Laserfiche WebLink
<br />N <br />= <br />~ <br />00 <br /> <br />where <br /> <br />C <br />e <br /> <br />= estimated dissolved-solids concentration at the record- <br />extension site, <br />= dissolved-solids concentration at the base-station site, and <br /> <br />~ <br />a and b = <br /> <br />regression coefficients. <br /> <br />Flow-weighted monthly concentrations were used for all sites. <br /> <br />Results <br /> <br />The three techniques were tested using monthly dissolved-solids data from <br />the 13 record-extension sites in this study. Weighted regression on stream- <br />flow was applied for seven sites that had streamflow data during the entire <br />period of dissolved-solids record extension. Weighted regression on stream- <br />flow at a base-station site was applied for the five sites that did not have <br />streamflow data during the record-extension period. Simple regression on <br />dissolved-solids data at a base-station site was applied for six sites that <br />had suitable base-station sites with water-quality data. Results are listed <br />in table 4. Calibration periods used to fit the models were selected to <br />decrease the effects of flow regulation upstream. The only sites affected by <br />regulation during the calibration period were site 12, Duchesne River near <br />Randlett, Utah; and site 23, Colorado River below Parker Dam, Ariz.-Calif. <br />Streamflow at these sites was regulated during both the historical period of <br />record and the entire record-extension period. <br /> <br />For sites at which two record-extension techniques could be applied, the <br />appropriate ,model was selected based on a criterion of minimizing the standard <br />error. Weighted regression on streamflow yielded better results for sites <br />upstream from Lake Powell (sites 1, 9, and 17). Simple regression on <br />dissolved-solids concentration at a base-station site was more accurate for <br />main-stem sites downstream from Lake Powell (sites 19 and 24). Standard <br />errors of the selected models ranged from 3 to 27 percent of the observed mean <br />monthly dissolved-solids concentration. <br /> <br />COMPARISON OF EXTENDED AND HISTORICAL RECORDS <br /> <br />A summary of techniques used to extend streamflow and dissolved-solids <br />records for the 13 record-extension sites is listed in table 5. This table <br />also lists the periods, in water years, during which the record was extended. <br />Graphs of the extended and historical records are shown in figures 2-19. The <br />monthly values of streamflow and dissolved-solids concentration computed for <br />the extension periods are presented in tables 6-19 in the "Supplemental Data" <br />section at the back of this report. <br /> <br />Seasonal and annual variations in streamflow and dissolved-solids <br />concentrations were comparable for the extension and historical periods at all <br />sites. Most of the values estimated for the extension period were within the <br />range of values observed during the historical period. Estimated streamflow <br />values that were not within the historical range occurred at sites 5 (Dolores <br />River near Cisco, Utah) and 9 (Green River near Greendale, Utah). Estimated <br /> <br />11 <br /> <br />