The North Santiam River Basin drains approximately 690 mi2 upstream from the City of Salem's water-treatment facility. The basin includes Detroit Lake, a controlled reservoir with 436,000 acre-ft of storage capacity. Big Cliff, a small reservoir with 2,430 acre-feet of usable storage capacity, is used to stabilize water releases from Detroit Dam. Besides providing recreation and flood control, both dams are used for power generation. Detroit and Big Cliff Dams have 100,000 and 18,000 kilowatt powerhouses, respectively, neither of which has capabilities for selective withdrawal at varying lake depths.
The City of Salem's water-treatment facility uses a slow sand filtration system to supply water to approximately 155,000 customers in the Salem metropolitan area. The sand and gravel layers of the filter flocculate inorganic clay and larger particles. Protozoa, algae, and other invertebrates on the filter surface form a mat that helps to remove biological and other organic contaminants. In the next 20 years, the Salem-area water-service population is projected to grow to 230,000; at peak water demand, including a 10 percent reduction for conservation, this population would require approximately 90 million gallons per day (City of Salem, 1999).
The floods of 1996-97 caused increases in turbidity that persisted for several months and surpassed the ability of the treatment plant to filter the water, subsequently disrupting normal delivery of water to the Salem area for about a month. During the February 1996 flood, 8-15 inches of precipitation fell in a 4-day period, shutting-down Salem's water treatment facility for 8 days, forcing operations to acquire about 4 million gallons of water per day from the City of Keizer (Cotton et al., 1998). Since 1996, Salem's water intakes also have been closed for as many as 10 high-flow episodes at up to 48 hours per closure due to storm-related, silt-laden water. In 1997 a pretreatment facility was constructed to process highly turbid water, although its use results ina significant increase in operating and personnel costs.
During the February 6, 1996, storm event, turbidity values measured on the North Santiam River at the water-supply intake began to rise and eventually peaked on February 15 near 140 NTU (Nephelometric Turbidity Units), declining to 10 NTU by March 10 (Ruffing et al., 1997). A slow sand-filtration system cannot treat water with turbidity values higher than 10 NTU. The highest turbidity values in the North Santiam River Basin during this storm were recorded on February 14-15 in Detroit Lake, ranging from 55 NTU at the surface to 389 NTU at a depth of 257 ft. In contrast, on February 14-15 at 0.5 miles upstream from Detroit Lake on the North Santiam River at river mile 70, turbidity ranged from 8 to 21 NTU; on the Little North Santiam River near the mouth at Mehama, the turbidity ranged from 12 to 21 NTU. On February 21, Detroit Lake readings varied from 30 NTU at the surface to 388 NTU at a depth of 290 ft, no readings were taken from the tributaries on that day.
There is a lack of streamflow, turbidity, and suspended-sediment data from the main-stem North Santiam River and its tributary inflows to Detroit Lake and below, thus making management decisions regarding water supply to the City of Salem difficult. Reductions in water clarity in this basin, primarily caused by altered geomorphic features or land disturbances, may result in a revision of water-quality standards for the North Santiam River watershed. The State of Oregon Administrative Rules specify that a greater than 10 percent increase in turbidity for all streams, as measured from a control point upstream of the turbidity causing activity, is not allowed (OAR 340-41-445(2)(c)). Further, water suppliers are legally required to monitor turbidity and to provide their users with treated water having average turbidity levels at or below 1 NTU, with no exceedance above 5 NTU allowed (OAR 333-061-0030 (3) (B)). In the Umatilla River Basin, the Oregon Department of Environmental Quality (ODEQ) has developed a Water Quality Management Plan with Total Maximum Daily Loads directly linking turbidity to instream suspended sediment. In determining load calculations for the Umatilla River basin, ODEQ stated that a turbidity of 30 NTU (not be exceeded for a duration of more than 48 hours) was protective of aquatic life (Oregon Department of Environmental Quality, 2000). In addition, the U.S. Environmental Protection Agency ambient water-quality criteria (U.S. Environmental Protection Agency, 2000) recommend a turbidity reference condition in the Cascades Ecoregion of 0.25 NTU; they urge States to determine their own reference sites for rivers and streams within each ecoregion, at different geographic scales, and to compare them with EPA's reference conditions .
Some studies have suggested using turbidity as a surrogate for suspended-sediment concentrations (SSC), total suspended solids, and soil loss (Christensen et al., 2000; Costa, 1977; Truhlar, 1976). Background data from the North Santiam River Basin on real-time, storm-driven turbidity fluctuations causing increased long-term (persistent) turbidity were not available prior to establishing the USGS water-quality monitoring network in 1998. The turbidity measurements taken during the February 1996 flood were point readings made at differing times, not continuous data showing the peak turbidity during the event or any trend through the flood hydrograph. Because of the event-dependent nature of erosion and sediment-transport processes, the need to monitor real-time turbidity and other water-quality constituents in the North Santiam River Basin is critical to both effectively manage downstream water-treatment facilities and inform water-quality managers in the upper basin. A direct correlation between turbidity and suspended-sediment concentration has been documented in other studies (Beschta, 1980; Kunkle and Comer, 1971) and has been observed from data collected at sites within the North Santiam River Basin (fig. 1). This turbidity-SSC relationship might be one method land and water resource managers could use to estimate sediment loads to water treatment systems and assess overall basin response to peak-flow and storm events.
