Multi-Objective Modeling Framework for Environmental Flow Optimization in a River-Reservoir System Using Histogram Comparison Approach for Estimation of Hydrologic Alteration

In the proposed Parameterization-Simulation-Optimization (P-S-O) framework for multi-objective E-flow optimization for long-term planning of reservoir operation, Histogram Comparison Approach (HCA) is adopted for the estimation of Hydrologic Alteration (HA) and the resulting performance is compared with that of Range of Variability Approach (RVA) in terms of the trade-off solutions obtained, optimal monthly E-flow targets and the individual alterations achieved in the river. A limiting constraint on the maximum average monthly E-flow deficits is included in the formulation in order to ensure realistic optimal E-flow targets at the reservoir level. The parameterization of the reservoir operation rules includes rule curves concerning carryover storage, Irrigation, E-flows and their transitions and fuzzified hedging factors as the decision variables with monthly water availability as the hedging trigger. The E-flow targets are derived as fractions of the respective mean monthly flows for the three hydrologic year types (dry, normal and wet), accounting for the intra- and the inter-annual flow variability. The meta-heuristic Borg-MOEA is employed as the search engine in the framework. The highly regulated Bhadra reservoir in Southern India is employed as the application case example. The individual alterations of the various hydrologic indicators appear to be reasonable for all the solutions along the P-O front of HCA, thus establishing its reliability in estimating HA and consequently the optimal E-flow targets derived and the reservoir operating policy. While, the RVA underestimates the individual alterations due to which the HA estimate, the optimal E-flow targets derived and the reservoir operating policy may not be reliable.