Downriver Water Quality Predicting in River Segment with Multi-Pollution Source via ANN

Abstract

Cutting the external waste loads can improve the water quality. The correlation analysis of discharged pollutant and water quality of the river is useful auxiliary for effective water quality management. ANN models were used to link the NH3-N concentration and emission loads and to predict downriver water quality in planned water quality functional section (WQFS) of Harbin, northeast of China. Besides conventional point and non-point pollution sources, polluted river branches were also considered as an influencing factor for downriver water quality. Downriver NH3-N concentration was predicted based on the loads and upriver water quality by ANN model. In general, it indicates that an acceptable agreement between the observations and the predictions has been achieved. All predicted results and analysis should provide useful suggestion to decision-makers for effective water quality management in the Harbin region.