The North Santiam River Basin is characterized by a temperate marine climate, normally with dry summers and wet winters. Streamflow is primarily precipitation driven, although in the upper basin some perennial snow fields and springs help sustain spring and summer flows. Hot springs occur in the basin, but flow from these is insignificant compared to the total streamflow in the North Santiam River. Approximately 80 percent of the precipitation falls between October and May, with mean annual rainfall of from 45 inches at Jefferson, located 15 miles south of Salem on the Santiam River (RM 10), to 75 inches at Detroit, with extremes approaching 200 inches at the crest of the Cascades (Laenen and Hansen, 1985). The annual high flow from 1908-1987, for a 1-day period with a 50-year recurrence interval, is 15,000 cfs (cubic feet per second) at North Santiam River above Boulder Creek near Mehama (USGS Station ID 14178000) (Wellman et al, 1993).
The North Santiam River Basin is bounded by the Cascades Range to the east and the Willamette Valley to the west. The basin drains west. The eastern boundary extends from Olallie Butte in the north to Three Fingered Jack mountain in the south. Elevations along the 25-mile eastern edge surpass 8,000 feet, with Mt. Jefferson the highest at 10,495 feet. Thirty-five percent of the North Santiam River Basin is composed of older volcanics and tuffaceous sedimentary rocks, and approximately 30 percent is younger basalts. (Sherrod and Smith, 2000; fig. 2). The older, weathered basalts and volcaniclastic rocks may be responsible for the colloidal suspensions and resulting high turbiditities in Detroit Lake.
Studies have identified assemblages of colloidal minerals in both the water column and in alluvial fan and delta areas of Western Cascade streams and reservoirs. Smectite and other amorphous clays have been identified as the primary minerals responsible for the persistent turbidity in the North Santiam and other Western Cascade basins (Ambers, 1998; Glasmann, 1998; Pearch, 2000). The upper basin is composed principally of High Cascade volcanics of the Pliocene age, with timber harvesting as the primary land use activity. The lower basin, below Mehama, is a large alluvial plain, except where volcanic and marine sedimentary rocks of the Eocene and Pliocene age are exposed in the foothills. Agriculture is the primary land use activity in the lower basin.
The river reach above Detroit Lake (RM 59) is a steep-channeled pool and riffle system having stream gradients greater than 80 ft/mi. Erosion potential from the banks is high, particularly where mature vegetation is not present, causing landslides and mass wasting during extreme precipitation events. The middle reach, from Detroit Dam (RM 49) to Mehama (RM 27), generally is in a canyon, with a stream width of about 150 ft and an average gradient of 30 ft/mi. The lower reach, from Mehama to Stayton (RM 17), flows through an alluvial valley at an average width of 225 ft and an average gradient of 17 ft/mi (the above two paragraphs are from Laenen and Hansen, 1985).
Other sources of erosion could be related to land use, as well as to landscape and geomorphic features. For example, land cover upstream from the water treatment plant is approximately 90 percent forest (U.S. Geological Survey, 1990), with timber harvesting the primary industry; associated activities, such as road building and logging operations, may alter storm hydrographs and increase sediment delivery to the basin (Harr et al., 1975; Reid, 1984; Swanson and Dryness, 1975). Approximately 21 percent of U.S. Forest Service lands have been harvested above Detroit Lake (Cotton et al, 1998). In addition, the North Santiam River Basin extends eastward to the crest of the Cascade Mountain Range to elevations above 8,000 feet. These areas are glaciated and prone to landslides, debris flows, mass wasting, and moraine-lake outbursts causing channel alterations that can markedly affect the sediment carrying capacity of the local streams, as well as the suspended-sediment to streamflow relations.
Long-term USGS stream-gaging stations (table 1) include the North Santiam River at Mehama (MEHAMA, site 14183000; period of record 1921-present)1, the North Santiam River at Niagara (NIAGARA, site 14181500, period of record 1938-present), the North Santiam River below Boulder Creek near Detroit (N SANTIAM, 14178000, period of record 1928-present) and the Little North Santiam River near Mehama (LITTLE NORTH, 14182500, period of record 1931-present). All sites provide hourly discharge data; in addition, historic hourly temperature data2 are available for all sites, except Blowout Creek near Detroit (BLOWOUT, 14180300) and French Creek near Detroit (FRENCH, 14179100). The Breitenbush River above French Creek near Detroit site (BREITENBUSH, 14179000) was discontinued; daily streamflow record from this site is available for 1933 through 1987.
The primary goals of this investigation are to:
1. Define the spatial extent of suspended-sediment loads and turbidity in the North Santiam River Basin.
2. Estimate suspended-sediment loads for the major subbasins and main-stem North Santiam River using correlations developed between real-time turbidity readings and SSCs.
3. Identify the relative contribution of turbid water (both short-term and long-term persistent turbidity) from major subbasins and other primary sources to Detroit Lake and the North Santiam River at and above the City of Salem's water intake.
4. Establish a network of real-time streamflow and water-quality stations to monitor short-term (daily, monthly, storm-to-storm, and seasonal) and long-term (year-to-year and decadal) temporal changes in stream discharge, stream temperature, specific conductance, pH, and turbidity.
5. Establish an early warning system to monitor high streamflow and turbidity events in the North Santiam River Basin that may impact operation of the City of Salem's water treatment plant.
Streamflow and water-quality data collection activities and publication of interpretive reports are planned for October 1998 through September 2003 (5 years), although this may be extended. The persistent turbidity and suspended-sediment data collection started October 1998 and will continue through September 2002, although the persistent-turbidity data collection will probably end by June 2001, as high turbidity events are not common from July to September and because at least 15 persistent-turbidity samples per site will have been collected by June.
