Estimated Decreases in Dissolved-Solids USGS Loads in Four Tributaries to the Colorado River science for a changing world in the Grand Valley, Colorado, 1973-96

The Colorado River is used extensively for municipal, agricul­ tural, and industrial purposes in seven Western States. Since water development began in the 1870's, the average dissolved-solids con­ centrations in the lower Colorado River has more than doubled when compared to pre-development con­ centrations (U.S. Department of

the Interior, 1995). The dissolved- -^^^-

solids concentration, commonly p DENVER COLORADO referred to as salinity, is a measure HJBS?=» . Colorado springs of salts, such as sodium chloride, Pueblo calcium bicarbonate, and calcium sulfate, that are dissolved in water. Dissolved solids can have adverse 0 5 KILOMETERS effects on agricultural, municipal, Figure 1. Location of the Grand Valley, extent of irrigated area, and locations of study sites. and industrial uses, especially in the lower part of the basin. The fields. The salinity control project Valley were reported by Butler Colorado River Basin Salinity in the Grand Valley consisted of (1996) However, dissolved-solids Control Act, passed by the U.S. two parts: (1) The BOR lined parts concentrations and loads also Congress in 1974, authorized the of the Government Highline Canal had significant downward trends in Bureau of Reclamation (BOR) and (fig. 1) and placed some laterals in the Colorado River upstream the U.S. Department of Agriculture pipe to decrease distribution-sys­ from the Grand Valley, which (USDA) to plan and construct tem losses, and (2) the USDA introduced some uncertainty in salinity-control projects to (through the Natural Resources determining the effect of the Grand decrease dissolved-solids loading Conservation Service) was respon­ Valley salinity-control projects on to the Colorado River. One of the sible for on-farm improvements to the dissolved-solids trends in the authorized projects was a salinity- decrease water losses from on-farm Colorado River. control project to address dis­ ditches and to decrease deep perco­ Examining dissolved-solids solved-solids loading from irri­ lation from fields. loads in tributary streams and gated agricultural lands in the In 1994, the BOR requested washes that dissect the Grand Grand Valley (fig. 1). Because the U.S. Geological Survey Valley (fig. 1) could be a more ground water that was recharged (USGS) to examine trends in dis­ direct measure of effects of salin­ primarily from irrigation practices solved-solids data for the Colorado ity-control projects than are dis­ was the major source of dissolved- and Gunnison Rivers to examine solved-solids loads in the Colorado solids loading from the Grand the effectiveness of the salinity- River. These streams and washes Valley (Bureau of Reclamation, control projects in the Grand Valley carry much of the irrigation- 1978), the salinity-control project (U.S. Department of the Interior, induced dissolved-solids load was designed to decrease the 1995). Significant downward resulting from subsurface irriga­ recharge to this ground water from trends in dissolved-solids concen­ tion return flow to the Colorado canal, lateral, and ditch leakage and trations and loads in the Colorado River. If salinity-control projects in from deep percolation in irrigated River downstream from the Grand these relatively small watersheds were effective, there could be mea­ the Grand Valley. By 1996, suffi­ rized in table 1. It is more surable decreases in dissolved-sol- cient streamflow and chemical data appropriate to compare dissolved- ids loads in the streams in these had been collected to estimate solids loads between time periods areas. dissolved-solids loads for four of if loads were calculated using the During 1973-83, the USGS, the sites (fig. 1) at former gaging same method. Because only peri­ in cooperation with the BOR, stations: Leach Creek (site LC), odic water-quality and streamflow operated several gaging stations on Big Salt Wash (site BSW), Reed data were collected during 1991- selected streams in the Grand Wash (site RW), and Salt Creek 96 (table 1) at the four sites, only Valley. Daily streamflow and spe­ (site SC). Dissolved-solids loads periodic data were available for cific conductance were recorded, during 1973-83 then could be computing dissolved-solids loads and water-quality samples were compared to dissolved-solids for 1991-96. Therefore, to com­ collected at some of the stations. loads for 1991-96 for the four pare the loads for 1973-83 to loads All the gaging stations operated by sites to determine the effects of for 1991-96, the loads for 1973-83 the USGS, where water-quality salinity-control projects in the also needed to be computed using data had been collected, were dis­ Grand Valley. only periodic data. continued by October 1983. In This fact sheet describes The method for estimating 1991, the National Irrigation Water changes in dissolved-solids loads loads using only periodic water- Quality Program (NIWQP) of the in four streams in the Grand Valley quality and streamflow data is U.S. Department of the Interior between 1973-83 and 1991-96 referred to as the periodic method. began a study in the Grand Valley. and the relation between changes In this method, loads are computed The focus of the NIWQP study was in loads and implementation of for individual samples by using the on selenium sources and pathways salinity-control projects. The fact dissolved-solids concentration and effects of selenium on fish and sheet also describes the methods from the chemical analysis of water birds (Butler and others, 1996). used to compute dissolved-solids samples and the streamflow mea­ Although no gaging stations were loads and the rationale for making sured when the sample was col­ reactivated in support of the the load comparisons between lected. The individual dissolved- NIWQP study, water samples for 1973-83 and 1991-96. solids loads then are separated into chemical analyses that included irrigation-season and nonirriga- dissolved solids were collected at tion-season samples, and means are several sites where the gaging sta­ Methods Used To Estimate computed for each set of seasonal tions had been located in 1973-83. Dissolved-Solids Loads data. A mean annual load is deter­ Data collected for NIWQP were to mined by computing a weighted be used to estimate selenium and Daily streamflow, daily mean of the two seasonal mean dissolved-solids loads at outflow specific conductance, and periodic loads. The weighting factors are sites on the major streams and water-quality data that were col­ based on the fraction of the year washes draining irrigated areas in lected at the four sites are summa­ each season encompasses (the

