Abstract

The U.S. Geological Survey cooperated with South Dakota Game, Fish and Parks to characterize hydrologic information relevant to management of water resources associated with Sheridan Lake, which is formed by a dam on Spring Creek. This effort consisted primarily of characterization of hydrologic data for a base period of 1962 through 2006, development of a hydrologic budget for Sheridan Lake for this timeframe, and development of an associated model for simulation of storage deficits and drawdown in Sheridan Lake for hypothetical release scenarios from the lake. Historically, the dam has been operated primarily as a "pass-through" system, in which unregulated outflows pass over the spillway; however, the dam recently was retrofitted with an improved control valve system that would allow controlled releases of about 7 cubic feet per second (ft³/s) or less from a fixed depth of about 60 feet (ft).

Development of a hydrologic budget for Sheridan Lake involved compilation, estimation, and characterization of data sets for streamflow, precipitation, and evaporation. The most critical data need was for extrapolation of available short-term streamflow records for Spring Creek to be used as the long-term inflow to Sheridan Lake. Available short-term records for water years (WY) 1991–2004 for a gaging station upstream from Sheridan Lake were extrapolated to WY 1962–2006 on the basis of correlations with streamflow records for a downstream station and for stations located along two adjacent streams. Comparisons of data for the two streamflow-gaging stations along Spring Creek indicated that tributary inflow is approximately proportional to the intervening drainage area, which was used as a means of estimating tributary inflow for the hydrologic budget. Analysis of evaporation data shows that sustained daily rates may exceed maximum monthly rates by a factor of about two.

A long-term (1962–2006) hydrologic budget was developed for computation of reservoir outflow from Sheridan Lake for the historical pass-through operating system. Two inflow components (stream inflow and precipitation) and one outflow component (evaporation) were considered. The hydrologic budget uses monthly time steps within a computational year that includes two 6-month periods—May through October, for which evaporation is accounted for, and November through April, when evaporation is considered negligible. Results indicate that monthly evaporation rates can substantially exceed inflow during low-flow periods, and potential exists for outflows to begin approaching zero-flow conditions substantially prior to the onset of zero-inflow conditions, especially when daily inflow and evaporation are considered. Results also indicate that September may be the month for greatest potential benefit for enhancing fish habitat and other ecosystem values in downstream reaches of Spring Creek with managed releases of cool water. Computed monthly outflows from Sheridan Lake for September are less than 1.0 ft³/s for 8 of the 44 years (18 percent) and are less than 2.0 ft³/s for 14 of the 44 years (32 percent). Conversely, none of the computed outflows for May are less than 2.0 ft³/s.

A short-term (July through September 2007) data set was used to calculate daily evaporation from Sheridan Lake and to evaluate the applicability of published pan coefficients. Computed values of pan coefficients of approximately 1.0 and 1.1 for two low-flow periods are larger than the mean annual pan coefficient of 0.74 for the area that is reported in the literature; however, the computed values are consistent with pan coefficients reported elsewhere for similar late summer and early fall periods. Thus, these results supported the use of variable monthly pan coefficients for the long-term hydrologic budget.

A hydrologic model was developed using the primary components of the hydrologic budget and was used to simulate monthly storage deficits and drawdown for Sheridan Lake using hypothetical release scenarios of 1, 2, 3, 4, 5, 6, and 7 ft³/s for the months May through October for 1962–2006. Five computational years (3 years that include 1987–90 and 2 years that include 2004–06) tend to be "defining" years for purposes of drawdown considerations for Sheridan Lake. These 5 years include the only years with storage deficits for the historical pass-through operating condition (zero-release scenario). Computed drawdowns for 1987–90 approach or exceed 1 ft for many months for a prescribed release of 1 ft³/s and approach or exceed 2 ft during many months for a prescribed release of 2 ft³/s. For prescribed releases of 1 and 2 ft³/s, the maximum computed drawdowns of 1.6 and 2.4 ft, respectively, occur in February of the 1988–89 computational year. For a prescribed release of 7 ft³/s, computed drawdown exceeds 12 ft for one-half of the months of the 1989–90 computational year. During the 2004–05 and 2005–06 computational years, storage deficits were computed for 16 of 24 months for a prescribed release of 2 ft³/s.

Computed storage deficits occur for 25 of 36 months during the 1987–90 computational years for a prescribed release of 1 ft³/s, as opposed to a combined total of only 13 months with storage deficits for the other 41 years. For a prescribed release of 2 ft³/s, storage deficits are computed for 60 months during 15 of the 44 years, and deficits exist for 22 of the 24 months during the 1988–90 computational years. Computed storage deficits become progressively larger and more frequent for each of the progressively larger release scenarios.

Posted April 1, 2009