The worth of streamflow data in water resources planning: Computational results

This paper presents computational results for evaluating the worth of streamflow data in water resources planning. Synthetic streamflow is generated with a synthetic streamflow generator, HEC-4 simulation model, where the optimal scheduling and capacity expansion of water supply projects are determined with a dynamic programing model. Penalty matrices associated with high-and low-yield estimates are then developed, using the results from the HEC-4 and the dynamic programing models. A multi-objective optimization problem is then formulated, minimizing the expected monetary penalty (in dollars) and the expected water shortage penalty (in million gallons). The solution of the multiobjective optimization problem (by using the surrogate worth trade-off (SWT) method) yields a Pareto optimal and preferred data collection policy.