The work will address several priority water-resource issues identified in the U.S. Geological Survey Water Resources Division strategic plan for 1998-2008, (http://water.usgs.gov/pubs/ofr/ofr99-249/html/exec.html).
The study includes both data collection and interpretive components that meet the USGS mission and strategic plan:
1. The network of North Santiam water-quality sites meets the long-term data collection goals of the USGS by adding three streamflow sites and four water-quality parameters to each site in the total streamflow network of seven sites, thereby enhancing the spatial and temporal data coverage for the basin.
2. The additional real-time streamflow sites will help minimize loss of life and property by supplying timely streamflow data for flood response and emergency decision making.
4. The turbidity and suspended-sediment correlations will provide insight into source areas of sediment that could be improved or protected in order to reduce fluvial sediment loads and allow fuller use of the water resources of the basin. Subbasins could then receive restoration and erosion abatement measures to enhance water clarity and quality. For example, land-use activities in subbasins with high erosion and sediment yields could be reviewed and modified to reduce sediment flux and subsequent increases in turbidity.
4. The North Santiam water-quality data directly relates to human and aquatic health issues, as the measured constituents may affect the drinking water for the City of Salem area and the endangered species present in the basin.
Streamflow, turbidity, and suspended-sediment concentration and load data will help provide a basis to better understand the overall health and status of the North Santiam watershed and could be used to determine long-term water-quality trends. In addition, because turbidity measurements can be used to estimate suspended-sediment concentrations (Kunkle and Comer, 1971), the turbidity monitoring network, as designed in this investigation, could provide data for estimating suspended-sediment loads, both into and out of Detroit Lake, which is of interest to the U.S. Army Corps of Engineers in the use of Detroit Lake Dam. Finally, the installation of telemetered water-quality probes to provide real-time turbidity data will facilitate the effective operation of the City of Salem water-treatment plant.
The expected audience both interested in and benefitting from this data and findings is widespread. A consortium of Federal, State, and local agencies and the academic community have been involved in the development, design, and implementation of this study and anticipate study data and results. Finally, the general public and watershed groups, specifically in the population using and drinking North Santiam River water and/or living within the North Santiam watershed, currently monitors the real-time data provided by this project.
The approach used in this study can be divided into six principle tasks:
1. Continued streamflow measurement and operation of real-time water-quality monitors
2. Cross-section measurements and datasonde calibration
3. Suspended-sediment data collection
4. Persistent turbidity and storm sampling
Operation of telemetry and data-logging instrumentation for stream level and water-quality measurements will continue at the seven North Santiam River Basin sites (table 1). Four sites are located above Detroit Lake (one on the main-stem North Santiam River [N SANTIAM] and three tributaries [BREITENBUSH, BLOWOUT, FRENCH]) and three below the lake (two on the main stem [NIAGARA, MEHAMA] and one on a tributary [LITTLE NORTH]); all are above the water treatment plant. Daily streamflow, water temperature, conductivity, pH, and turbidity data will be reviewed and published in USGS annual data reports. Real-time data is available through either the North Santiam Turbidity Project Web site or the U.S. Army Corps of Engineers (COE) interactive database query page.
For all sites in this study, real-time station streamflow and water-quality data are available in 30 minute increments. The following parameters are being collected at the seven sites (table 2):
|
Cross-sectional |
||||||
|---|---|---|---|---|---|---|
|
Sept. 19214 |
||||||
|
LITTLE NORTH5 |
||||||
|
Oct. 19386 |
||||||
|
BLOWOUT7 |
||||||
|
BREITENBUSH8 |
Oct. 19989 |
|||||
|
Oct. 192810 |
||||||
All sites have separate stream-level and YSI11 multi-parameter water-quality instrumentation installed. All sites are telemetered via direct phone lines, except the Blowout Creek station, which has a cellular system. Data are logged every 30 minutes. The direct-line sites are called at 3-hour intervals (8 times per day). The Blowout Creek station requires more power to operate the cellular system, so it is called less often to maintain battery reserves at 6 times per day in 4-hour intervals. Streamflow and water-quality data for all six sites are being collected, entered, and stored in the USGS Automated Data Processing System (ADAPS) and National Water Information System (NWIS) database. The four-parameter daily water-quality data will be published for all seven sites as daily maximum, minimum, and mean values. A depth sensor is also available on the YSI datasonde for all sites, except BLOWOUT, which supplies back-up stage data. Campbell CR10X data loggers, powered by deep-cycle batteries and kept charged by solar panels, record all station parameters for downloading to USGS computers connected through the telemetry. The YSI datasonde also operates as a data logger, providing redundancy to the Campbell unit. All equipment and instrumentation is serviced on a regular basis.
In addition, the City of Salem has installed a Hach 1720D flow-through turbidity meter in Big Cliff Dam power station, located approximately 0.8 miles upstream of the North Santiam River at the Niagara site (NIAGARA). A Campbell CR10X data logger is connected to it to record turbidity readings at 30 minute intervals. The USGS maintains and calibrates this turbidity monitor every 1-2 months and downloads the data. The City's monitor provides backup turbidity data to the station probe at NIAGARA and can be used for data verification or estimating turbidity when records are missing at NIAGARA. The turbidity data from Big Cliff Dam is not published, but has been correlated to within +/- 10 percent of turbidity recorded at NIAGARA.
