WSPC12535 (2) - Laserfiche WebLink

Home Browse Search

002520 Draft: Test Flood Effects on Lake Powell 4 \ was in transition to a 20W interflow 135 Ian uptake. These conditions were similar to those in 1994 (155 krn) and 1995 (110 Ian) (Fig. 4). A second component of the freshening lOWU is the advective force it applies to the hypolimnion. While rarely able to penetrate the chemocline. the advective forces of the lOWI are often sufficient to depress the hypolimnion, creating a slow internal seiche or rocking motion across the hypolimnion. Internal seiches can be of much greater amplitude than surface seiches (Wetzel 1978, Cole 1994), and result in the chemocline ascending the face of the dam for a period of weeks to months. This effect can be seen in the 3-year forebay isopleths, with the "upwelling" effect typically beginning in February, peaking in March, and diminishing by May (Fig. 5, and the animation sequence). Prior to the test flood, upwelling had already peaked by mid.February and was subsiding. The upwelling effect is diminished: I) by discharge through the dam, and, 2) subsidence of the seiche as the advective forces of the 20WI dissipate. The synoptic channel profiles demonstrate annual winter upwelling cycles evident near the dam from 1994 to 1996 (Fig. 4). The upwelling pattern maximizes hypolimnetic discharge through the penstocks and ROW. However, the interflow pattern can confuse the interpretation of test flood impacts with seasonal hydrodynamics already underway. By late 1995, the 20WU had shifted to a 20WI, and its descent along the thalweg of the lake slowed as it impinged on the . pycnocline and diverted horizontally downlake toward the penstocks. From the onset of the test flood, inflow hydrodynamics actively affected reservoir limnology at the penstock elevation. Therefore, distinguishing test flood effects from existing seasonal change required an examination of rates of change on water quality and the impacts from the ROW. Effects on stratification and hydrodynamics Test flood effects on Lake Powell were observed through; shifts in chemoclines with consequent changes in strata volume, and through shifts in water quality. The synoptic channel profiles (Figs. 4 and 6) and temporal Wahweap forebay isopleths (Fig. 5) demonstrate the descending migration of the chemocline and DO gradients during the test flood. Comparisons with the previous year's upwelling and subsidence patterns show the test flood effects were most pronounced at the ROW depth, where the freshening effects of the 20WI were most dramatic. Prior to the test flood, 3 distinctive strata were distinguished from SC and DO concentrations at the Wahweap forebay station (Figs. 4, 6): 1) an upper convectively mixing (CM) epilimnion underlain by a distinct chemocline 7.5 m above the penstock outlets; 2) a middle 24 m~thick 20WI underlain by a second chemocline 13 m above the ROW; and 3) a lower 66 m-thick monimolimnion. Changing the elevation and magnitude of discharge restructured these layers. In general, increases in discharge result in a 3rd power increase in kinetic energy available for mixing, as KE 0: Q3 (Thorton et a/. 1990); this extends the vertical draw of the outlets. Hence, the increase from the nonnal penstock discharges of 392 m'!s to bi-level discharges of 849 m3/s and 424 m3/s from penstocks and ROW increased mixing and destratification by at least a factor of II, while total discharge only increased by 3.3~fold. The addition of sub-hypolimnetic discharge intensified vertical mixing. With the onset of the bi-Ievel high releases, the upper chemocline weakened as the penstocks drew more heavily from the CM epilimnion and the 20W!. Profile data at the dam demonstrated refreshment at the penstocks as they drew from the CM epilimnion (Fig. 3). But below the dam at Lees Ferry, co-mingled penstock and ROW releases show an overall increase in ionic concentrations, reflecting the dominance of ROW hypolimnetic output (Fig. 7). The chemocline below the 20WI and between the outlet ports initially sharpened, then weakened and descended more than 12 m to the level of the ROW at the conclusion of the flood. The 20WI stratum thickened from 24 m to 40 m as it drew from the thicker wedge uplake, entraining the CM epilimnion and hypolimnion and weakening the associated chemoclines as it moved downlake. Isopleths indicate the withdrawal zone extended from 50 to 100 or more km up~lake, even accounting for vertical error produced by localized seiche oscillations (Fig. 4 & 6). Chemical data collected near the dam before and after the test flood show consistent decreases in ionic concentrations by an average 4.40/0, demonstrating the refreshment of the forebay, particularly in the hypolimnion (fable I). The most pronounced shifts surrounding the test flood occurred near the ROW. This was not unexpected due to the meromictic conditions, the source of fresher conditions provided by the 20WI, and higher discharge. Surface and bottom samples demonstrated the least change. Calculations of the load of salt ions and DO versus relative discharge from the ROW and penstocks illustrate the disparity in discharge versus meromixis (Table II). Although the ROW only accounted for a third of the flood discharge, they contained 23% higher conductance and 33% less DO than is found at the penstocks. Consequently, the ROW discharges had a disproportionate freshening effect upon the hypolimnion compared with penstock withdrawals. Continued dilution of the hypolimnion was apparent (Fig. 5) following the test flood through 1997. This resulted from another high inflow year and continued high releases from February to June 1997, again, commenced during upwelling. Rates of change (in unit/day) for T"C, SC, and DO from modeled profile data spanning 28 February to 21 April indicate the greatest changes between 2-5 April immediately following the test flood (Fig. 8). The next highest rates of change were observed from March 22nd to April 2nd, during the test flood. Withdrawal Zone and in the Tailwaters Effects During the experiment, wind and discharge conditions contributed to water quality oscillations from the Wahweap forebay station to Lees Ferry (Figs. 3 and 7). We frequently observe internal seiche oscillations (Wetzel 1975) in high temporal resolution data sets during the winter at GCD. These oscillations are most evident when the reservoir's chemocline impinges on the penstock elevation, such as during the ascending and descending limbs ofhypolimnetic upwelling. Isolated wind events such as those on 20 April 1996 (National Climatic Data Center 1996), initiate wind~induced seiche oscillations ofToC, SC, DO and pH at the dam. As mentioned previously, changes in release rates also create rapid water quality shifts at the penstock level due to the strength and dimensions of the withdrawal plume (Hart and Sherman 1996, Hueftle and Vemieu 1998). Although at least 4 strong wind events occurred during the test flood and created complex