USGS Home
Contact USGS
Search USGS

Detroit Lake Temperature and Suspended Sediment Model

Introduction | Objectives & Approach | Analysis & Results | Significant Findings
Model | Data | Downloads | Report | References

Introduction

Detroit Lake: overview looking east (Photo by M. Uhrich, 29-Sep-2004)

Impounded behind Detroit Dam, Detroit Lake is a man-made reservoir located on the North Santiam River in the Willamette River Basin in northwestern Oregon (location map). Built by the U.S. Army Corps of Engineers (USACE) and completed in 1953, Detroit Dam is one in a system of 13 dams operated by the USACE in the Willamette River Basin to provide flood control, power generation, recreation, irrigation, and minimum flows for navigation on the Willamette River. Detroit Dam is 141 meters (463 feet) tall, and stores 561 million cubic meters (m³) (455,000 acre-feet) of water in the reservoir at full pool, with a surface area of 14.5 square kilometers (km²) (5.6 square miles). The water-surface elevation of the lake varies by 35 meters (115 feet) or more over the course of a year; it is kept high in summer for recreation, and drawn down in winter to increase the available storage for flood control.

Power generation at Detroit Dam often results in large variations in release flows over the course of a day. To dampen and reregulate these flows, a smaller dam (Big Cliff) was built 5 km (3.1 miles) downstream of Detroit Dam. Big Cliff reservoir, also completed in 1953, holds only 8 million m³ (6,450 acre-feet) of water at full pool, with a surface area of 0.6 km² (0.2 square miles).

Major inflows to Detroit Lake include the North Santiam River, Breitenbush River, Blowout Creek, French Creek, Box Canyon Creek, and Kinney Creek. The U.S. Geological Survey (USGS) operates flow and water-quality monitors on several of these inflows, as well as at locations downstream, most notably at Niagara, just downstream of Big Cliff Dam.

The North Santiam River is one of several major rivers flowing west from the Cascade Range in Oregon. These rivers are a source of high-quality drinking water; the City of Salem, Oregon, relies on the North Santiam River as its primary source of municipal water. The North Santiam River is also prime habitat for the rearing and spawning of anadromous fish such as salmon. Water temperature and suspended sediment, however, are issues of concern. Portions of the North Santiam and Santiam Rivers downstream of Detroit Lake periodically exceed Oregon's maximum water temperature criteria, and these reaches were included on Oregon's 2004/2006 303(d) list of impaired waterbodies. Suspended sediment is a potential problem due to its effect on salmon habitat and because it increases the cost and complexity of water treatment for municipal uses.

The USGS has worked in partnership with the City of Salem since 1998 to monitor and study sediment and turbidity throughout the North Santiam River watershed (Uhrich and Bragg, 2003; Bragg and Uhrich, 2004). As part of that assessment, a focused effort was undertaken to examine the effect that Detroit Lake has on temperature issues and sediment transport. In particular, developing a model that simulates the transport and fate of suspended sediment and the dynamics of water temperature in Detroit Lake was deemed to provide an important component of understanding how the lake affects suspended sediment and temperature in the North Santiam and Santiam Rivers downstream.

Objectives & Approach

The objectives of this study were to:

1. Develop a model of Detroit Lake to simulate circulation, water temperature, total dissolved solids, and suspended sediment in both the reservoir and its outflow,
    
2. Understand processes affecting suspended sediment, quantify sediment sources and transport to the lake outlet, and quantify sediment deposition in the lake, and
    
3. Understand processes controlling water temperature in Detroit Lake and its outflow, and demonstrate the water temperature effects of a hypothetical selective withdrawal device.

To meet these objectives, USGS personnel constructed, calibrated, and tested a model of circulation, water temperature, and suspended sediment in Detroit Lake. The model was constructed to simulate conditions that occurred in the entire calendar years of 2002 and 2003, as well as the period December 1, 2005, through February 1, 2006, in order to simulate some large winter storm events. During January 2006, about 70 centimeters (27.6 inches) of precipitation were recorded at Detroit Dam, making it the wettest January ever recorded and breaking the previous record set in 1970. Processes occurring in Big Cliff reservoir, the reregulating reservoir below Detroit Lake, were not included in this model.

After the model was constructed, calibrated, and tested, it was used to: (a) examine the sources of sediment to Detroit Lake and the lake's ability to trap those sediments, (b) estimate the amount and pattern of sediment deposition, and (c) simulate the in-lake and downstream effects of adding a hypothetical selective withdrawal to Detroit Dam to control release temperatures.

