Optimizing environmental flow regime by integrating river and reservoir ecosystems

The present study develops a novel form of optimization framework to assess environmental flow in the reservoirs in which upstream and downstream river ecosystems and the lake ecosystem are assessed simultaneously. Physical habitat simulation was applied to compute ecological suitability in the river ecosystems for the fish habitats. Moreover, an adaptive neuro fuzzy inference system was utilized to simulate normalized population of the fish in the reservoir ecosystem. The ecological models were used in the structure of the reservoir operation optimization. Different measurement indices were used in the system performance measurement including reliability index, vulnerability index and root means square error. Three evolutionary algorithms were applied to optimize release including particle swarm optimization, biogeography-based optimization, and differential evolution algorithm. Based on the results in the case study, particle swarm optimization was the best algorithm to optimize ecological flow of the reservoir. The average physical habitat loss at upstream rivers as well as downstream river is less than 20%, which implies the proposed method is robust in terms of ecological flow assessment in rivers. Furthermore, normalized population in the lake is more than 50% that demonstrates the capability of the proposed method to balance ecological needs in the river and lake ecosystems.

Automated summary

The present study develops an optimization framework to assess environmental flow in reservoirs, considering the upstream and downstream river ecosystems as well as the lake ecosystem. The study uses physical habitat simulation and an adaptive neuro fuzzy inference system to evaluate ecological suitability and fish population in the ecosystems. Particle swarm optimization is identified as the best algorithm for optimizing ecological flow in the reservoirs.