9493 - Laserfiche WebLink

Home Browse Search

FINAL REPORT, November 2003 High-jlow Requirements for the Duchesne River METHODS Hydrology Historical stream-flow gaging records were obtained from US Geological Survey online databases. USGS gaging station 09302000 provides a continuous record of daily mean discharge for the lower Duchesne River from water year 1943 to the present. We extended this record back in time to water year 1912 by correlation with the USGS gaging station 09295000 (Duchesne River at Myton, UT). The Myton gage, which is located approximately 35 km upstream from the Uinta River, has been in operation since July 1911. We generated synthetic daily mean stream- flow records for the Duchesne River near Randlett prior to water year 1943 using MOVEl methodology (Hirsch 1982) applied to overlapping data from the two gages spanning October 1, 1942, to September 30,2000. Hirsch (1982) demonstrated that maintenance of variance extension (MOVE) techniques perform significantly better than ordinary least squares regression in producing synthetic stream-flow records with statistical properties similar to those of the longer-running time series. Daily mean discharge measured at the gage near Randlett is linearly correlated with corresponding daily mean discharge at the Myton gage (R2 = 0.9539), although a subset of the larger discharges at Randlett increase at a faster rate relative to this relationship, probably because oflarge inflow from the Uinta River in some years (Figure 3). Transformation ofthe data does not substantially improve the relationship, hence extension analysis was performed on untransformed daily mean data. Many of the predicted daily mean discharge values near Randlett are less than the lower boundary of the 90-percent confidence envelope for flood events greater than about 7,000 ft3/s, suggesting that the magnitude of the largest events may be underestimated in the synthetic time series (Figure 4). We used the extended daily mean discharge record for the gage near Randlett to generate flow duration curves, average hydrographs, flood frequency analyses, and an annual flood discharge time series. Annual flood discharge for each water year was calculated as the total discharge in excess of a given flood discharge threshold. Thus, days for which the daily mean discharge was less than the threshold discharge contribute nothing to annual flood discharge. Days for which the daily mean discharge is greater than the threshold contribute the difference between the daily mean discharge and the threshold discharges in units of ft3 Is-days. 10