Analysis & Results

The Detroit Lake model was constructed using CE-QUAL-W2 version 3.12 (Cole and Wells, 2002). CE-QUAL-W2 is a laterally averaged, two-dimensional flow and water-quality model developed by the U.S. Army Corps of Engineers and Portland State University. The simulated dimensions are longitudinal (along the length axis of the waterbody) and vertical (from water surface to bed sediment). The model simulates flow, horizontal and vertical velocities, water temperature, and a suite of water quality constituents, including any number of suspended sediment groups. It has been applied to more than 400 lakes and reservoirs around the world, and is well documented and supported. It also is the same model used to simulate flow and water temperature in the North Santiam and Santiam Rivers downstream of Detroit Dam (Sullivan and Rounds, 2004).

The version of CE-QUAL-W2 used in this project was modified by USGS personnel to:

1. fix coding errors either posted by CE-QUAL-W2's development team or found by USGS,
    
2. add new model output fluxes related to sediment deposition, and
    
3. enhance model capabilities through the addition of a new subroutine to automatically blend outflows from multiple reservoir outlets to match a user-supplied downstream temperature target.

Input data for the Detroit Lake model included lake bathymetry, meteorological conditions (air temperature, dew-point temperature, wind speed and direction, cloud cover, and solar radiation), tributary inflows, tributary temperature and water quality, and lake outflows. Simulated constituents included water temperature, total dissolved solids, and two suspended sediment class sizes (sand/silt, clay), as well as lake stage and circulation.

The details of Detroit Lake model construction and calibration have been documented in a USGS Scientific Investigations Report:

Sullivan, A.B., Rounds, S.A., Sobieszczyk, S., and Bragg, H.M., 2007, Modeling hydrodynamics, water temperature, and suspended sediment in Detroit Lake, Oregon: U.S. Geological Survey Scientific Investigations Report 2007-5008, 40 p.
[Online at http://pubs.water.usgs.gov/sir20075008]

This report describes the objectives and results of the modeling work, including a quantification of model performance, a discussion of the processes that influence temperature and suspended sediment in the lake, and the results of model sensitivity tests. The report also documents several model applications geared toward a better understanding of sediment sources, deposition, and fate as well as the effects of a hypothetical selective withdrawal tower to control release temperatures from Detroit Dam.

Some details of the calibration, particularly the methods used to determine light-extinction coefficients from measurements of light profiles and secchi depth, are documented in:

Sullivan, A.B. and Rounds, S.A., 2006a, Modeling suspended sediment and water temperature in Detroit Lake, Oregon, in Proceedings of the Eighth Federal Interagency Sedimentation Conference, April 2-6, 2006, Reno, NV: Subcommittee on Hydrology of the Interagency Advisory Committee on Water Information, ISBN 0-9779007-1-1.
[Online at http://or.water.usgs.gov/santiam/detroit_lake/Sullivan_FISC_paper.pdf]

Significant Findings

1. An annual pattern of water temperature exists in Detroit Lake that was similar in all time periods modeled. The reservoir typically begins the year isothermal and cold. In spring, the lake surface warms and a thermocline develops by summer, isolating cold, dense water at the reservoir bottom. In fall, the water surface cools, and eventually the reservoir mixes, or "turns over," and becomes isothermal again.
    
2. Detroit Lake has an important influence on downstream water temperature in the North Santiam River. Reservoir outflow water temperature reaches an annual maximum in fall, at times exceeding the water temperature criterion. In the absence of Detroit Dam, the annual water temperature maximum would occur in midsummer. Water released from Detroit Lake also has less daily temperature variation compared to what would occur in the absence of the lake.
    
3. Model results demonstrated that if a selective withdrawal device were installed at Detroit Dam, the outflow from Detroit Lake could remain below Oregon's maximum water temperature criteria for the North Santiam River all year. A more natural seasonal temperature pattern could be produced through most of the year, but in fall the lake did not have enough stored cold water to match this hypothetical seasonal temperature pattern downstream of the dam.
    
4. Total dissolved solids (TDS) had an annual cycle in Detroit Lake. During spring storms, the inflowing TDS concentrations were relatively low. As the lake was not yet strongly stratified, these inflows acted to decrease TDS throughout the lake. As summer progressed, TDS concentrations in the inflows typically increased. The summer temperature stratification acted to keep summer inflows, with their higher TDS concentrations, in the epilimnion, preventing these inflows from mixing into the colder, denser hypolimnion. With the breakdown of stratification in the fall, waters with higher TDS concentrations in the epilimnion were mixed throughout the lake.
    
