[Objective] Current research on waterlogging control in plain tidal river networks predominantly relies on numerical simulations, often lacking specific rule-based studies on control water level thresholds under extreme hydrological conditions. This deficiency frequently leads to excessive drainage from polders, exacerbating waterlogging risks within larger hydraulic zones. By simulating river network hydrodynamics, this study systematically diagnoses the deficiencies in existing dispatching methods and proposes a hierarchical optimization scheme centered on “pump activation water levels.” The aim is to enhance refined flood control dispatching and risk response capabilities in Shanghai and similar plain tidal river networks. [Methods] Taking the Qingsong Region in Shanghai during Typhoon “In-Fa” as a case study, we analyzed rainfall intensity, water level dynamics, and the status quo of the “two-level control and two-level drainage” dispatching mode. A river network hydrodynamic model was employed to simulate the region’s response under various rainfall scenarios, serving as the basis for proposing the hierarchical optimization scheme. [Results] The analysis indicates an inverse relationship between pump activation levels and flood risk: lower activation water levels in polders increase water storage capacity, effectively suppressing water levels in the main Qingsong Hydraulic Zone and enhancing overall flood safety. Reasonable control thresholds were identified to balance the restriction of excessive drainage with the maximization of river channel regulation and storage. Specifically, a pump activation level of 2.3 m is recommended when the forecasted daily rainfall is ≤100 mm, while a level of 2.0 m is advised when rainfall exceeds 100 mm. [Conclusion] To better coordinate polder management, we recommend conducting further dispatching experiments to validate these findings, refine control parameters, and facilitate the translation of research outcomes into regulatory guidelines. While this novel approach is applicable to other regions, the critical pump activation thresholds must be recalibrated for different hydraulic zones based on their specific external drainage capacities (pump/sluice scales) and river network storage volumes.