Lastly, a telemetered tipping-bucket precipitation gage was installed at BLOWOUT in March 2001. Real-time precipitation data is available from BLOWOUT in both 30 minute increments and as daily totals.
Cross-sectional measurements generally will accompany each calibration visit, which includes attaining readings with a separate "backup datasonde" from at least three points in the cross-section. The USGS will perform all calibration work. Periodic depth-integrated turbidity samples also will be collected and composited from points along the cross-section and from dip samples near the station datasonde location. Turbidity in these samples is measured onsite using a Hach 2100P turbidimeter and compared to readings from the station datasonde during the calibration process. If the turbidity value is greater than 5 NTU (OAR 333-061-0030 (3)(B) exceedance limit) the sample is transported to the USGS lab, refrigerated, and tested for persistent turbidity (see Section 4 below for persistent-turbidity procedures), otherwise the sample is discarded after the turbidity reading. All calibration information and notes are entered on separate station water-quality monitor field forms. Corrections are applied as appropriate and used to compile the daily water-quality records. Individual malfunctioning probes are repaired or replaced as needed. Maintenance and calibration of the multiparameter datasondes is expected to extend at least 5 years.
As explained above, the multiparameter monitors supply real-time turbidity data at 30-minute intervals, providing 17,520 readings per year for analysis. Because these 30 minute turbidity readings have shown a close correlation to suspended sediment concentration in the North Santiam River Basin (fig. 1), a periodic sampling regime has been designed to collect suspended-sediment samples both routinely and during storm events at all seven sites. This sampling effort will develop linear regression equations (Gunst and Mason, 1980) between turbidity and suspended-sediment concentration. When used in conjunction with continuous discharge, the estimated SSC's will be used to compute estimates of daily and annual suspended-sediment loads. This load data will aid the City of Salem in the operation and maintenance of the water treatment plant by providing predictive capabilities in estimating the sediment load entering their intakes, settling ponds, and pretreatment plant. Sediment load information also can be used to estimate possible reservoir siltation (Ewing and Mohrman, 1989, Gippel, 1989, Sidle and Campbell, 1985). A 1953 report by the U.S. Army Corp's of Engineer's estimated that the recently constructed Detroit Reservoir would fill with sediment at the rate of 175 acre-feet per year (U.S. Army Corps of Engineers, 1953). The estimated suspended-sediment load data from this project could help verify these shoaling amounts.
Fine clay particles have been shown to proliferate in the upper tributary watersheds and in the water column of Detroit Lake and delta areas (Bates et al., 1998). To better estimate the larger (heavier) to smaller (lighter) portion of the suspended-sediment load, and therefore the faster to slower particle settling ratio, the suspended-sediment results include a sand/silt partitioning. This process separates particles smaller than 62 micrometers nominal diameter (clay and silt) from those that are larger than or equal to 62 micrometers (sand and larger particles) (Colby, 1963). Coarse to fine separation also may help to explain fluvial sediment transport dynamics and could pinpoint sources of high and persistent turbidity in specific subbasins in the North Santiam tributary-reservoir system. Samples will be analyzed for sediment concentration in milligrams per liter and sand/silt partitioning when sand is present. The clay fraction will be determined for one or two samples per site per year, or more if sufficient fine particles are present. Data are provided in millimeter sieve sizes of .031, .016, .008, .004, .002, and less than .002. Results from these samples will help characterize types and amounts of colloidal material for each subbasin and provide a correlation to the persistent turbidity information.
Duplicate samples will be collected and analyzed for quality assurance and to check results of the sediment lab. The suspended-sediment sampling may either coincide with the datasonde calibration, or occur as a separate trip. Suspended-sediment load estimates are tabulated in spreadsheet format using simple linear regression to determine correlations. The 30-minute-turbidity unit values will be used in the individual station regression equations to derive unit values of sediment concentration. Sediment load estimates will then be computed in 30-minute intervals using streamflow unit values (17,520 SSC values per year). The resulting 48 estimates per day will be averaged and provided as the estimated mean daily suspended-sediment load.
In conjunction with suspended-sediment sampling, persistent-turbidity samples will be collected at all seven sites during periods of high turbidity and/or streamflow. This will entail collecting back-to-back near-identical depth-integrated samples along with the suspended-sediment samples at all seven sites. Because the persistent-turbidity tests use a smaller, glass, 1-quart jar, instead of the 3-liter containers, less sample is collected; hence four to six EWI vertical point samples from the cross-section are all that is required to sufficiently fill the jar. Fewer sampling-station points are possible because lateral measurements of turbidity taken through the cross-sectional at all seven sites, under all flow conditions, have shown turbidity to vary by less than 5 NTU for 95 percent of the readings12. This variance is less at lower flows and hence is acceptable for our persistent-turbidity samples, since the initial turbidity (or first reading of turbidity before the particles settle) should be 5 NTU or higher (see below for procedures). Initial sample turbidity is measured in the field at the time of collection, using a Hach 2100P portable turbidimeter, and repeated in the lab as the first reading of the persistent-turbidity test. As mentioned in Section II, if the turbidity reading is under 5 NTU, the value is recorded in the field notes and the sample is usually not processed for persistent turbidity.
