Hydrologic analysis and model development for cannonsville reservoir

The hydrology of Cannonsville Reservoir, a water supply and flow augmentation reservoir for the City of New York, is analyzed. Measurements of components of the hydrologic budget of the reservoir over the entire 30 years of operation are reviewed. The largest tributary is gauged near its discharge to the reservoir; all others are ungauged. Water surface elevation and outflow, in the form of spillway flow, direct releases for flow augmentation, and drinking water withdrawal are monitored. Multiple uses of the reservoir and variations in runoff common to the region result in strong seasonal and interannual variations in hydrology. Spillway flow typically occurs in late winter and spring, and is rare in summer and early fall. Releases, to meet minimum flow requirements in the Delaware River downstream, generally occur only in summer and early fall. Drinking water withdrawal is relatively uniform throughout the year. On average, spill, dam releases, and withdrawal for the water supply have represented 40%, 32%, and 28%, respectively, of the outflow of the reservoir over the 30 years. Reservoir water surface elevation is generally greatest in the spring and lowest in the fall. Reductions in maximum depth of 40% and in storage volume of 75% have occurred in dry years. A hydrologic model was developed to maintain a hydrologic balance and estimate surface inflows from the ungauged portion of the watershed. While estimates based on a simple ratio of ungauged flow to gauged flow may be used for annual averages, a more complex budget calculation based on a 10-day averaging period was used to provide a time series of daily average ungauged inflow. The average annual flushing rate for the 30 years was 2.6 y⁻¹ while the range was 1.9 to 3.6 y ⁻¹. Use of the entire 30 years hydrologic record is recommended to support forecasting with hydrothermal and water quality models.