Identifying the Relationship between Assignments of Scenario Weights and their Positions in the Derivation of Reservoir Operating Rules under Climate Change

In order to mitigate the adverse impacts of climate change, adaptive operating rules (AOR) are generally derived using an ensemble of General Circulation Models (GCMs). Up to date, most of related literatures only focus on one fold of the following issues concerning the derivation of AOR using the GCMs ensemble, including: (1) consideration of different scenario weighing methods, or (2) analysis of different positions to locate scenario weights. And less concern is given to the latter compared with the former. However, few studies identify the relationship between (1) and (2) in the derivation of AOR based on the GCMs ensemble. In this study, we attempt to investigate where to use Equal and REA scenario weights in the derivation of reservoir operating rules under climate change. Equal weights (EW) and unequal weights based on the reliability ensemble average (REA) method are used in two positions: (I) the optimization objective of the reservoir operation model, which is to maximize the weighted average hydropower generation for all future scenarios; and (II) the incorporation of GCMs ensemble climate projections into the weighted climate conditions, and then it is input into the reservoir operation model with the objective of maximizing annual hydropower generation. Four AORs, including EW-AOR(I), REA-AOR(I), EW-AOR(II) and REA-AOR(II), are derived, and their optimized parameters are obtained by the simulation-based optimization (SBO) method with the Complex algorithm. The case study in the Jinxi Reservoir in China indicates that REA-AOR(I) outperforms the other three operation schemes, and EW-AOR(II) performs better than REA-AOR(II). Therefore, equal weights are preferably used to incorporate climate conditions, while unequal weights based on REA method can improve the performance of the reservoir operation model. Generally, REA-AOR(I) and EW-AOR(II) are suggested for adaptive reservoir management under climate change.