Introduction

Background

It is widely recognized that dams have important effects on streamflow, water temperature, and fisheries (Collier and others, 1996). Often managed to provide protection from floods, reservoirs capture large inflows and release them over a longer time period, thereby decreasing annual peak flows. During dry seasons, reservoir releases for navigation, irrigation, municipal uses, and flow augmentation increase the annual minimum streamflow above pre-dam levels. Because deep reservoirs tend to thermally stratify in summer and keep less-dense warmer water at the surface, releases from deep in a reservoir typically are colder in mid-summer compared to nearby rivers without dams. Similarly, when a reservoir is drawn down in autumn to provide storage for flood control, the warm surface layer can be brought down to the outlet elevation, resulting in warmer-than-normal discharges. These alterations in streamflow and seasonal temperature patterns downstream of dams can create problems for fish related to the timing of migration, spawning, and egg hatching (Caissie, 2006). Dams also can be a barrier to fish passage, further disrupting the patterns and success of migration and spawning.

The Willamette River basin in northwestern Oregon has a system of dams that were built to provide flood control, power production, storage for agricultural and municipal uses, flow augmentation for downstream navigation, and recreation. The U.S. Army Corps of Engineers (USACE) operates the largest system of dams in the basin, totaling 13 storage and flow-regulating dams. Portland General Electric (PGE) owns and operates several hydropower projects, many of which are located in the Clackamas River basin, a tributary to the Willamette River. Fish passage and downstream thermal effects have long been an issue for many of these dams. A retrofit at the USACE Cougar Dam on the South Fork (SF) McKenzie River was completed in 2005 to allow for better control of downstream temperatures; the resulting seasonal temperature pattern now is more natural and has improved conditions for fish spawning and fry emergence in the reach downstream of the dam (Greg Taylor, U.S. Army Corps of Engineers, oral commun., 2010).

As a result of many factors including the dams, winter steelhead (Oncorhynchus mykiss) and Chinook salmon (Oncorhynchus tshawytscha) were listed as threatened under the Federal Endangered Species Act in March 1999. To protect these species, the National Marine Fisheries Service issued a “Biological Opinion” in July 2008, providing guidelines and timetables that address fish passage and the flow and temperature alterations caused by the dams, among other recommendations (National Marine Fisheries Service, 2008). Some of the recommendations already are being implemented, such as modified operations at the USACE Detroit Dam to better manage downstream temperatures.

Recognizing that water temperature is a critical factor affecting the survival and viability of anadromous fish, the Oregon Department of Environmental Quality (ODEQ) created a maximum water-temperature standard under the authority of the Federal Clean Water Act. To protect fish and aquatic resources and meet water-quality standards in the Willamette River and several of its largest tributaries, ODEQ issued a water-temperature Total Maximum Daily Load (TMDL) in September 2006 (Oregon Department of Environmental Quality, 2006a, 2006b). Relying on models and data, ODEQ attempted to quantify the thermal effects of several important factors on the river system. The TMDL specified heat wasteload allocations for point-source dischargers, called for revegetation of degraded riparian zones, and established temperature targets downstream of the major dams. The cumulative heating effects of point-source discharges and decreased riparian shading were determined using detailed calibrated models. Models were not available, however, to quantify the thermal effects of the major dams at dam sites, and the analysis of those effects in the TMDL was limited primarily to a comparison of measured temperatures upstream and downstream of the reservoirs.

Thermal effects downstream of tall dams can be large and commonly have characteristic seasonal and spatial patterns. These patterns and characteristics have been described and modeled by Risley and others (2010) for several idealized stream, reservoir, and climate conditions, demonstrating that altered release temperatures have an effect that dissipates with downstream distance, but that flow modifications have an increasing thermal effect downstream. In the Willamette River basin, a simple comparison of measured temperatures upstream and downstream of Detroit and Cougar Dams reveals temperature changes as much as 6 °C (cooler downstream in mid-summer, warmer in October) prior to actions taken to mitigate such effects (Oregon Department of Environmental Quality, 2006b; Rounds, 2007; Sullivan and others, 2007). If dam releases are not managed to provide for some daily variation in water temperature, a downstream pattern in the daily range of water temperature also can develop, with distinct nodes of minimal daily variation at daily travel-time distances (Lowney, 2000). These downstream thermal effects, and their mitigation through changes in dam operations, can be simulated accurately with models. Examples from the Willamette River basin include models of the reservoirs impounded by Detroit Dam and Scoggins Dam, which were used to evaluate how changes in dam operation can be used to meet target downstream water temperatures (Sullivan and Rounds, 2006; Sullivan and others, 2007). The downstream thermal effects of the Cougar Dam retrofit have been tracked downstream using the Willamette River TMDL models (Rounds, 2007).

Description of Study Area

The Willamette River flows for 187 mi from south to north past some of the largest cities in northwestern Oregon, including Eugene, Corvallis, Albany, Salem, and Portland (fig. 1). The river basin covers an area of approximately 11,500 mi² and is home to more than 2.6 million people, about 70 percent of the population of Oregon (U.S. Census Bureau, 2008). Bounded by the Cascade Range to the east and the Coast Range to the west, the basin has a modified maritime climate characterized by cool, wet winters and warm, dry summers. Eastward-moving Pacific storms deliver plentiful rain as well as snow in the mountains during the wet winter season; about 70–80 percent of the annual precipitation falls between October and March, and less than 5 percent in July and August (Wentz and others, 1998). Rainfall can total as much as 130 in/yr in the mountains, but typically 35–40 in/yr in the valley lowlands. Approximately 70 percent of the basin is forested and 20 percent, primarily in the valley bottom, is agricultural land (Hulse and others, 2002).

