Simulations of a hypothetical temperature control structure at Detroit Dam on the North Santiam River, northwestern Oregon

Water temperature models of Detroit Lake, Big Cliff Lake, and the North Santiam River in northwestern Oregon were used to assess the potential for a hypothetical structure with variable intake elevations and an internal connection to power turbines at Detroit Dam (scenario SlidingWeir) to release more natural, pre-dam temperatures year round. This hypothetical structure improved outflow temperature control from Detroit Dam while meeting minimum dry-season release rates and lake levels specified by the rule curve specified for Detroit Lake.

A water temperature target based on long-term, without-dams temperature estimates was developed and used to guide the Detroit Lake model to blend releases from the user-defined outlets at Detroit Dam. Simulations that included warm surface water releases during the spring and summer, and cool, deep hypolimnetic water releases later during autumn typically met the temperature target. Immediately downstream of Detroit Dam, these simulations resulted in temperatures within the range of the without-dams temperature estimates for most of the year until about November. The minimum release rates of flow imposed at Detroit Dam during late summer and early autumn exceeded unregulated, without-dams flow estimates. This higher flow led to temperatures near the low end of the without-dams temperature range 46.3 river miles downstream at Greens Bridge from July to September; the high flows released from Detroit Dam were less susceptible to downstream warming than the low unregulated flows. Simulations that blended warm and cool water from different outlets at Detroit Dam resulted in less daily temperature variation compared to the without-dams scenarios as far downstream as Greens Bridge.

Estimated egg-emergence days for endangered Upper Willamette River Chinook salmon (Oncorhynchus tshawytscha) and Upper Willamette River winter steelhead (Oncorhynchus mykiss) were assessed for all scenarios. Estimated spring Chinook fry emergence under SlidingWeir scenarios was 9 days later immediately downstream of Big Cliff Dam, and 4 days later at Greens Bridge compared with existing structural scenarios at Detroit Dam. Despite the inclusion of a hypothetical sliding weir at Detroit Dam, temperatures exceeded without-dams temperatures during November and December. These late-autumn exceedances likely represent the residual thermal effect of Detroit Lake operated to meet minimum dry-season release rates (supporting instream habitat and irrigation requirements) and lake levels specified by the current (2014) operating rules (supporting recreation and flood mitigation).