Page images

Study Design and Methods

Below its Sierra Nevada headwaters, the San Joaquin River extends 192 miles from Friant Dam in the foothills, to Vernalis, just upstream from backwater influence of the Sacramento-San Joaquin Delta (fig. 20). The first 65 miles of river between Friant Dam and Mendota generally has intermittent flow, and often no river water at all reaches Mendota Pool near Mendota. Most of the next 67 miles of river between Mendota and Stevinson is also intermittent. Flow in the remaining 60 miles of river from Stevinson to Vernalis is perennial and increases downstream as Salt and Mud Sloughs enter from the west, the Merced, Tuolumne, and Stanislaus Rivers enter from the east, and smaller tributaries, irrigation-return flows, and ground water enter along the entire reach. This study focused on the San Joaquin River between site 1 near Stevinson and site 11 at Vernalis.

During the study period, streamflow at Vernalis was below normal for October to January. The average of the monthly mean flows for October 1985 to January 1986 was in the lower 20 percent of flows for the same months for 1970–85. The

monthly mean flow at Vernalis was 2,126 cubic feet per second (ft/s) in October, 1,892 fto/s in November, 2,125 ft/s in December, and 1,935 ft /s in January.

Samples were collected twice monthly at each of the 11 study sites (fig. 20) for analysis of dissolved and total-recoverable selenium, as well as many other constituents. At each site, continuously recorded measurements were used to compute daily mean streamflow throughout the study period.

Selenium sources and transport were evaluated from selenium loads computed for each site and sampling. The method used to compute selenium load at each site is based on the concept of evaluating water at each site that will arrive at the farthest downstream site (Vernalis) at the same time as water from all other sites. Ideally, if the time of travel for water from each site to Vernalis were known, then samples could be collected at each site the same amount of time before the sample was collected at Vernalis. However, logistical constraints prevented this from being accomplished on a regular basis. Instead, the order of site sampling within each sampling week was generally matched with the order of travel times. Travel times were

[ocr errors][ocr errors][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][ocr errors][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][ocr errors][ocr errors][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small]
[ocr errors]



Table 2. Site names and summary of dissolved-selenium concentrations

Dissolved selenium concentration (ug/L) Site No. Site name

Minimum Median Maximum San Joaquin River near Stevinson


2.3 Salt Slough at State Highway 165

5.2 San Joaquin River at Fremont Ford Bridge

9.9 Mud Slough near Gustine


Merced River near Stevinson
San Joaquin River near Newman

San Joaquin River near Patterson
Tuolumne River at Modesto
San Joaquin River at Maze Road Bridge
Stanislaus River at Ripon

.2 11 San Joaquin River near Vernalis


[ocr errors]

Ecco woser A CON

[ocr errors]
[ocr errors]
[ocr errors]

the median percentage was used to characterize the relative importance of each site.

Concentrations of Selenium

estimated from measured velocities and distances between sites. Daily mean streamflow was determined for each site at the time necessary to result in the simultaneous arrival at Vernalis of water from all sites. We assumed that the selenium concentration was constant at each site during the typical 1- to 2-day period of time between the actual sampling date and the date on which the flow was measured.

The dissolved-selenium load was computed at each site, for each sampling, by multiplying the flow at the time estimated for simultaneous arrival of water from all sites at Vernalis by the concentration measured in the sample collected the same week. By estimating the load in this way, the sum of selenium loads at each site on primary tributaries to Vernalis (sites 1, 2, 4, 5, 8, and 10) is expected to equal the selenium load computed for Vernalis, if there are no gains or losses of selenium between these sites and Vernalis. (Similarly, loads from any other combination of upstream and downstream sites would be additive if there are no within-reach gains or losses.) In reality, analytical errors, errors in travel-time estimates, and natural variability will prevent perfect agreement between sites even when there are no within-reach changes. Therefore, such calculations are best used to identify broad patterns of gains and losses and to estimate only the larger within-reach changes. Consistent with these applications, median values of flows and loads relative to Vernalis are emphasized for characterizing sources and transport during the study period. Streamflow and selenium load for each site and sampling were divided by the corresponding value at Vernalis and multiplied by 100 to convert to percentage, and then

Selenium in samples from all sites generally was in dissolved forms, and there was usually no measurable amount in particulate forms. Concentrations of dissolved selenium were highest in Salt and Mud Sloughs (sites 2 and 4), and in the San Joaquin River downstream of Salt Slough (site 3) (table 2, fig. 20). Among these three sites, median concentrations ranged from 3.6 to 5.2 ug/L with a minimum of 0.9 ug/L and a maximum of 16 ug/L. Concentrations were lowest in the three eastside tributaries, the Merced (site 5), Tuolumne (site 8), and Stanislaus (site 10) Rivers. The median dissolved-selenium concentrations in all three of these streams, which contribute most of the San Joaquin River streamflow during low flow, was 0.1 ug/L. The minimum was less than 0.1 ug/L and the maximum was 0.2 ug/L for all three sites. Selenium concentrations in the San Joaquin River at sites 6, 7, 9, and 11 decrease downstream of the Merced River as the selenium from Mud and Salt Sloughs is diluted by low-selenium water from the eastside tributaries. At Vernalis, the site farthest downstream, the median selenium concentration was 0.8 ug/L, with a minimum of 0.3 ug/L and a maximum of 1.4 ug/L.