The entire sample is collected or composited in a 1-quart glass jar. A glass container was selected for optimum clay-particle settling because there is less cohesion to the container walls. The 1-quart size also provides for better transport, refrigerator storage, and testing. Persistent-turbidity data collection and analysis will be completed in fiscal year 2001. The USGS lab in Portland will complete the persistent-turbidity work, which entails measuring turbidity over time as the sediment settles under quiescent conditions. Analysis of the persistent-turbidity data will be presented in an interpretive report.
The method used to determine persistent turbidity of fine sediments is similar to the pipet method outlined by Guy (1969) for particle-size analysis, except that dispersion agents and mechanical agitation are not used and the settling medium is native water. Persistent turbidity is measured as follows: Samples are refrigerated at 4-6° C13, removed, and gently shaken for 1-2 minutes to re-suspend the settled material, depending on the material present, and then a 10-11 ml aliquot is withdrawn by use of a wide-mouth pipette14 at 2.75 cm below the sample surface in the container. An apparatus both holds the pipette vertically level and provides uniform lowering through the center of the bottle opening. Four to five turbidity readings are taken from the same aliquot and averaged for each withdrawal. The sample is then returned to 4-6° C refrigeration, being careful not to re-agitate the sample, and prepared for the next set of readings, taken at intervals of approximately 30 seconds, 30 minutes, 2 hours, 4 hours, 8 hours and 28 hours (table 3). These time intervals coincide with the fall velocities for defined silt- to clay-sized particles as computed from Stokes's law using the 2.75 cm fall distance in the sample container (see Equation 1 below). The 2.75 cm aliquot depth was selected based on the total sample depth and volume, along with the actual height of the 1-quart sample container.
Fall time (in sec) = (0.1113) (viscosity at sample temp, in °C) (fall distance, in mm)/(diameter of spherical particle, in mm)2
If the turbidity value remains above 5 NTU, then additional readings are obtained until the readings drop below 5 NTU. The general time frame for the approximate 10 storm samples is designed for (1) one to two fall samples (first flush), (2) five to six winter storm samples, and (3) three to four spring runoff samples, with additional samples collected during high flow periods. Duplicate samples will be collected for quality assurance and checks.
Turbidity versus time curves will be developed for all sites using storm-event samples or from sites with turbidity sample readings greater than 5 NTU, although colloidal particles in the North Santiam system may be so small as to not settle according to Stoke's Law. Environmental factors such as velocity currents, wind, wave action, and solar heating also will alter the viscosity and quiescent settling conditions of fluvial sediment particles in a water column, causing particles in the river systems and water bodies, such as Detroit Lake, to remain in suspension longer than theoretically calculated (Rinella and McKenzie, 1982).
Other studies have shown that a correlation exists between persistent turbidity and the clay-sized fraction (less than 0.002 mm) of the suspended sediment over a range of concentrations (Curtiss, 1982; Loper and Wetzel, 1988). Using Stoke's law, the amount of time for different-sized sediment particles to fall 2.75 cm can be determined. As mentioned above, aliquots are timed for withdrawal at specific sediment-size classes (Vanoni, 1975) using table 3. These fall times can also be converted to approximate fall times in Detroit Lake assuming quiescent settling conditions. Mean winter depth to the Detroit Lake penstock outlets is 120 ft.
By comparing the settling times of the different turbidity samples from various samplings, sites and subbasins with the highest persistent turbidity can be determined. That is, the longer the settling time, the longer the persistent turbidity and higher the amount of clay-sized particles that would occur in a particular subbasin. Since settling times are directly correlated to specific sediment sizes, then each clay-particle size will correspond to an approximate turbidity in time during the testing sequence. It may be possible to compute clay loads or volumes of persistent turbidity using these correlations and the corresponding streamflow. Another method would use rates of change. That is, since particles settle at different rates with some samples having more particles than others, these rates of change can be compared and computed as change in turbidity per unit of time. The sample is assigned a rate of decrease that would equate higher rates of change per unit of time with higher persistent turbidity, when compared with other samples collected during the same sampling period.
The February 1996 flood caused turbidity to persist for several months, with raw turbidity values of 10 NTU recorded into the summer at the intakes of the water-treatment plant. By using table 3 and the February 1996 data we can define persistent turbidity in Detroit lake as the time it takes .002 mm size particles and smaller to settle the 120 ft to the Detroit Lake penstock outlet port, approximately 6 months or longer. A third analysis method would use turbidity decay curves from the laboratory tests, tabulated so that the lab turbidity value for the .002 mm size particle and smaller decay time would represent the persistent turbidity for that site. Persistent turbidity values from each site and storm event can then be compared.
For major peak storm events only, concurrent samples will also be collected from Detroit Reservoir by the U.S. Army Corps of Engineers (USACE). Three locations in Detroit Reservoir are identified for sampling: (1) close to Detroit Dam near the log boom, (2) near the Blowout Creek tributary inflow, and (3) upstream and adjacent to Piety Knob Island near the tributary inflows of the North Santiam River, Breitenbush River, and French Creek. The USACE will acquire latitude/longitude locations and vertical water-quality profiles of the reservoir at these three sampling points. The water-quality parameters will include water temperature, conductivity, pH, dissolved oxygen concentration, percent dissolved oxygen saturation, and turbidity. Samples for persistent turbidity will be collected using a VanDorn sampler. Sampling depths generally are at three points in the vertical profile: near the surface, at mid-depth, and close to the lake bottom. Samples will be delivered to the USGS office in Portland. The USGS will notify the USACE when storm sampling has begun, and the USACE will follow 1-3 days later with sampling in Detroit Reservoir. Because of sampling logistics and resources, only peak storm events will be sampled in the lake. These data are entirely contingent on the USACE collecting the samples. USACE is not funded under this project.
