<<

Evaluation of Data from Two Gunnison Temperature Monitoring Stations Discontinued in 2018 by the Program Director’s Office

Jana Mohrman and Don Anderson Upper River Recovery Program January 2020

Purpose of Report

The Upper Endangered Fish Recovery Program (Recovery Program) has been monitoring river water temperatures at multiple sites in the basin since as early as 1992 (Figure 1 and Table 1) for multiple purposes, including evaluating the influence of the Bureau of Reclamation’s Aspinall Unit on downstream river temperatures and the commensurate effect on fish habitat conditions. This report documents the rationale for discontinuing, in 2018, two sites originally established in 1996, and provides some analysis of the information those sites provided.

Gunnison River Temperature Monitoring

The Bureau of Reclamation’s Aspinall Unit consists of a series of on the mainstem Gunnison River including Blue Mesa (established in 1966; storage capacity 940,800 acre- feet) and two smaller reregulating reservoirs downstream, Morrow Point (1968) and Crystal (1977). Temperatures in the lower Gunnison River are influenced to some degree by the temperature of water passing through the Aspinall Unit. The Recovery Program has annually monitored 12 sites in the Gunnison River basin below the Aspinall Unit; those sites have been monitored between 13 and 23 years.

Some temperature monitoring sites have been maintained by the Recovery Program Director’s Office (PDO) based in Lakewood, Colorado; others are maintained by the U.S. Fish and Wildlife Service’s (FWS) Grand Junction Fish and Wildlife Conservation Office based in Grand Junction, Colorado. Several of these sites became redundant to temperature monitoring sites installed nearby in 2009 by the U.S. Geological Survey (USGS), and for that reason were discontinued in 2014. The USGS sites report real-time data1. Data from the PDO sites are collected continuously on data loggers, reviewed, and posted to a web site once a year2 .

In 2018, two sites were discontinued at the recommendation of the PDO. The two discontinued sites are (1) below and (2) above the North Fork Gunnison River Confluence – highlighted in red in Table 1, and denoted as sites #3 and #4 in Figure 1. Crystal Reservoir, which is just upstream of the first discontinued site, is the most downstream reservoir of the Aspinall Unit.

Importance of River Temperatures

The Gunnison River downstream of Delta, Colorado is designated critical habitat for two endangered fish species: the (Ptychocheilus lucius) and the (Xyrauchen texanus). Currently, the observed range of the Colorado pikeminnow in the Gunnison

1 Data can be accessed at https://waterwatch.usgs.gov/wqwatch 2 Data can be accessed at www.fws.gov/mountain‐prairie/riverdata/temperatures.html 1

River extends only about 33 miles upstream of its confluence with the Colorado River, to a point near Delta, Colorado.

Cool water temperatures may be a factor limiting the Colorado pikeminnow’s range in the Gunnison River. Osmundson (1999) assessed the potential for further extending the range of the Colorado pikeminnow and determined that river temperatures limit the distribution of this fish in the Gunnison River, in spite of good habitat and prey conditions upstream. His 1999 study indicated that cool water temperatures upstream of the pikeminnow’s current range can interfere with its reproduction and can reduce its growth rate. That study examined the potential for extending the range of the Colorado pikeminnow in the Gunnison River by improving temperature conditions for this warm-water fish. It noted that relatively cool water releases from the reservoirs comprising the Bureau of Reclamation’s Aspinall Unit on the upper Gunnison River cause summer temperatures in the critical habitat to be about 3°C cooler than in other river reaches that have relatively large populations of Colorado pikeminnow. Osmundson proposed that efforts to recover the Colorado pikeminnow seek to increase the temperatures in the Gunnison River at Delta by 1oC in May, June, September and October, and by 2oC in July and August (Boyer and Cutler 2004). Osmundson’s work indicated that increasing mean water temperatures at Delta by these amounts would increase the mean annual thermal units (ATU) from 32 to 46, thus putting stream temperatures at Delta at a level similar to sites on the Yampa and Colorado which have abundant populations of pikeminnow (Osmundson 1999).

Figure 1: Temperature Monitoring Sites in the Gunnison River basin, Colorado

2

Table 1: Recovery Program temperature monitoring sites in the Gunnison River basin, CO.

