Tracking heat in the Willamette River system, Oregon

Abstract

The Willamette River Basin in northwestern Oregon is home to several cold-water fish species whose habitat has been altered by the Willamette Valley Project, a system of 13 dams and reservoirs operated by the U.S. Army Corps of Engineers. Water-resource managers use a variety of flow- and temperature-management strategies to ameliorate the effects of upstream Willamette Valley Project dams on the habitat and viability of these anadromous and native fish. In this study, new capabilities were added to the CE-QUAL-W2 two-dimensional flow and water-quality model to inform those flow- and temperature-management strategies by tracking the quantities and ages of water and heat from individual upstream sources to downstream locations in the Willamette River system. Specifically, the fraction of water and heat attributable to upstream dam releases or other water inputs, and the fraction of heat sourced from environmental heat fluxes across the water and sediment surfaces, were tracked and quantified in the river at all locations and times simulated by the model. Applying the updated CE-QUAL-W2 models to the Willamette River system for the months of March through October in the years 2011 (cool and wet), 2015 (hot and dry), and 2016 (warm and somewhat dry) demonstrated that the influence of upstream dam releases on downstream water temperature diminished within a few days as water moved downstream. At sites that are roughly two or more days of travel from upstream dams (Albany and downstream), the July–August fraction of riverine heat content that could be tracked back to upstream dam releases typically diminished to less than 20 percent, despite the fact that roughly 50 percent of July–August streamflow could be attributed to upstream dam releases at the same sites. In contrast, the fraction of riverine heat content that could be attributed to environmental energy fluxes continued to increase with downstream distance, from about 59 to 67 percent at Albany during July–August to 62 to 73 percent at Keizer and 68 to 79 percent at Newberg. At locations sufficiently far downstream, upstream dam releases affect water temperature mainly through a decrease in travel time (less time for environmental heat fluxes to warm the river during summer) and an increase in thermal mass (more water to dilute and buffer incoming heat fluxes) rather than through the simple transport of heat content (water temperature) released from the dams. This concept was explored not only for the baseline conditions that occurred in March–October of 2011, 2015, and 2016, but also for a hypothetical high-flow release during August 2016 and an actual high-flow release during August 2017. In these high-flow releases, an extra 2,500 cubic feet per second (roughly) was released from Dexter Dam on the Middle Fork Willamette River, and downstream effects were measured (2017, actual) and simulated (2016, hypothetical). Results of the simulations were consistent with insights gained from the baseline conditions, such that temperature changes caused by flow augmentation were substantial in upstream reaches (measured cooling of about 1.5 °C near Harrisburg [43 miles downstream] and Albany [84 miles downstream] in 2017, and cooling of about 0.5 °C near Albany in 2016) and diminished farther downstream, but still measurable (more than a few tenths of a degree Celsius) even at Newberg, which is about 154 miles downstream. The direct downstream effects of dam releases on the river heat content attributable to those releases were increased by the hypothetical flow augmentation, with increases of 20 percent at Harrisburg and 12 percent at Keizer. Even with a decreased influence of environmental energy fluxes on river heat content, however, the fraction of heat content attributable to such fluxes was still more than 50 percent at and downstream of Albany and more than 70 percent at Newberg, where the river temperature was less affected by upstream dam-release temperatures and instead was affected primarily by a decreased travel time and increased thermal mass.