Streamflow in the Willamette River follows seasonal rainfall patterns, with the highest flows during winter storms and the lowest flows in late summer. A system of flood-control dams moderates annual peak flows. The mean annual peak flow in the Willamette River at Salem (USGS station 14191000) for the 1970–2008 post-dam time period was 111,000 ft³/s but as high as 244,000 ft³/s in 1996. Annual low flows are augmented by releases from upstream reservoirs during summer, with typical target minimum flows of 6,000 ft³/s at Salem and 4,000 ft³/s at Albany.

Dams in the Willamette River basin are owned and operated by various entities and for a wide range of purposes. The USACE operates a system of 13 dams, including the tallest dams (463 and 452 ft, Detroit and Cougar Dams, respectively) and the largest storage reservoirs (about 455,000 acre-ft, Detroit and Lookout Point Dams) in the basin (table 1). On the Clackamas River, PGE operates a series of dams and diversions for power generation; the most downstream dam is River Mill Dam.

Streamflow and temperature characteristics in Willamette Basin streams are linked to the physiographic region of the stream’s headwaters. The basin has four major physiographic regions distinguished by their predominant location, elevation, and geologic characteristics (fig. 2). Streams with significant source areas in the High Cascades derive their flow from rainfall and snowmelt, where relatively young and permeable volcanic rocks conduct snowmelt to large spring complexes that provide cold and consistent flows year round (Conlon and others, 2005). In contrast, streams in the Western Cascades and the Coast Range are more driven by rainfall and lack the large spring complexes of the High Cascades, resulting in more variable flows and lower flows and warmer temperatures in summer (Tague and others, 2007).

Temperatures in Willamette Basin streams generally follow the seasonal climate pattern, have characteristics derived from their morphology and physiographic source area, and are affected by several anthropogenic influences such as upstream dams, water withdrawals, point-source discharges, and modifications to riparian shading. ODEQ’s analysis for the Willamette River water-temperature TMDL showed that cumulative heating effects from point sources amounted to less than 0.2 °C and that loss of riparian shading accounted for a warming of 0.5–1.0 °C (Oregon Department of Environmental Quality, 2006b). Downstream of tall dams, the temperature can be modified by as much as 6–8 °C, both warmer and cooler depending on the time of year (Sullivan and others, 2007). The thermal effects of the dams diminish with distance downstream, as demonstrated by an analysis of the effects of the retrofit to Cougar Dam, where near-dam effects were as large as 6.0–6.5 °C, but decreased to a fraction of a degree in the Willamette River because of dilution and heat exchange with the atmosphere (Rounds, 2007).

Purpose and Scope

Dams have a significant effect on downstream flows and temperatures in the Willamette River basin (Oregon Department of Environmental Quality, 2006b; Rounds, 2007). The primary purpose of this study was to quantify the thermal effects of Willamette River basin dams, concentrating on the following two objectives:

• Develop and apply methods to estimate the water temperatures that would occur at selected Willamette River basin dam sites in the absence of those dams, and
• Simulate the downstream flow and thermal effects of Willamette River basin dams.

To take advantage of a suite of existing CE-QUAL-W2 flow and water-temperature models of the Willamette River and its major tributaries, the time period for this study was aligned with the calibrated time periods for those models—spring through autumn in 2001 and 2002. Although the models were run to determine downstream effects only for June through October 2001 and April through October 2002, without-dam water temperatures at the dam sites were estimated for the entire years of 2001 and 2002.

This study focused on 13 USACE dams that make up the Willamette Project in northwestern Oregon, as well as the combined effects of multiple dams on the Clackamas River, a tributary to the lower Willamette River (fig. 1, table 1). Other dams in the Willamette River basin, such as Scoggins Dam in the Tualatin River basin and the Carmen-Smith Project in the upper McKenzie River basin, were not included in this analysis because these dams and tributary reaches were not included in the models used to assess water-temperature issues for the Willamette River water-temperature TMDL; these dams also are small enough and far enough upstream that their effects on Willamette River flows and temperatures are minimal. Downstream effects were simulated only for the reaches included in the TMDL models. Where two or more dams are located on the same tributary, such as Green Peter and Foster Dams in the South Santiam River basin, the downstream effects were simulated starting at the most downstream dam only, although without-dam water temperatures still were estimated for the upstream dam sites.

In addition to water temperatures, without-dam streamflows were estimated at each of the dam sites. Those flow estimates were not a major focus of the study, but were a critical model input for the simulation of downstream thermal effects. Although an assessment of the flow effects of the dams was not a major objective of this study, those effects are substantial and the results provide useful data for future resource management.

Finally, the methods and techniques documented in this report are not meant to provide definitive equations for estimating without-dam water temperatures. Instead, the methods are meant to illustrate the types of estimation methods that can be used and to provide reasonable starting points for similar analyses or models. In the absence of more detailed models of water temperature for a without-dams condition, the methods documented in this report can be used to produce useful estimates.

First posted August 17, 2010