicheterotrophs, and first-stage nitrifiers were the most influential. Reactive soluble organic substances included in untreated gray waters, in addition to a prevalence ratio of the most advanced on-site treatment facility, strongly contributed to the model output variability. A remediation analysis revealed that a renewal to the most advanced on-site treatment facility by 20% increment was almost equivalent to the 70% decrease in the effluent concentration from an on-site treatment facility producing the highest pollutant load in terms of a BOD concentration decrease in the stream. This study provided baseline data assisting in policy implementation regarding the management of effluents from on-site treatment facilities. HIGHLIGHTS The impact of wastewater discharge from on-site treatment facilities was studied.; Sensitivity-identifiability analyses provided effluent management perspectives.; A prevalence ratio of the most advanced on-site treatment facility was the most influential parameter.; Easily soluble organic matter in untreated gray water should be prioritized as a wastewater constituent to be treated in the studied area.;

Keywords

• fisher matrix information
• identifiability analysis
• parameter collinearity
• river
• sensitivity measures
• water quality modeling