Modeling Opportunities to Improve River Temperatures for Endangered Fish In response to recommendations from Osmundson (1999) to increase the temperatures in designated critical habitat, the Recovery Program sponsored Jean Marie Boyer and Amy Cutler to study alternatives for improving temperatures in the Gunnison River. Their conclusions were published in a 2004 report (Gunnison River / Aspinall Unit Temperature Study – Phase II, Final Report) hereafter referred to as the “Boyer Report”. The Boyer Report evaluated the potential to improve temperature conditions by manipulating instream water temperatures using temperature control devices (TCD’s) at . As part of that work, Boyer and Cutler developed a model to simulate reservoir release water temperatures. They used the CE-QUAL-W2 model version 3.1, developed by the U.S. Army Corps of Engineers, to simulate temperature in Blue Mesa, Morrow Point, and Crystal Reservoirs. The QUAL2K model (Chapra and Pelletier 2003) was used to simulate river temperatures in the Gunnison River from below Crystal to Delta, Colorado. The model results characterize the temperature regimes which are probable over a given series of years, the effect of controlling the reservoir elevations from which water is released (“selective withdrawals”) on the reservoir’s heat budget, and the range of achievable temperatures to meet downstream targets identified by Osmundson (1999). Data provided by the Recovery Program’s monitoring between 1997 and 2001 were used to calibrate the model, including temperature data from the two sites that are the focus of this report, plus a third monitoring site located below the confluence with the (site #11 in Figure 1). The Boyer Report assumed for modeling purposes that the total volume released annually from Crystal Reservoir would remain the same, only the timing would change. The study simulated “high-”, “average-”, and “low-flow” scenarios for each month, and evaluated multiple release scenarios:

3

1. Historical releases: Releases corresponded to the 1997-2000 historic conditions, in which flow varied from 0 to 6,015 cubic feet per second (cfs).

2. Flow recommendation releases: Simulated temperatures reflected likely operations if flows proposed by the U.S. Fish and Wildlife Service (McAda 2003) to promote the recovery of the endangered fish in the Gunnison River were adopted. Meeting these recommendations would require higher spring peak flow releases.

3. Releases under multiple temperature control device (TCD) scenarios: Two hypothetical selective withdrawal structures at Blue Mesa Reservoir were modeled: a fixed and a multiple-level TCD.

4. Releases affected by low reservoir storage volumes: This scenario assessed how low storage conditions in Blue Mesa Reservoir (2003 conditions) would affect release temperatures.

Figure 2: Temperature releases from the Crystal Reservoir modeled by Boyer and Cutler (2004) under the above scenarios for the years 1997 through 2000.

Conclusions from the Boyer Report included:  The best location for a temperature control device would be .  Large releases from combined with increased release temperatures could result in as much a 5o C warming in the Gunnison River at Delta. Smaller releases can result in 1o C warming at Delta and as much as 3o C warming upstream of the North Fork confluence.

4

 These increased release temperatures are achievable from June to October with the TCD option.  In general, there was not much variation in reservoir stratification and potential release temperatures for the modeled years 1997 through 2001. The PDO’s Two Discontinued Monitoring Sites As noted above, temperature data from the now-discontinued Below Crystal Reservoir and Above North Fork Gunnison River monitoring sites were used by Osmundson (1999) to characterize temperature conditions in the Gunnison River downstream of the Aspinall Unit and by Boyer and Cutler (2004) to calibrate their models and simulate temperature changes likely to occur as the result of specified changes in Aspinall Unit operations. Despite these sites being important for both of these reports, we describe the rationale for discontinuing them below. First Site: Gunnison River below Crystal Reservoir The Gunnison River below Crystal Reservoir temperature site was discontinued in 2018 because data collected at this location for 18 years (1997 through 2014)3 satisfactorily confirms a strong correspondence between the observed temperatures and Boyer and Cutler’s 2004 model predictions, and provides a robust characterization of seasonal temperatures and temperature variations at this location. Furthermore, as no temperature control devices have been installed at Blue Mesa Reservoir, and none are anticipated in the foreseeable future, it does not appear worthwhile to continue collecting temperature data at this site at this time. As illustrated in Table 2, the range of temperatures observed over this period was broader than the “Without TCD” range predicted in the Boyer Report in each of the six months. This is not surprising, as the 18-year period of monitoring would be expected to capture a much greater range of flow and temperature conditions than the four-year period considered by Boyer and Cutler. Table 2 shows the Boyer Report’s model predictions that releases from Crystal Reservoir with a TCD operational at Blue Mesa Reservoir would provide up to 18oC water in the late summer, while without the TCD maximum summer temperatures would be 13o C. The recent 21 years of data indicate a late-summer maximum of 14 o C under a somewhat below-average flow regime4, while minimum monthly temperatures were 1 to 4oC cooler than predicted by the Boyer Report under the ‘without TCD’ scenario..

3 Data were collected at this site through September of 2018, however for purposes of this report, this 18-year period was used, rather than the full 21-year record available, because this overlaps with the temperature monitoring record for the Gunnison River at Delta site located at the top of designated critical habitat for the Colorado pikeminnow. 4Average annual streamflow during the PDO monitoring period (1997-2014) was 12% less than during the entire 42- year period after the Aspinall units were completed (1977-2018) and during the Boyer study period of 1997-2001.

