عنوان مقاله [English]
Given the global shortage of water, optimized water usage and pollution prevention are vital issues. In this study, a part of the surface water collection network in the North East and South of Tehran has been studied. Two stations, one in each area, have been chosen for water quality monitoring. The time epochs for monitoring are not equidistant and thus, a Bayesian approach has been employed. Results showed that a Bayesian dynamic regression model had a proper fit to the most collected data. Total Dissolved Solids (TDS) have been modeled as a function of Electric Conductivity (EC), and Dissolved Oxygen (DO) that of Temperature (T). The predictions were reliable. For acidity parameters, a second degree polynomial time series showed a better fit. These models can be reliably used for short period predictions.