WSP12616 - Laserfiche WebLink

Home Browse Search

001925 645 BAUCH & SPAHR: SALINITY TRENDS IN SURFACE WATERS OF THE COLORADO RIVER BASIN crease with the same yearly percentage that streamflow increases. For example, if streamflow at the Eagle River at Gypsum site increased 20% from 1982 to 1983, the dissolved-solids load for the same time period would increase at a rate <20%. Concentrations of dissolved solids are low in snowmelt runoff. When this runoff is added to the existing stream discharge, the total salinity load in the stream increases but at less than the percent- age that the stream discharge increases. Butler (19%) gives additional possibilities for variable annual dis- solved-solids loads: (i) errors associated with compu- tational methods; less uncertainty may be present in annual dissolved-solids loads calculated with stream dis- charge and specific conductance as compared to loads calculated only with stream discharge; (ii) errors in dis- solved-solids load calculations arising from errors in water-quality data, streamflow data, daily specific-con- ductance data, and regression equations; and (iii) year- to-year variations in precipitation, runoff, irrigation practices, and land use. time-period comparisons of concentration or monthly load data. Because these values are not very useful for direct com- parisons between stations, trend-slope magnitude (the relation of concentration or load to time) and percent rate of change between stations will not be compared for dissolved-solids concentrations and monthly loads. Trends for annual loads were calculated using linear regression. The flow-adjusted an- nual dissolved-solids loads were regressed against time to de- termine whether the regression slope significantly differed from zero. Because slopes from linear regression show a linear rate of change, comparison of slope magnitude and percent change in annual dissolved-solids loads between stations will be made. Further discussions on statistical tests for trends in water quality are in Hirsch et aI. (1982), Hirsch et aI. (1991), and Helsel and Hirsch (1992). RESULTS Annual Dissolved-Solids Loads The annual dissolved-solids loads for sites in the Up- per Colorado River Basin in Colorado for water years 1970 to 1993 are listed in Table 2. Also included are the mean annual dissolved-solids load and standard de- viations (SD) for each site. Annual dissolved-solids loads varied in the VCOL study unit during the 1970 to 1993 period. In general, the annual dissolved-solids load was highest when the total annual stream discharge was highest; for example, at Eagle River at Gypsum and Colorado River near Dotsero (Fig. 2 and 3). As explained by Butler (1996), though, there is not a 1:1 or linear relationship between dissolved-solids load and stream discharge. Dissolved-solids load does not in- Temporal Trends Results of trend tests on periodic dissolved-solids con- centrations and monthly and annual dissolved-solids loads for each site and time period analyzed are listed in Tables 3 to 5. Results include the trend-slope magnitude, the percent rate of change in the trend slope per year, the P value of the test, the significance level of the slope, and trend direction. Trend-slope magnitude and the per- cent rate of change are shown when the P value is <0.10, 700,000 3,500,000 600,000 3,000,000 ...\ \ \ \ \ \ ... " 500,000 '<ii' c 9 '-' 400,000 ~ 2,500,000 ~ e ~ 2,000,000 o ~ .c ~ is 1,500,000 ! ~ ""..., I \ '" '0 := o ~ 300,000 "S .e ~ is 200,000 1,000,000 100,000 -Stream Discharge - - - Dissolved-Solids Load 500,000 o o (\~ ~ g,1o g,'tI ~ ~ .# ...~ ~'tJ'V ~~ (\10 ~ fb~ ~ (\'tI ~ ~'\'V ~'\~ Water Year Fig. 3. Annual stream discharge and dissolved-solids loads for Colorado River near Dotsero (6), water years 1970-1993. Number after site name refers to site number in Fig. 1.