WRRCTMR No.64 Water Quality Simulation in Wahiawa Reservoir, O'ahu, Hawai'i

Abstract

Wahiawa Reservoir, a multiple-use facility, has historically experienced significant water quality problems, especially low dissolved oxygen (DO) concentrations which are spatially and temporally variable. To assist decision-makers in selecting among alternative water quality management strategies, the Water Quality for River-Reservoir Systems (WQRRS) model, developed for the U.S. Army Corps of Engineers, is applied to the Wahiawa Reservoir. The model is calibrated and verified to adequately represent dynamic behavior of vertical profiles of water temperature and DO. Data collected during December 1972 to November 1973 is used for calibration; data from July 1972 through November 1972 is the basis for verification. Although statistical analysis of calibration results shows no significant differences (at a 0.05 significance level) between observed and simulated water temperatures and DO, a variety of qualitative discrepancies are evident in these results. Simulated temperature, show a consistent positive bias of about 2°C; and simulated DO values tend to be too low during the winter and spring, although DO results correspond well with observed values during the critical low-flow period in the summer and fall. Model verification results show several important discrepancies (the source of which is unknown) from observed data. Hydraulic representation of the reservoir is questionable, as demonstrated by differences between simulated and observed water surface elevations. Temperature and DO results are statistically and significantly different from observed values. Observed data show more thermal stratification in the reservoir than is predicted by the model, which tends to over predict surface DO values, but which corresponds well with measured deeper water values. Although the results demonstrate a need to further refine the model, three preliminary specific alternative strategies are simulated: phosphate removal from WWTP effluent, diversion of WWTP effluents, and removal (dredging) of sediment organics. Simulation results suggest that none of these strategies by themselves are sufficient to eliminate anaerobic conditions in the reservoir. The occurrence of low DO is related to high surface productivities and the large reservoir of oxygen demanding sediments. Principal recommendations are to refine model calibrations, modify the model to allow simulation of artificial aeration, and further investigate alternative management strategies, including artificial aeration and combined management strategies.