Table 1. Sampling sites and summary of daily and periodic data collected in 1973-83 and 1991-96

[FOR, period of record; N, number of samples]

Period of record, daily Sampling USGS streamflow and Periodic water-quality sampling site station Stream name conductance (fig-1) number 1973-83 1991-96 1973-83 1991-96 POP N POP N LC 09152650 Leach Creek 04/73-09/77 None 08/73-07/77 28 03/91-12/96 32 BSW 09153270 Big Salt Wash 03/73-09/77 None 08/73-07/77 29 03/91-12/96 20 RW 09153300 Reed Wash 04/73-09/83 None 08/73-10/82 48 03/91-12/96 39 SC 09163490 Sail Creek 04/73-09/83 None 08/73-09/83 53 03/91-11/96 30 irrigation season encompasses solids load to streamflow, and the in annual loads ranged from about about 59 percent of the year, the second equation relates dissolved- 4 to 8 percent (table 2). Using the nonirrigation season about 41 per­ solids load to streamflow and periodic method, the annual dis­ cent of the year). The reason for specific conductance. The equa­ solved-solids loads for Leach using a seasonally weighted mean tions then are used to compute Creek and Big Salt Wash were less is because of the large differences daily loads for the time period of than the annual loads computed in streamflow and dissolved-solids interest using the daily streamflow using the SLOAD method; the loads between the irrigation and record and, if available, daily spe­ opposite was true for Reed Wash nonirrigation seasons in many cific conductance. The daily loads and Salt Creek (table 2). For the Grand Valley streams and washes. are summed to give monthly and four stations, the annual loads If the periodic sampling was biased annual dissolved-solids loads. based only on periodic data are not to one of the seasons (for example, Dissolved-solids loads computed substantially different than the if most of the samples were col­ using the SLOAD method should annual loads computed using lected in the irrigation season), be a more accurate estimate of SLOAD. The implication is that then an arithmetic mean on all the dissolved-solids loads when com­ the periodic dissolved-solids-load samples might result in an unrealis­ pared to the periodic method data were collected over the gen­ tic mean dissolved-solids load, if because much more detailed data eral range of streamflow and dis­ that mean is assumed to represent (daily streamflow and specific solved-solids concentrations for the daily load throughout the year. conductance) are used instead of several years at these four sites and Using a seasonally weighted mean only periodic dissolved-solids data. that fairly reliable estimates of removes seasonal sampling bias. A comparison of the two methods annual dissolved-solids loads can Once the seasonally weighted is provided in the next section of be obtained without having daily mean is calculated, the resultant this fact sheet. streamflow data. A disadvantage of mean load is assumed to represent using only periodic samples is that the mean annual dissolved-solids runoff from large storms may be load and is multiplied by 365.25 Comparison of Methods missed. In the Grand Valley, such to estimate the total annual dis­ storms are relatively infrequent and solved-solids load for a multi-year Annual dissolved-solids often of short duration. At least for period in which the sampling was loads calculated by the two meth­ the periods when the four gaging done. ods are compared in table 2. The stations were operated, results indi­ To determine if annual dis­ loads are applicable to the periods cated that using only periodic data solved-solids loads computed when the gaging stations were for computing dissolved-solids using only periodic data are repre­ operated at the four sites during loads provides reasonable esti­ sentative of actual loads, the loads 1973-83 (table 1). The differences mates of the annual loads. for 1973-83 also were computed using another