The USACE uses a Hydrolab datasonde for lake water-quality measurements. The USGS uses YSI datasondes at all stations. The turbidity probes are somewhat different in design. The Hydrolab uses a lens shutter, and the YSI uses a lens wiper. Both however, use infrared LED's and are calibrated using the same type standards. Comparison tests between the two instruments will be conducted to ascertain any significant differences.
In addition to the YSI datasondes in the field, the USGS uses a Hach turbidimeter for the persistent turbidity lab tests and onsite checks. The Hach instrument uses a tungsten lamp in a dark, laboratory environment instead of an infrared LED in an open, in situ environment. In addition, the two instruments measure turbidity differently. The YSI unit deploys a single 90° scatter-light detector, whereas the Hach unit deploys a second transmitted-light (pass-through) detector in addition to the 90° scatter-light detector and ratios between the two, based on an algoritm, to acquire a signal output. For these reasons, the readings from these instruments are not directly comparable. The difference in readings can vary by as much as 20 percent, especially during storm events. However, since the persistent-turbidity work uses only the Hach unit, the persistent-turbidity readings are all comparable. The initial Hach reading is considered the whole-water turbidity value; the Hach readings after particle settling provide a measure of persistent turbidity.
As a means of quality assurance, standard USGS methods and protocols will be used for all measurement, sampling, and data-management activities. These include stream-discharge measurements, suspended-sediment sampling and analysis, data management and review, and training of personnel. Quality-control samples for this study will constitute 15 percent of the total number of suspended-sediment samples collected and will consist primarily of replicate samples collected throughout the year at all sites. A subset of the quality-control samples will consist of replicate samples analyzed at different laboratories to evaluate laboratory performance.
Various statistical packages and GIS applications will be used to relate landscape features, such as land use, land cover, hydrogeology, and soils data layers, in a watershed framework, to the suspended-sediment loads and persistent turbidity results. Watershed data of this type might include subbasin percentages of forest types, roads, landslide and earthflow areas, geology, and soil types, along with maps of slope, aspect, precipitation and other geomorphic terrains and hydrologic characteristics (fig. 3). GIS has been used extensively as a water-quality analysis tool in the USGS National Water Quality Assessment program (Uhrich and Wentz, 1997).
Paired basin studies in forested watersheds require establishing a correlation between basins before and after a land use change occurs to one basin, such as logging or road building. The other basin serves as a control basin in an unaltered landscape. A project of similar design and scope will be proposed, starting in 2002, between upper North Santiam subbasin sites (yet to be established) with altered and unaltered terrain and landscape. Ideally, a small forested basin with a complete stand of at least 25 year old timber (control basin) and a small adjacent watershed (altered basin) with some past or proposed alteration, such as in USFS and ODF matrix lands, should be selected. The primary land use activity in the upper North Santiam Basin is timber harvesting and road building. Data from this study will help document the effects these activities have on water-quality, specifically relations between streamflow versus suspended sediment and suspended sediment versus turbidity from unaltered and altered basins of similar terrain and geology (Ponce et al., 1982).
There are several components of the field work and data collection program that require safety considerations. These are in two categories:
1. Sampling and calibration- USGS sampling protocols are followed for all suspended-sediment samples (Edwards and Glysson, 1999). Employee cableway use and sampling training is provided before an employee conducts their first independent field trip. Approved life jackets as described in the USGS Occupational Hazards and Safety Procedures Handbook 445-2-H, Chapter 16; D (http://www.usgs.gov/usgs-manual/handbook/safety/hand16.html) are used during all wading, bridge, and cableway sampling. Cableways and gaging stations are inspected annually. Any required repairs or maintenance is performed before the next use or visit. Lighter-weight samplers are used whenever possible to alleviate back injuries. Heavier samplers, used during major storm event, require two personnel to assemble the equipment and conduct sample collection. During harsh weather and high flow conditions, sampling crews work in pairs of two, usually a crew of two at the three upper sites and another two person crew at the three lower sites. Water-quality standards are transported, handled, and disposed of properly. Station pathways, trails and embankments have been upgraded with stairs, and brush and logs are removed as needed.
2. Vehicles and transportation- All vehicles are kept properly serviced and maintained. A four-wheel drive, 2000 model year Ford Excursion SUV is the primary field vehicle. A steel cage has been installed to separate the passengers from the cargo. If snowmobiles or snowshoes are required for accessing sites, two personnel with separate snowmobiles and/or snowshoes travel together at all times. All vehicles carry cellular phones, first-aid kits, emergency and survival gear, flood lights and an AC/DC converter unit. A call-in procedure is used at the end of the day to verify employee safety and accident-free conditions. Extended working days are kept to a minimum by overnight stays at local lodging establishments. U.S. Forest Service radios are available for most emergency situations in the Willamette National Forest.
Ambers, R.K.R., 1998, Effects of land use on the concentration and clay mineralogy of suspended sediments in streams of the Dorena Lake Watershed, Western Oregon, American Geophysical Union, Abstracts for 1998 Fall Meeting, San Francisco, California: http://earth.agu.org/meetings/waisfm98.html, accessed March 1, 2001.
Austin, R.W.,1973, Problems in measuring turbidity as a water quality parameter; Seminar on Methodology for Monitoring the Marine Environment: Seattle, Washington, 11 p.
