REGIONALIZATION OF NATURAL FLOW REGIME: APPLICATION TO ENVIRONMENTAL FLOW OPTIMIZATION AT UNGAUGED SITES

In this work we use the regionalization approach (RA) to derive the natural flow regime at an ungauged site. The derived natural flow regime, expressed by the regional cdf models of 32 indicators of hydrologic alteration (IHA) is used in the histogram matching approach (HMA) to seek the optimal environmental flows for a proposed multiobjective diversion weir in Taiwan. The results reveal that the outcomes associated with the planning constant scheme are significantly improved by the optimal time-varying scheme. The histogram dissimilarities of 32 IHA associated with the planning scheme are consistently greater than those associated with the optimal scheme, especially for low-flow IHA. Despite the inherent discrepancy between different approaches to generating flow data at the ungauged site, the optimal outcomes resulting from the RA-baged natural flow regime appear to be plausible and consistent with those reported in the previous work, thus validating the RA used in this work. We also explore the effect of weighting factors on the optimal outcomes. The results reveal that the weighting factor of the ecosystem needs objective dominates all optimal outcomes, while those of the agricultural demands and interbasin transfers objectives have minor effects on the optimal outcomes. The global optimal solution is obtained with a full or null weighting assigned to the ecosystem needs objective, while the least optimal solution is obtained as the interbasin transfers objective is given a null weighting and the agricultural demands objective is weighted more than the ecosystem needs objective. River managers and decision makers may select more balanced weir operation strategies based on the results presented in this work. Copyright (C) 2008 John Wiley & Sons, Ltd.