5

Table 2: Predicted monthly temperature ranges modelled by Boyer 2004 and measured by USFWS below Crystal Reservoir, in oC. The first two columns are with a Temperature Control Device (TCD) in Blue Mesa Reservoir, middle columns are predictions without the TCD, and last two columns are actual ranges in mean daily temperatures measured in those months over 21 years.

Figure 3 illustrates the generally close correspondence between the monthly average temperatures predicted by Boyer and Cutler under both high- and low-flow conditions5, and the temperatures observed over the 18-year period below Crystal Reservoir. For the months of May through October, mean monthly temperatures in both cases range from 6o to 12oC. Under all but two of the six month/flow-condition pairs, the observed mean monthly temperatures were within 0.68oC of the Boyer Report predictions. Larger temperature departures were apparent in only May low-flow months (1.08oC difference) and October high-flow months (1.86oC difference). The October high- flow temperature departure from the Boyer Report may be attributable to the fact there was only one month meeting the high-flow condition (October 1999), rather than multiple months corresponding to this hydrologic condition.

5 Boyer and Cutler defined the thresholds for “low flow” releases from Crystal Dam to be less than 2,400 cfs in May, less than 2,600 cfs in June, less than 2,500 cfs in July, less than 1,900 cfs in August, less than 1,600 cfs in September, and less than 1,300 cfs in October. 6

Boyer 4 yr and FWS 18 yr Water Temperatures for the Low Flows in the Gunnison River below Crystal Reservoir

14

12 o C 10

8

6

4 May Jun Jul Aug Sep Oct

Below Crystal Boyer Study 4yrs Below Crystal 18 yrs

Boyer 4 yr and FWS 18 yr Water Temperatures for the High Flows in the Gunnison River below Crystal Reservoir

14

12

o 10 C

8

6

4 May Jun Jul Aug Sep Oct

Below Crystal Boyer Study 4yrs Below Crystal 18 yrs

Figure 3. Comparison of average monthly mean daily water temperatures during low flow and high flow years predicted in Boyer 2004 for four years, versus PDO’s 18 years of measurement on the Gunnison River below Crystal Reservoir. Flows measured at USGS 09128000 Gunnison River below the Gunnison Tunnel, CO

7

Second Site: Gunnison River above the North Fork Confluence This site just upstream of the confluence of the North Fork of the Gunnison River, which is approximately 26.6 RM downstream from the Below Crystal Reservoir site, was also discontinued in 2018. The PDO felt that sufficient data had been gathered to adequately characterize the temperature regime at this site and to characterize the warming in river temperatures observed over the intervening 26.6 mile reach. No major tributaries join the river between these sites, suggesting that tributary inflow typically has little effect on water temperatures along this reach, or on year-to- year variations in river warming. River temperatures at the Above North Fork Confluence site are on average 1.97o C warmer than Below Crystal Reservoir, ranging from a 4.67o warmer average temperature in July, to a 0.89o lower average temperature in December (Figure 4; Table 3) The long monitoring record at both sites and the relative consistency in monthly mean temperature differences would likely support fairly reliable estimates of water temperature at the confluence when the temperature of water below Crystal Dam is known.

Average Monthly Temperatures Two Discontinued Sites below the Aspinall Unit on the Gunnison River

20 18 16 14 12 o

c 10 8 6 4 2 0 jan feb mar apr may jun jul aug sep oct nov dec

Below Crystal Reservoir Above NF of Gunnison

Figure 4. 18 years of measured monthly average water temperatures in the Gunnison River below Crystal Reservoir and in the Gunnison River above the Confluence of the North Fork of the Gunnison River. Bars illustrate one standard deviation around the monthly means.

8

Table 3. Difference in mean monthly river temperatures at the two discontinued FWS monitoring sites, in oC.

River Temperatures at Delta, Colorado Because one objective of establishing these two data monitoring sites was to assist in evaluating options for creating warmer river conditions for endangered fish downstream, we also considered the river temperatures observed at the Gunnison River at Delta, Colorado, location over this same 18-year time period. Table 4 illustrates the monthly average temperatures measured at that location (column 3), as compared to the targets suggested by Osmundson (1999) (column 2) to benefit endangered fish. Under low-flow conditions in May, June and July, and under the high-flow conditions in May and September, average river temperatures at Delta actually met or exceeded Osmundson’s suggested temperature targets. Temperatures fell more than 0.2o short of those targets only for June high-flow conditions (0.7o short), August low-flow conditions (1.2o short), and October high-flow conditions (0.7o short). This is a significant finding, as it suggests that water temperatures are frequently higher than they were expected to be in 1999, and are not as much of a limiting factor to Colorado pikeminnow’s use of river habitat upstream of Delta as was believed at the time of Osmundson’s report. That said, interpretation of these data is hampered by the fact that there were no high-flow releases from Crystal Reservoir in July or August of the 18-year period evaluated, and thus these monthly comparisons are incomplete. Interestingly, the temperatures the Boyer Report suggested were ‘achievable’ at Delta with a TCD structure in place at Blue Mesa Dam in May, June, and July of low-flow years were, on average, actually met and exceeded during the 18-year monitoring period without a TCD (Figure 5). This suggests that installation of a TCD at Blue Mesa may provide less temperature benefit during these months than was originally assumed, and perhaps less temperature-control value overall.

