Evaluating changes to reservoir rule curves using historical water-level data

Abstract

Flood control reservoirs are typically managed through rule curves (i.e. target water levels) which control the storage and release timing of flood waters. Changes to rule curves are often contemplated and requested by various user groups and management agencies with no information available about the actual flood risk of such requests. Methods of estimating flood risk in reservoirs are not easily available to those unfamiliar with hydrological models that track water movement through a river basin. We developed a quantile regression model that uses readily available daily water-level data to estimate risk of spilling. Our model provided a relatively simple process for estimating the maximum applicable water level under a specific flood risk for any day of the year. This water level represents an upper-limit umbrella under which water levels can be operated in a variety of ways. Our model allows the visualization of water-level management under a user-specified flood risk and provides a framework for incorporating the effect of a changing environment on water-level management in reservoirs, but is not designed to replace existing hydrological models. The model can improve communication and collaboration among agencies responsible for managing natural resources dependent on reservoir water levels.

Development and Evaluation of the Hydropower Reservoir Rule Curve for a Sustainable Water Supply

“…In Sweden, Bosona et al developed a rule curve using reservoir simulation to improve the performance of hydropower reservoirs [25]. In the United States, a study was conducted on the use of historical water level data to evaluate reservoir rule curves [26]. Previous studies developed and evaluated a rule curve for a single reservoir, while this study developed a rule curve for the Hwacheon Reservoir while considering the status of other reservoirs in the Han River basin.…”