method called Table 2. Comparison of annual dissolved-solids loads computed using two methods SLOAD. A computer program named SLOAD (Salt LOAD) was [Periodic method, only periodic water-quality and streamflow data are used to estimate loads; SLOAD developed by Liebermann and method, daily streamflow and specific conductance are used to estimate loads: loads for Leach Creek and Big Salt Wash are based on data collected during 1973-77, and loads for Reed Wash and Salt Creek are based on others (1987) for estimating dis­ data collected during 1973-83; loads are in tons per year] solved-solids loads in streams and rivers. In the SLOAD method, Annual dissolved- Difference in loads, Stream solids load periodic minus SLOAD daily streamflow, daily specific- Periodic SLOAD Load Percent conductance, and periodic Leach Creek 41,400 44,900 -3,500 -7.8 dissolved-solids data are used to Big Salt Wash 83,200 88,200 -5,000 -5.7 estimate loads. Two equations are Reed Wash 139,600 130,000 9,600 7.4 computed using the periodic data. Salt Creek 131,300 125,500 5,800 4.6 One equation relates dissolved- Comparison of Dissolved- 140,000 Solids Loads for Two Study 130,000 Periods ^ 120,000 O \ \ LOAD FOR EARLY PERIOD £ 110,000 l=l LOAD FOR 1991-96 ^ Estimated dissolved-solids 0 100,000 2 90,000 P loads in Leach Creek, Big Salt c/) § 80,000 Wash, Reed Wash, and Salt Creek 0 «? 70,000 were lower in 1991-96 than in the 0 earlier period (table 3; fig. 2). Only > 60,000 W 50,000 the dissolved-solids loads com­ W ° 40,000 < puted using the periodic method 1 30,000 are compared for the earlier < 20,000 period and 1991-96. Decreases 10,000 in the annual dissolved-solids 0 LEACH BIG SALT REED SALT loads ranged from about 13.3 to CREEK WASH WASH CREEK 20.6 percent in the four streams. Figure 2. Annual dissolved-solids loads in four streams in the Grand Valley for two study periods. Early periods are prior to Bureau of Reclamation The largest decreases in annual dis­ salinity-control projects and are 1973-77 for Leach Creek and Big Salt solved-solids loads were in Reed Wash; 1973-80 for Reed Wash; and 1973-83 for Salt Creek. Wash (27,600 tons) and Salt Creek in 1983. Because BOR salinity- work was started in 1991 and was (27,000 tons). control work was occurring in completed by early 1993. Because The changes in loads for the Reed Wash during 1981-83, the of the overlap of the salinity-con­ four streams are of sufficient mag­ dissolved-solids load for the earlier trol work with the 1991-96 sam­ nitude to indicate that dissolved- pling period in Big Salt Wash, the solids loading from the Grand period was determined for 1973- 80 (table 3; fig. 2) instead of 1973- loads for 1991-96 might not com­ Valley has decreased over the last pletely reflect effects of salinity 23 years. The largest single known 83 In the Salt Creek Basin, the BOR did salinity-control work control. In the Leach Creek Basin, effect on dissolved-solids loads in the BOR salinity work did not start the Grand Valley was the imple­ from 1986 to 1991. Decreases in until 1995. mentation of salinity-control dissolved solids resulting from The second part of the Grand projects. In the Grand Valley, the BOR work in the Reed Wash and Valley salinity-control project is BOR salinity-control projects Salt Creek Basins would have the on-farm work done by the began in late 1980 with the Stage I occurred before the 1991-96 USDA. The USDA implemented project in the Reed Wash Basin. sampling period. In Big Salt Wash the on-farm program in 1979, and The Stage I project was completed Basin, the BOR salinity-control work has been done at various farms throughout the Grand Valley Table 3. Comparison of annual dissolved-solids loads for the 1973-83 and 1991-96 study periods since 1979. Because of the variable nature of the on-farm program, the [Annual dissolved-solids loads for earlier periods are prior to Bureau of