SPDSS_Task2_IDKeyStreamflowGages

• Title: SPDSS Task 2 - Identify Key Streamflow Gages and Estimate Streamflows for Missing Records
• Description: The objective of this task is to identify key streamflow gages to use in the SPDSS modeling efforts and to develop a method for filling missing data. Revised memo to account for new information regarding the Balzac Gage. Revised February 2007.
• Decision Support - Doc Type: Task Memorandum
• Date: 2/10/2007
• DSS Category: Surface Water
• DSS: South Platte
• Basin: South Platte
• Contract/PO #: C153954
• Grant Type: Non-Reimbursable
• Bill Number: SB01-157, HB02-1152, SB03-110, HB04-1221, SB05-084, HB06-1313, SB07-122
• Prepared By: Leonard Rice Engineering

Together, the two Balzac gages contain historical data for the entire SPDSS study period.
However, the Prewitt Reservoir inlet canal, the Johnson and Edwards Ditch, and the Tetsel Ditch
diversions are located between the old and new Balzac gage locations (the Prewitt Inlet Canal
and Johnson and Edwards Ditch share a headgate but are gaged separately.) To combine the
Balzac gages, historical diversion data for the Prewitt Inlet Canal, Johnson and Edwards Ditch,
and Tetsel Ditch were added to the South Platte River at Balzac historical gage data for the
period 1950 to 09/30/1987, effectively adjusting the South Platte River at Balzac gage data
upstream to the current location. This adjustment was completed using the following data:
Johnson and Edwards Ditch (0100526) -HydroBase records complete for 1950 through 2005
Tetsel Ditch (0100525) -HydroBase records complete for 1950 through 2005
Prewitt Reservoir inlet ditch -HydroBase records are stored under reservoir ID 0103552
from 1970 through 1996 and inlet canal ID 0100829 starting in 1996. HydroBase does not
contain any records prior to 1970. Historical monthly Prewitt Reservoir diversion records
were obtained from the Prewitt Reservoir Operating Committee for the period 1926 through
1987 (this data has been provided to the State for review to determine if it is appropriate to
add to HydroBase). Comparison of the Committee records to HydroBase shows a very
strong correlation (R~ value of 0.95). Therefore, HydroBase records for reservoir ID
0103552 were used for 1970 through 1987 and supplemented with Committee records for
1950 through 1969.
Approach and Results -Fill Missing Records for Key Streamflow Gages to be used in the
Ground Water and Water Budget Modeling
Missing streamflow data can be filled using techniques developed in previous DSS modeling
efforts, which generally involve using regression with upstream, downstream, or nearby gages.
TSTooI was used to fill missing streamflow data using simple linear regression with the best
available streamflow data at other gage locations. A detailed review of filling methods is
provided in the Attachment A, "Methods for Filling Missing Streamflow Records (In Support of
Task 2 -Identify Key Streamflow Gages and Estimate Streamflows for Missing Records)".
Table 3 (sorted alphabetically) provides the regression analysis results for the ground water and
water budget gages. For each regression, the average annual streamflow, coefficient of
determination (R~), and square root of the mean of the sum of the square errors (RMSE) are
provided. The RMSE is another measure of the difference between estimated values and the line
of best fit. For a perfect fit, R~ = 1 and RMSE = 0. The filled results are shown in Figures 6
through 11 below. Although monthly logarithmic regression is commonly used to fill missing
streamflow data, monthly linear regression produced the best correlations for most of the
dependent gages being filled under this task. An exception is the Sand Creek at Colorado-
Wyoming State Line (06659580) gage, where annual linear regression and monthly historical
averages were used (described further below).
Task2.doc 20 of 30