5. The largest suspended sediment loads entered Detroit Lake during storm events. During the record-breaking precipitation between December 1, 2005, and February 1, 2006, more mass of suspended sediment entered and was deposited in the reservoir than in the entire calendar years of 2002 and 2003 combined. In summer, when storms were few, the inflow of suspended sediment into the lake was small, and resultant lake concentrations also were low.
    
6. Most of the mass of sediment entering Detroit Lake was in a size class designated "sand and silt." Sediment in that size class comprised 85 percent of the inflowing mass in calendar year 2002, 83 percent in calendar year 2003, and 92 percent during the modeled 2005-2006 storm events.
    
7. Although the sand and silt component made up most of the mass of suspended sediment entering the reservoir, it comprised only a small portion of the suspended sediment exiting the reservoir. It constituted only 9 percent of the outflowing sediment mass in calendar year 2002, 7 percent in 2003, and 16 percent during the modeled 2005-2006 storms. Most of the mass of sediment leaving Detroit Lake was composed of clay-sized particles.
    
8. Assuming a bulk density of 1.89 g/cm³ (grams per cubic centimeter), 14,300 m³ (cubic meters) (11.6 acre-feet) of sediment was deposited in Detroit Lake in 2002, 11,820 m³ (9.6 acre-feet) in 2003 and 34,900 m³ (28.3 acre-feet) in storms from December 1, 2005, to February 1, 2006. Each of these sediment volumes is less than 0.01 percent of Detroit Lake's full pool volume of 561 million m³ (455,000 acre-feet). The model results indicate that most sediment deposition occurred in the upper reaches of the reservoir, near the inflows of Breitenbush River and the North Santiam River.
    
9. All inflows contributed suspended sediment to the reservoir outflow. The North Santiam River was the largest contributor, followed by Breitenbush River in calendar year 2003. The North Santiam River was unique in that it contributed sediment to the outflow in October and November, when contributions from other tributaries decreased. Tributaries situated that fed directly to the lake closer to the dam were more likely to contribute suspended sediment that was exported to the North Santiam River downstream of the dam.

Model

The model code, program, and all data necessary to run it for the years 2002, 2003, and 2006 are available from the links below. The model was compiled to run on the Windows operating system, but can be compiled and run on any system that has a FORTRAN 95 compiler. Although CE-QUAL-W2 is available with a graphical user interface, the copy distributed in these packages is a generic version that does not include the interface. The generic version produces the same results, but simply provides less feedback to the model user during the run. The generic version of the model was used exclusively by USGS personnel in this investigation.

The USGS model was a modification of version 3.12 of the CE-QUAL-W2 source code. Modifications included the correction of coding errors, the addition of new model output fluxes related to sediment deposition, and the addition of a subroutine to automatically blend outflows from multiple reservoir outlets to match a user-supplied downstream temperature target. The model was compiled using the Compaq Visual Fortran compiler (version 6.6.A) with the following options:

  /fast /nodebug /real_size:64
  /warn:(argument_checking,nofileopt,unused,nousage)

Data

Many sources of data were used to construct and run the USGS Detroit Lake model. All of the data necessary to run the model are included in the downloadable packages (see the download section below). Additional data related to the project may be found online:

USGS Flow, Stage, and Water-Quality Data
	North Santiam R. below Boulder Creek (14178000)	NWISweb | Data Grapher
	Breitenbush River above French Ck (14179000)	NWISweb | Data Grapher
	French Creek near Detroit (14179100)	NWISweb | Data Grapher
	Blowout Creek near Detroit (14180300)	NWISweb | Data Grapher
	North Santiam River at Niagara (14181500)	NWISweb | Data Grapher
	Detroit Lake near Detroit (14180500)	NWISweb
Elevation, Outflow, Inflow, Power Data and more
	USACE Detroit Dam (DET)	DET | Summary Tables
	USACE Big Cliff Dam (BCL)	BCL | Summary Tables
Meteorological Data
	Bureau of Reclamation Agrimet site (DTRO)	DTRO
	ODF RAWS site at Stayton (SYNO3)	SYNO3
	USACE Detroit Dam weather station (DTWO)	DTWO
	USACE Detroit Dam (DET)	DET

Downloads

Detroit Dam: aerial view (Photo: US Army Corps of Engineers, 11-Jul-1990)

Download reports:

