WSPC12535 (2) - Laserfiche WebLink

Home Browse Search

Draft: Test Flood Effects on Lake Powell 0025i.8 which may pose hazards in-lake and downstream to both living organisms and to metal surfaces, such as the powerplant turbines. Drought conditions have resulted in several episodes of pronounced meromixis in Lake Powell since 1963, including the years preceding the test flood. Likewise, stagnation and DO demand has produced hypolimnetic hypoxia (as low as 1.4 mg DOlL near the dam). Extremely low DO concentrations have not yet reached discharge elevations (a minimum of 4.5 mg DOlL has been recorded at the penstock elevation). Meromictic stagnation is one result of dam design. Another is that the river outlet works are located deeper in the hypolimnion, although they are seldom used since they bypass power generation. Their location and operation affects meromixis as they draw entirely from the hypolimnion except during the lowest lake stage. Data suggest that higher flow-through, and ROW withdrawals may diminish the extent ofhypolimnetic meromixis. The existence and operation of Glen Canyon Dam (OCD) has significantly altered post-dam water quality in Glen and Orand Canyon (Stevens ef al. 1997). Seasonal variations in river flow, temperature, turbidity and ionic concentrations have been reduced to uniformly cold. clear, low nutrient waters that vary more on a daily and weekly basis than historic seasonal patterns. Water quality and discharge below the dam is now dictated by reservoir water quality and the dam operations (Stanford and Ward 1986, 1991, Angradi ef al. 1992). Interactions between the magnitude, duration, frequency, timing, and location of discharges from the dam influence uplake water quality which. in turn, determine downstream water quality. The effects of the large discharges and deep withdrawals occur in the context of seasonallimnological processes. obscuring cause and effect relationships. Historical data allow comparisons of similar antecedent conditions without large discharges. The 1996 test flood provided an opportunity to quantify these effects and elucidate the linkage between reservoir and downstream water quality. In this paper we address the following objectives: (1) describe the historical development of Lake Powell limnology; (2) determine whether the test flood's larger penstock discharges and releases from alternate structures affect Lake Powell limnology; (3) determine the extent of discharge required to produce measurable effects and how far uplake such effects are detected; and (4) determine the impacts on down-river water quality. The large historical database (1964 through 1997), and the large size of this reservoir make it an excellent study site. Analysis of the limnological changes associated with a single, large discharge event may contribute to improved management of this and other large reservoirs that develop meromixis or hypoxia, in addition to improving the linkage to downstream water quality. METHODS Study Area Glen Canyon Dam was completed in 1963, part of a series of dams resulting from the 1922 Colorado River Compact and the 1956 Colorado River Storage Act, providing for allocation and storage of water across the arid Colorado River basin. OCD is a 216.4 m high arch construction dam (Fig. 1). It provides 3 routes of release for the reservoir's water. Eight penstocks located 70 m below full pool elevation are the primary release structures. These can release a maximum of 940 mlls to the 8 turbines for 2 power generation, but are constrained to 892 ml/s. The penstock draft tubes release below the surface of the tailwater pool. limiting aeration effects. Two alternate release structures may be used for greater discharge capacity; but both bypass power generation and their use is avoided. The ROW are located 99 m below full pool (29 m below penstock outlets) and can discharge 424 -566 m'/s. Their greater depth facilitates hypolimnetic discharge. and they have been used on 7 occasions since 1963. The spillways draw from the epilimnion near the lake's surface at a depth of 16 m below full pool, although the lake has been below the spillways' operational levels over half the lake's history. The spillways have a capacity of 5890 m'/s to accommodate a 1 DO-year flood event, and have only been used in 1980, 1983 and 1984 (USSR 1970, 1995). QeD is one of the largest U.S. reservoirs; located in southern Utah and northern Arizona of southwestern USA. (Fig. 2). It first reached full pool in 1980, and has a maximum depth of 160 m. a surface area of 653 lan1, a length of 300 km, a volume of 32.1 lan' and approximately 3200 km of shoreline at the full pool elevation of 1128 m amsl (USSR 1970, 1995). The region has an arid continental climate-- annual precipitation is 200 rom/year and pan evaporation is 1800 mmlyear (Potter and Drake 1989). Lake Powell is an oligotrophic lake (potter and Drake 1989) with low nutrient levels; mean total phosphorus is 0.01-0.02 mg-PIL, total kjeIdahl nitrogen ~ 0.16-0.2 mg- NIL. Results from the 30+ year long-term Lake Powell monitoring program (LPMP) identify Powell as a warm meromictic reservoir; it has never completely mixed since its fonnation. It has a chemocline that persists near the depth of the penstock withdrawals. This meromictic hypolimnion, or monimolimnion, contains relatively stagnant water with elevated salinity (750 J.1S/cm to 1200 itS/em), cold temperatures (6-90C) and depressed DO (1.5-7 mg/L). A previous period of meromixis at Lake Powell was disrupted by high inflows and multiple. level discharges in the 1980's during 5 years of exceptionally high inflows. The spillways (near the surface) and the ROW were operated on several occasions for extended periods in 1980 and from 1983 to 1986. Combined with 3 years of high flow.through and multiple level withdrawals, the lake achieved a unique level of homogeneity in June 1985, with a conductance gradient 2.8 times less than the average for the lake's history. Data collection in the 1980's, however, was sporadic, with only 2 to 5 lake-wide collections/yr. Trends were discerned, but relationships between darn operations and uplake processes were less clear. We expected that analyses of the test flood results would clarify some of the effects observed in the 1980's. Data Collection and Sampling Design We utilized historical and ongoing data from our LPMP, augmented with higher spatial and temporal resolution data near the dam surrounding the test flood (Fig. 2). The LPMP includes 25 long-term monitoring stations. eight which have been sampled since 1964. The test flood was bracketed by 2 full-lake quarterly LPMP sampling trips in the weeks of 1 March and 6 June 1996. These included 25 stations in the Colorado, Escalante and San Juan River anns of Lake Powell. Using a Hydrolab Surveyor H20T... multi-parameter submarine sonde, we collected profiles of temperature (TOe), specific