To explore the spatial and temporal response of water quality to external load reduction in Lake Taihu, Jiangsu Province, China, and clarify the exogenous load reduction under different water inflow and pollution conditions, a water quality model was constructed and the inflow boundaries were divided into seven groups based on the EFDC model. Taking COD and ammonia nitrogen as output targets, the sensitivities of Taihu Lake water quality boundaries were analyzed using a local sensitivity analysis. The results showed that COD and ammonia nitrogen concentrations of each lake area were more sensitive to the boundary load of the lake area than the rest of the lake area, and the sensitivity index was the highest in the Northwest Lake area. Furthermore, the improvement rates of mean COD concentrations in the whole lake decreased by 28.40%-34.71% in the dry season relative to the wet season, and the ranked sensitivity order of the boundaries was as follows:Northwest Lake boundary > Zhushan Lake boundary > Gonghu Lake boundary > Meiliang Bay boundary > Southwest Lake area boundary > Eastern Lake area boundary > East Lake Taihu boundary. The average improvement rates of ammonia nitrogen concentrations in the whole lake were 41.59%-42.34% higher in the dry season relative to the wet season, and the ranked boundary sensitivity order was as follows:Northwest Lake boundary > Meiliang Bay boundary > Zhushan Lake boundary > Gonghu Lake boundary > Southwest Lake boundary > East Lake Taihu boundary > Eastern Lake area boundary. This difference was affected by algal growth and metabolism, and artificial water diversion and drainage. Therefore, it is necessary to consider the reduction period and inflow location according to different water-quality indicators when planning external prevention and control measures in large lakes.
Keywords: EFDC model; Lake Taihu; external pollution load; local sensitivity analysis; water quality boundary conditions.