Bates, D., Willis, K., Swanson, F., Glasmann, J. R., Halemeier, D., Wujcik, H., 1998, North Santiam River turbidity study, 1996-1997: Berkley, California, Watershed Management Council Networker, University of California, p. 1-17.
Beschta, R.L., 1980, Turbidity and suspended sediment relationships, Proceedings of Watershed Management Symposium, Irrigation and Drainage Division, American Society of Civil Engineers, Boise, Idaho, July 21-23, 1980: p. 271-282.
Christensen, V.G., Xiaodong, J., Ziegler, A.C., 2000, Regression analysis and real-time water-quality monitoring to estimate constituent concentrations, loads, and yields in the Little Arkansas River, South-Central Kansas, 1995-99: U.S. Geological Survey, Water-Resources Investigations Report 00-4126, 36 p.
Colby, B.R., 1963, Fluvial sediments--A summary of source, transportation, deposition, and measurement of sediment discharge: U.S. Geological Survey Bulletin 1181-A, 47 p.
Costa, J.E., 1977, Sediment concentration and duration in stream channels: Journal of Soil and Water Conservation, v. 32, p. 168-170.
Cotton, C.S., Dominicci, A.J., Feldman, D.S., Pilatzke, C'L., 1998, Oregon Watersheds, Many activities contribute to increased turbidity during large storms: Washington, D.C., U.S. Government Accounting Office GAO/RCED-98-220, 69 p.
Curtiss, D.A., 1982, An evaluation of suspended sediment and turbidity in Cow Creek, Oregon: U.S. Geological Survey Water-Resources Investigations Open-File Report 82-364, 21 p.
Edwards, T.K., Glysson, G.D., 1999, Field methods for measurement of fluvial sediment: Techniques of Water-Resources Investigation of the U.S. Geological Survey, Book 3, Chapter C2, 89 p.
Ewing, R. and Mohrman, J., 1989, Suspended sediment and turbidity from Northern Yellowstone Park, Wyoming, 1985-1987, Proceeding of the Symposium on Headwaters Hydrology, American Water Resources Association, Bethesda, Maryland: p. 213-222.
Gippel, C.J., 1989, Use of turbidimeters in suspended sediment research: Hydrobiologia HYDRB 8 Vol. 176/17 p. 465-480.
Glasmann, J.R., 1998, Soil erosion and water quality in the Western Cascades, Oregon, Annual Meeting Abstracts, American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Baltimore, Maryland, October 18-22, 1998: p. 344.
Gunst, R.F. and Mason, R.L., 1980, Regression analysis and its application, A data-oriented approach: New York, Marcel Dekker, Inc., 402 p.
Guy, H.P., 1969, Laboratory theory and methods for sediment analysis: U.S. Geological Survey Techniques of Water-Resources Investigations, Book 5, Chapter C1, 58 p.
Guy, H.P., 1970, Fluvial Sediment Concepts: U.S. Geological Survey Techniques of Water-Resources Investigations, Book 3, Chapter C1, 52 p.
Harr, R.D., Harper, W.C., Krygier, J.T., and Hsieth, F.S., 1975, Changes in storm hydrographs after road building and clear-cutting in the Oregon Coast Range: Water Resources Research, v. 11, p. 436-444.
Koeppen, J., 1974, Turbidity measurement--an overview, in Proceedings of Turbidity Workshop, U.S. National Oceanic Instrumentation Center, Washington, D.C.: p. 11-12.
Kunkle, S.H. and Comer, G.H., 1971, Estimating suspended sediment concentrations in streams by turbidity measurements: Journal of Soil and Water Conservation, v. 36, p. 18-20.
Laenen A., and Hansen, P.R., 1985, Preliminary study of the water-temperature regime of the North Santiam River downstream from Detroit and Big Cliff Dams, Oregon: U.S. Geological Survey Water-Resources Investigations Report 84-4105, 45 p.
LaHuzen, R.G., 1994, Variations in turbidity in streams of the Bull Run watershed, Oregon, Water Years 1989-90: U.S. Geological Survey Water-Resources Investigations Report 93-4045, 28 p.
Loper, C.A. and Wetzel, K.L., 1988, Sediment loads, discharges, and yields in the East Branch Mahoning Creek Basin, Clearfield and Jefferson Counties, Pennsylvania, June 1979 through September 1981: U.S. Geological Survey Water-Resources Investigations Report 85-4213, 26 p.
Oregon Administrative Rules (OAR), 2000, Chapter 340--Department of Environmental Quality, Division 41--Water Pollution, Section 445--Willamette River Basin (2)(c).
Oregon Administrative Rules (OAR), 2001, Chapter 333--Department of Human Services, Oregon Health Division, Division 61--Public Water Systems, Section 30--Maximum Contaminant Levels and Action Levels (3) (B) (i) (ii).
Oregon Department of Environmental Quality, 2000, Umatilla River Basin TMDL and WQMP: http://waterquality.deq.state.or.us/wq/TMDLs/TMDLs.htm, accessed March 1, 2001.
Paustian, S.J., and Beschta, R.L., 1979, The suspended sediment regime of an Oregon Coast Range stream: Water Resources Bulletin, v. 15, p. 144-154.
Pearch, J.M., 2000, Relationships between landscape stability, clay minerology, and stream turbidity in the South Santiam Watershed, Western Cascades, Oregon: Corvallis, Oregon State University, M.S. Thesis, 199 p.