- Detroit Lake model construction, calibration, & use (USGS SIR 2007-5008)
 

Download GIS coverages:

- Detroit Lake bathymetry
 

Download the model and the data to run it:

- Calibrated model for the year 2002 [ZIP, 2.68 Mb]
- Calibrated model for the year 2003 [ZIP, 2.60 Mb]
- Calibrated model for the year 2006 [ZIP, 1.14 Mb]
 

View animations of calibrated model output:

- Water temperature, 2003
- Suspended sediment, 2003
 

View fly-by animations of Detroit Lake topography/bathymetry:

- Fly-by of topography, draped with aerial photography [AVI, 12.2 Mb]
- Fly-by of lake bathymetry [AVI, 14.7 Mb]
 

Download Detroit Lake water-quality data:

- Lake profiles, Kinney station, 2002 [XLS, 116 kb]
- Lake profiles, Blowout station, 2002 [XLS, 112 kb]
- Lake profiles, Mongold station, 2002 [XLS, 114 kb]
- Lake profiles, Kinney station, 2003 [XLS, 130 kb]
- Lake profiles, Blowout station, 2003 [XLS, 135 kb]
- Lake profiles, Mongold station, 2003 [XLS, 130 kb]
- Temperature profiles, at log boom, 2003 [XLS, 20.2 Mb]
- Lake profiles, four stations, January 2006 [XLS, 50 kb]
- N/P data from USGS samples, Jun-Sep, 2003 [XLS, 43 kb]
- Detroit Lake algae data, Jun-Sep, 2003 [XLS, 351 kb]

References

Bragg, H.M. and Uhrich, M.A., 2004, The North Santiam River, Oregon, water-quality monitoring network: U.S. Geological Survey Fact Sheet FS-2004-3069, 6 p.
[Online at http://pubs.water.usgs.gov/fs20043069]

Cole, T.M. and Wells, S.A., 2002, CE-QUAL-W2: A two-dimensional, laterally averaged, hydrodynamic and water quality model, version 3.1,: U.S. Army Corps of Engineers Instruction Report EL-02-1 [variously paged].

Rounds, S.A. and Sullivan, A.B., 2006, Development and use of new routines in CE-QUAL-W2 to blend water from multiple reservoir outlets to meet downstream temperature targets, in Proceedings of the Third Federal Interagency Hydrologic Modeling Conference, April 2-6, 2006, Reno, NV: Subcommittee on Hydrology of the Interagency Advisory Committee on Water Information, ISBN 0-9779007-0-3.
[Online at http://or.water.usgs.gov/tualatin/fihmc3_w2_modifications.pdf]

Sullivan, A.B. and Rounds, S.A., 2004, Modeling streamflow and water temperature in the North Santiam and Santiam Rivers, Oregon: U.S. Geological Survey Scientific Investigations Report 2004-5001, 35 p.
[Online at http://pubs.water.usgs.gov/sir20045001]

Sullivan, A.B. and Rounds, S.A., 2006a, Modeling suspended sediment and water temperature in Detroit Lake, Oregon, in Proceedings of the Eighth Federal Interagency Sedimentation Conference, April 2-6, 2006, Reno, NV: Subcommittee on Hydrology of the Interagency Advisory Committee on Water Information, ISBN 0-9779007-1-1.
[Online at http://or.water.usgs.gov/santiam/detroit_lake/Sullivan_FISC_paper.pdf]

Sullivan, A.B. and Rounds, S.A., 2006b, Modeling water-quality effects of structural and operational changes to Scoggins Dam and Henry Hagg Lake, Oregon: U.S. Geological Survey Scientific Investigations Report 2006-5060, 36 p.
[Online at http://pubs.water.usgs.gov/sir20065060]

Uhrich, M.A., and Bragg, H.M., 2003, Monitoring instream turbidity to estimate continuous suspended-sediment loads and yields and clay-water volumes in the upper North Santiam River basin, Oregon, 1998-2000: U.S. Geological Survey Water-Resources Investigations Report 03-4098, 43 p.
[Online at http://pubs.water.usgs.gov/wri034098]

Questions? Comments? For more information about the Detroit Lake modeling project, contact:

Stewart Rounds or Annett Sullivan
U.S. Geological Survey
2130 SW 5th Avenue
Portland, OR 97201
503-251-3280, 503-251-3260
sarounds@usgs.gov, annett@usgs.gov

Oregon Water Science Center Home page
Oregon Water Science Center Hydrologic Studies page
North Santiam Suspended Sediment Study