Supersaturated total dissolved gas (STDG) caused by the flood discharge can result in bubble disease and even fish death, posing potential risks to aquatic ecosystems. In order to reduce the impacts of STDG on such ecosystem, a dynamic multi‐objective STDG management model (DMO‐STDGM) with the flood pulse discharge pattern is developed based on the prototype observation and laboratory test analysis, which consists of STDG prediction, fish safety assessment, multi‐objective optimization and 2‐D STDG transport diffusion simulation modules. The proposed model can support generation of flood pulse discharge patterns to reduce the level of STDG and minimize the maximum residence time of STDG water. The developed model is applied to optimize the discharge pattern of Xiluodu project to reduce the effects of STDG on fish. Simulation results of the worst scenario with the flood pulse discharge pattern show that the maximum residence time of STDG for T120%, T125%, T130%, and T140% are 12, 9, 4, 3 and 1 h, which are less than the corresponding LT50s (Mean Lethal Time). It means that the physiological function of fish can recover from the affect of STDG, thus fish can survive in the worst scenario. The established model can also be employed for increasing power generation. Simulation results indicate that the flood pulse discharge pattern with equal weight of ecology and power generation can ensure full load operation of hydropower station with 12,600 MW. The established model is widely applicable and can provide an important theoretical and technical support for water quality and ecosystem management under multiple complexities.