Pickering, R.J., 1976, Measurement of "turbidity" and related characteristics of natural waters: U.S. Geological Survey Open-File Report 76-153, 13 p.
Ponce, S.L., Schindler, D.W., and Averett, R.C., 1982, The use of the paired-basin technique in flow-related wildland water-quality studies: U.S. Forest Service, WSDG Report WSDG-TP-00004, 21 p.
Porterfield, G., 1972, Computations of fluvial-sediment discharge: Techniques of Water-Resources Investigations of the U.S. Geological Survey Book 3, Chapter C3, 66 p.
Reid, L.M., and Dunne, T., Sediment production from forest road surfaces: Water Resources Research, v. 20, no. 11, p. 1753-1761.
Rinella, J.F., McKenzie, S.W., 1982, Method for relating suspended-sediment concentrations to suspended-sediment particle-size classes in storm-water runoff: U.S. Geological Survey Water-Resources Investigation Report 82-39, 43 p.
Ruffing, F.E., Britton, J.L., Flint, N.A., 1997, Effect of the February flood on the City of Salem water supply, in The Pacific Northwest Floods of February 6-11, 1996-- Proceedings of the Pacific Northwest Water Issues Conference, Portland, Oregon, October 7-8, 1997: American Institute of Hydrology, p. 31-37.
City of Salem, 1999, Annual report of Salem's drinking water, Volume 2: Salem, Oregon, Public Works Department, 8 p.
Sherrod, D.R., and Smith, J.G., 2000, Geologic map of upper Eocene to Holocene volcanic and related rocks of the Cascade Range, Oregon: U.S. Geological Survey Miscellaneous Investigations Map I-2569, scale 1:500,000.
Sidle, R.C. and Campbell, A.J., 1985, Patterns of suspended sediment transport in a coastal Alaska stream: Water Resources Bulletin WARBAQ Vol. 21/6, p. 909-917.
Swanson, F.J., and Dryness, C.T., 1975, Impact of clear-cutting and road construction on soil erosion by landslides in the western Cascade Range, Oregon: Geology, v. 3, no. 7, p. 393-396.
Truhlar, J.R., 1976, Determining suspended sediment loads from turbidity records, in Proceedings of the Third Inter-Agency Sedimentation Conference, Water Resources Council, Denver, Colorado: p. 65-74.
Uhrich, M.A., Wentz, D.A., 1997, Water quality study design for the Willamette Basin, Oregon, using a Geographic Information System, in International River Water Quality, Best, G.A., Bogacka, T., and Niemirycz, E., eds.: New York, E & FN Spon, p 181-189.
U.S. Army Corps of Engineers, 1953, Detroit Lake Water Control Manual, Detroit Dam Project: p. 33-34.
U.S. Environmental Protection Agency, 2000, Ambient water quality criteria recommendations, rivers and streams in nutrient Ecoregion II: EPA 822-B-00-015, 120 p.
U.S. Geological Survey, 1990, Land use and land cover digital data from 1:250,000- and 1:100,000-scale maps---Data user guide 4: Reston, Virginia, 25 p.
Vanoni, Vito A. (editor), 1975, Sedimentation Engineering: New York, American Society of Civil Engineers, 743 p.
Walker, G.W. and MacLeod, N.S., 1991, Geologic map of Oregon: U.S. Geological Survey, 2 sheets, scale 1:500,000.
Wellman, R.E., Gordon, J.M., Moffatt, R.L., 1993, Statistical summaries of streamflow data in Oregon, Volume 2--Annual low and high flow, and instantaneous peak flow: U.S. Geological Survey Open-File Report 93-63, 406 p.
1. Collection of continuous streamflow data at these sites, except Little North Santiam River (LSMO), actually began during 1905-08, although the record was intermittent until the start year listed.
2. Temperature data is available as follows: N SANTIAM and BREITENBUSH, 1951-1987; NIAGARA, 1954-1997; MEHAMA and LITTLE NORTH, 1986.
3. Method and location of cross-sectional sampling; DSSB = downstream side of bridge, USSB = upstream side of bridge, C = cableway, W= wading
4. Discharge record also available July 1905 to March 1907 and October 1910 to September 1914
5. Gaging station relocated in 1999
6. Discharge also available December 1908 to January 1920 and October 1921 to March 1922
7. Site located at river mile 5.5
8. Cableway at BREITENBUSH was reconstructed in January 2000 at an abandoned cableway location, approximately 0.25 miles downstream of gage. From October 1998 until this time high flow measurements and samples were collected from the main highway bridge (DSSB), 2 miles upstream of the gage
9. Discharge record also available June 1932 to September 1987
10. Discharge record also available January 1907 to October 1910
11. The use of trade products or firm names in this proposal is for descriptive purposes only and does not imply endorsement by the U.S. Geological Survey.
12. Lateral readings are not possible from the thalweg portion of the cross-section during high-flow conditions, due to high velocity.
13. This temperature is similar to most fall/winter stream temperatures and helps minimize algae growth.
14. A wide-mouth pipette was used to reduce vacuum effects during sample extraction.
North Santiam River Basin Suspended-Sediment and Turbidity Study Home Page
Contact: Mark Uhrich
U.S. Department of the Interior |
U.S. Geological Survey
URL: http://or.water.usgs.gov/santiam/proj_desc.html
Page Contact Information: Oregon Webteam
Page Last Modified: Monday - Mar 12, 2007 at 17:10:45 EDT