Sources and Transport

The sources and transport of selenium in the San Joaquin River depend on streamflow and dissolved-selenium concen

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][ocr errors][merged small][merged small][ocr errors][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][ocr errors][merged small]

tration, which determine selenium load. Figure 21 shows streamflow and selenium load at each site expressed as a median percentage of streamflow and load at Vernalis, site 11. Some key relations for the study period shown by these data are given below: • Most selenium loading to the San Joaquin River comes from Salt and Mud Sloughs. Salt Slough contributes about 55 percent and Mud Slough about 17 percent of the Vernalis selenium load, even though their combined flow was only about 9 percent of the Vernalis flow; • By the time flow reaches site 6, downstream of Salt and Mud Sloughs and other possible smaller sources of selenium, 88

percent of the selenium load at Vernalis is already present in the San Joaquin River. This selenium seems to be transported from site 6 to Vernalis with little or no within-stream loss along the way; • The Merced, Tuolumne, and Stanislaus Rivers contribute about 57 percent of the Vernalis flow but only 7 percent of the selenium load at Vernalis; • Within-reach gains of flow are substantial, particularly between sites 6 and 9. In this area, irrigation-returnflows and groundwater inflow appear to be substantial; and • Within-reach gains of selenium downstream from the Merced River are small, as shown by the median values in figure 21.


Salt and Mud Sloughs were the main sources of selenium transported to the San Joaquin River from October 1985 to January 1986, even though they contributed less than 10 percent of the streamflow at Vernalis. Water flowing into the river downstream of Salt and Mud Sloughs

diluted selenium in the river from a median of about 5 ug/L between the sloughs to less than 1 ug/L at Vernalis. Effective management of selenium concentrations in the San Joaquin River depends on managing sources of both selenium and dilution water, which will require detailed information on seasonal patterns and annual variability in streamflow and on the loading and transport of selenium.

Effects of Glen Canyon Dam Operations on Sand Bars and Rapids of the Colorado River in Grand Canyon National Park, Arizona

The USGS is carrying out a program of sediment studies, in cooperation with the U.S. Bureau of Reclamation and the National Park Service, designed to determine the effects of flow regulation at Glen Canyon Dam on the sediment-related resources of Grand Canyon National Park. Studies have defined geometric and hydraulic characteristics at major rapids, provided insight into the effects of debris flows in ungaged tributaries on sand supply and on changes in river form, led to a description of change in sand bars during various streamflow conditions, and provided an understanding of the controls on sand transport and storage in the main channel.

Although sand bars (which are used as campsites) are less susceptible to erosion than are other types of deposits, all sand deposits in narrow reaches are very susceptible to erosion by high flows. Powerplant releases are generally less than 31,500 cubic feet per second; frequent flows higher than these will deplete the sand stored in the main

channel and will result in significant loss of sand from the bars. Sand bars proba remain relatively stable as long as the flows remain within the powerplant-release range. Rapids, which are formed by coarse debris, change with the introduction of new debris and with the adjustment of older debris due to river flow. Therefore, rapids may become more difficult to navigate, because flows much greater than powerplant releases may be required to move the coarse debris that has been deposited by tributaries since construction of the dam.

These studies provide new insights into the behavior of both coarse and fine sediment in rivers in which much of the bed and banks are composed of boulders and bedrock. Results of these studies, together with results of studies of terrestrial and aquatic biology and recreation, will form a report that will be used as the basis for a Secretarial decision concerning future operations of the Glen Canyon Dam.

Information Systems Division

Highlights Communications By Jim Hott and Dan Devereaux

and mission-related responsibilities. GEONET permits the Survey to save $1 million annually on data communications services, while providing a high degree of computer communications capability for the Survey and for other bureaus of the Department of the Interior. As more non-Survey activities use the network, the overall costs to the Survey, both for individual characters transmitted and total data transmitted, are declining.

GEONET also supports the electronic transfer of earth science information from the Earth Science Data Directory and Earth Science Information Network to various public sources. GEONET support also includes electronic transfer of information among other agencies, such as the National Oceanographic and Atmospheric Administration and the National Aeronautics and Space Administration, and various State agencies.

Two major communications projects continued to grow during fiscal year 1987, in support of the U.S. Geological Survey mission: a local-area network, GEOLAN, began operation at Reston headquarters and the GEONET wide-area network came into use on a nationwide and departmentwide basis.

GEOLAN was developed to promote standardized data bases and to share information resources within USGS headquarters through communication between computers and computer systems. GEOLAN permits high-speed local area connections between terminals and between connected information resources such as minicomputers. This technology permits independent heterogeneous local-area networks (LAN) that are part of GEOLAN to be combined via the common foundation of the Ethernet LAN protocol. As compatibility standards for LANs are refined and implemented, the USGS will be able to use a common electronic communicating vehicle to the maximum benefit in accomplishing its critical missions.

GEONET has replaced a number of services that formerly connected users in two locations. Now, users not only may use GEONET to communicate with a single destination site but may reach more than 140 Department of the Interior host computers of various sizes, plus computers attached to various public networks. GEONET has delivered to USGS users an unprecedented capability that can be used to accomplish everyday data processing

« PreviousContinue »