Abstract: The poor water mobility of river channels in a plain river network region is the main cause of severe water environment pollution; water diversion is an important measure to improve the water environment in such regions. This paper constructs a hydrodynamic and water quality coupling model for river networks based on the measured data, and presents an application to the Sanshanwei area in Foshan, Guangdong. This model is verified against a 15-day continuous series of hydrodynamic and water quality monitoring data, and it is evaluated using the NSE and RMSE indexes. For this area, we have designed four gate control modes and seven landscape control water levels based on its terrain features, tidal motions, and landscape requirements. A total of 28 simulation conditions are used to simulate the river networks and analyze the mechanism of its all-sided responses to different conditions－including the improvement of its hydrodynamic water quality and the variations in the tidal levels of its inner and outer rivers; water diversion flow and drainage flow, and their spatial distributions; gate control modes and landscape water levels. The results show the model is reasonable and reliable. The study area is affected by its drainage path and has a significant difference in river flow mobility. We have achieved very significant improvement of the water environment through considering comprehensively the effects of river flow distribution, water flow path, and diversion effect of bifurcated river channels; considering the pollutant concentration in the outer rivers on the water quality; combining effectively the dynamic water environment capacity of the tidal river network and the spatial and temporal distribution of pollution source discharge. Compared with the low landscape control water level, our control modes of high landscape water levels can increase the dynamic water environment capacity of inner rivers, resulting in relatively low pollutant concentration. When the control level is raised from 0.2 to 0.8 m, the diversion flows in different diversion paths are increased by 28.0% - 64.7%, and the ammonia nitrogen concentration in the section reduced by 0.85 - 5.50 mg/L or a reduction ratio of 28.9% - 67.2%. This study presents a new idea for scheduling optimization and water environment in plain tidal river networks, useful for designing corresponding engineering measures.

Key words: SWMM, one dimensional river network model, water diversion scheme, landscape control water level, hydrodynamic-water quality couple model