Reclamation effects of that salinity-control work salinity-control projects in the basin, earlier periods are: 1973-77 for Leach Creek and Big Salt Wash; 1973-80 for Reed Wash; and 1973-83 for Salt Creek; loads for both study are cumulative since 1979. To doc­ periods were computed using periodic method; loads, in tons per year] ument the full extent of the effects of the salinity-control projects on Change in load, 1991 -96 to the dissolved-solids loads in the four Annual load, Annual load Stream early period, earlier period for 1991-96 study basins, additional sampling in tons and percent would be needed after all salinity- Leach Creek 41,400 35,000 -6,400 (-15.5%) control projects in the study basins Big Salt Wash 83,200 72,100 -11, 100 (-13.3%) are completed. Reed Wash 138,900 111,300 -27,600 (-19.9%) Other factors, in addition to Salt Creek 131,300 104.300 -27,000 (-20.6%) salinity-control projects, could Total 394,800 322,700 -72,100 (-18.3%) affect dissolved-solids loads. Changes in dissolved-solids loads in parts of the four study basins that to other uses during the study in dissolved-solids loads for the are upstream from the irrigated period. three basins would be greater than areas (north of the Government 53,000 tons per year if the salinity Highline Canal in fig. 1) could Relation of Salinity-Control decreases from the USDA on-farm affect loads measured at outflow Projects to Dissolved-Solids salinity-control projects were stations. However, on a long-term Loads included; however, these decreases basis, annual dissolved-solids have not been determined for spe­ The previously discussed loads in the four study basins from cific drainage basins. analysis of dissolved-solids data areas upstream from irrigated land Additional analysis of other collected by the USGS since 1973 in the Grand Valley might not be a water-quality and streamflow indicated that annual dissolved- major component of the loads mea­ data tends to substantiate the dis- solids loads have decreased in four solved-solids-load decreases sured at the outflow sites. Of the streams that drain irrigated areas in four study basins, the Salt Creek reported in this fact sheet. For the western Grand Valley. In the example, review of the annual Basin has by far the largest drain­ Big Salt Wash, Reed Wash, and age-basin area outside of the irri­ streamflow data for the USGS gag­ Salt Creek Basins, implementation ing station on Reed Wash at site gated area. Data collected by the of salinity-control projects by the USGS on tributaries of Salt Creek RW2 (fig. I) indicates a decrease in BOR and the USDA was the most annual mean streamflow after 1982 during the 1970's and early 1980's significant known change in these (U.S. Geological Survey, 1974-83) (fig. 3). The Stage I salinity-control basins that could cause decreases in project lined 6.8 miles of the Gov­ indicated that the annual dissolved- dissolved-solids loads. In those ernment Highline Canal and placed solids loads discharging into Salt three drainage basins, the projected 34.2 miles of open laterals in pipe Creek from the drainage basin decrease in dissolved-solids loads in the upper Reed Wash Basin from upstream from the Government from the BOR part (canal lining 1981 to early 1983 (Bureau of Rec­ Highline Canal are relatively small and piping of laterals) of the salin­ lamation, 1985). Some of the when compared to the load com­ ity-control project was about ground-water recharge from canal puted for the outflow site near 1-70 53,000 tons per year (U.S. Depart­ and lateral leakage in upper Reed (site SC in fig. 1). ment of the Interior, L995; Bureau Wash was eliminated, and the Another potential effect on of Reclamation, written commun., Stage I work should have dissolved-solids loads in streams 1997). The total projected decrease decreased ground-water discharge and washes draining the Grand " 60

