A framework for wastewater discharge estimation in data-scarce urban channels: integration of DREAM algorithm with pollutant transport simulation in the Cibarani channel

Urban channels in developing regions frequently suffer from severe water pollution due to poorly monitored domestic wastewater discharges. Conventional trial-and-error approaches for estimating lateral inflow rates are often time-consuming and inaccurate. This study explores a Bayesian inverse modeling approach using the DiffeRential Evolution Adaptive Metropolis (DREAM) algorithm, integrated with a pollutant transport model, to estimate discharge and reaction rates based on limited field-measured water quality data. The framework was applied to the Cibarani channel in Bandung, Indonesia, a polluted domestic sewage failing to meet Class II raw water standards. It simultaneously estimated twelve parameters – including lateral inflow and reaction rates – across both rainy and dry seasons. Model performance was validated using HEC-RAS simulations for DO, BOD, and NH3, achieving satisfactory agreement with observed data based on RMSE, R2, and NSE values. High-impact discharge locations were identified, supporting targeted and cost-effective pollutant control strategies. To our knowledge, this is among the first applications of DREAM-based inverse modeling to estimate lateral inflows in an under-monitored urban channel. The framework minimizes the need for high computational resources or advanced programming skills through the use of the DREAM Toolbox and MATLAB’s pdepe solver, offering a practical tool for water quality management in data-scarce environments.