Valley is conversion of agricultural O O MEAN ANNUAL STREAMFLOW, WATER YEARS 1976-82 land to residential and commercial UJ 56 CC ' development. Conversion of irri­ LU Q. 52 gated land to other uses could LU LU MEAN ANNUAL STREAMFLOW, U. 48 pi rn affect dissolved-solids loading in WATER YEARS 1983-96 U m the Grand Valley. However, the - 3 44 [-, effect of land-use changes on dis­ z solved-solids loads has not been 5 4° o documented for the Grand Valley. - E| 36 Conversion of agricultural land has < LU not been extensive in the Reed DC 32 ; Wash and Salt Creek Basins during Z < 28 1973-96. Some of the area around sHI ~ Fruita in the lower Big Salt Wash g 24 Z Basin has been affected by urban­ z ization, especially in the 1990's, ^ 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 WATER YEAR and some agricultural land in the Figure 3. Annual mean streamflow for USGS gaging station on upper Reed Wash Leach Creek Basin was converted (site RW2in fig. 1), water year 1976-96. into Reed Wash throughout the References Cited Liebermann, T.D., Middelburg, R.F., and year. Subsequently, annual stream- Irvine, S.A., 1987, User's manual for flow and dissolved-solids loads Bureau of Reclamation, 1978, Stage One estimation of dissolved-solids con­ have decreased in Reed Wash. development, Grand Valley Unit, centrations and loads in surface The dissolved-solids loads definite plan report Appendix B, water: U.S. Geological Survey presented in this fact sheet are con­ hydrosalinity, land resource, eco­ Water-Resources Investigations Report 86-4124, 51 p. sidered estimates. Assuming the nomics: Salt Lake City, five num­ bered sections plus attachments. errors in streamflow measurements U.S. Department of the Interior, 1995, and water-quality data and errors in __1985, Grand Valley Unit, Stage Two Quality of water Colorado River the computation method are rela­ development, final environmental Basin: Salt Lake City, Bureau of tively consistent for the study impact statement: Bureau of Recla­ Reclamation, Upper Colorado period, then the load differences mation Report INT FES86-10, Region, Progress Report 17, 96 p., between the 1973-83 and 1991-96 217 p. plus appendix. study periods should be reasonable Butler, D.L., 1996, Trend analysis of U.S. Geological Survey, 1974-83, Water approximations of changes in dis­ selected water-quality data associ­ resources data, Colorado, water solved-solids loads. Also, the ated with salinity-control projects in years 1974-83, v. 2, Colorado decreases in dissolved-solids loads the Grand Valley, in the lower Gun- River Basin above Dolores River: reported for the four streams can­ nison River Basin, and at Meeker U.S. Geological Survey Water- not be interpreted as a direct mea­ Dome, western Colorado: U.S. Geo­ Data Reports CO-74-2 through sure of the actual salinity logical Survey Water-Resources CO-83-2. (dissolved-solids) decreases Investigations Report 95-4272, 38 p. attained by Federal salinity-control David L. Butler projects in those four basins. There Butler, D.L., Wright, W.G., Stewart, K.C., Osmundson, B.C., Krueger, R.P., could have been some long-term Water Resources Division, and Crabtree, D.W., 1996, Detailed Grand Junction, Colorado changes in crops, water use, and study of selenium and other constitu­ irrigated acreage in these areas that ents in water, bottom sediment, were not related to the salinity- soil, alfalfa, and biota associated control program. However, results with irrigation drainage in the of this study indicate that, since the Uncompahgre Project area and in the implementation of salinity-control Grand Valley, west-central Colorado, projects, dissolved-solids loads 1991-93: U.S. Geological Survey have decreased in four streams that Water-Resources Investigations drain irrigated agricultural areas in Report 96-4138, 136 p. the Grand Valley.

For more information about this study, write to: Subdistrict Chief U.S. Geological Survey 764 Horizon Drive, Room 125 Grand Junction, CO 81506 email:[email protected]

January 1998 Fact Sheet FS-159-97