9

Table 4: Difference between target temperatures suggested by FWS flow recommendations (Osmundson 1999) for Delta verses the 18-year PDO period at Delta.

10

Predicted w/TCD and FWS 18 yrs Observed Water Temperatures During High flows on the Gunnison River 20

18

16 o

C 14

12

10

8

6 May Jun Jul Aug Sep Oct

Predicted Crystal w/TCD Below Crystal 18 yrs Delta 18 yrs Predicted Delta w/TCD

Predicted w/TCD and FWS 18 yrs Observed Water Temperatures During Low flows on the Gunnison River 20

18

16

o 14 C 12

10

8

6

4 May Jun Jul Aug Sep Oct Predicted Crystal w/TCD Below Crystal 18 yrs Delta 18 yrs Predicted Delta w/TCD

Figure 5: The upper graph is during high flow and lower graph during low flow conditions. The blue lines are predicted temperatures with a TCD versus 18 years of data below Crystal Reservoir. The red lines are predicted temperatures with a TCD versus 18 years of data below in Delta, CO.

Consideration of Hydrologic Differences All other factors being equal, increased reservoir storage, increased reservoir releases, and elevated rates of runoff tend to correspond to cooler river temperatures in the Gunnison River system. To

11 help evaluate whether differences in streamflow over the different periods described above were significant enough to have substantially influenced the river temperature regimes, average Gunnison River flows near the Colorado River confluence (USGS Gunnison River at Grand Junction gage #09152500) were examined for May through October for three different time periods since the completion of the last of the Aspinall Unit reservoirs in 1977 (Table 5). Mean flow during the period of PDO’s temperature monitoring (1997-2014) was about 12% lower than during the longer period of 1977-2018, and also about 12% less than the mean flow during the Boyer study period of 1997-2001 (Table 5). These substantially higher flows during the Boyer study period might partially explain why their forecasted temperatures downstream at the Delta, Colorado location were lower than what has since been observed.

Table 5: Hydrologic conditions in May through October during 3 periods; Boyer’s research, PDO’s monitoring and the period on the Gunnison River at Grand Junction after all Aspinall Units were operational.

Monthly Mean Daily (cfs) measured at the Gunnison River at Grand Junction (USGS 09152500)

Mean Daily Date Range Years Researcher May-Oct (cfs) 1997‐2001 5 Boyer 3,011 1997‐2014 18 PDO 2,635 1977‐2018 42 USGS 3,000

Summary The PDO’s collection of continuous water temperature data at two Gunnison River locations downstream of the Aspinall Unit (the first Below Crystal Reservoir, and the second Above the North Fork Confluence) have provided valuable data for assessing the effects of flows in this reach on water temperatures farther downstream, and for calibrating models to assess possible strategies for modifying Gunnison River water temperatures. The PDO concludes that sufficient data were captured at these two sites from 1997 to 2018 to adequately characterize the temperature regimes, provide a robust baseline of temperature data for future comparisons, and to verify Boyer and Cutler’s (2004) assumptions regarding river temperatures below Crystal Mesa Dam. Boyer and Cutler’s modeled temperatures at these sites provide sufficient accuracy that the Recovery Program has ceased to collect empirical data at these sites. Furthermore, in light of ongoing river temperature monitoring at other sites in the Gunnison River basin (Figure1), additional data collection at these two sites is not worth the cost and effort of maintenance. The data presented in this report are expected to be helpful in any further evaluations of potential temperature control modifications at Crystal Reservoir, should such a strategy be further pursued. 12

Section 3.2 of the 2011 Aspinall Study Plan (2011) concluded that, before pursuing such as strategy, a detailed analysis is needed to better understand the tradeoffs between operating to meet environmental flow targets in the Gunnison River and potential negative effects of these flows on the thermal regime for fish. The Recovery Program has tabled discussions of TCD installation at Crystal Reservoir until such time as there is interest in further evaluating such interactions. When and if those discussions are revived, these temperature data will help inform those evaluations.

13

Literature Cited

Aspinall Study Plan ad hoc Committee, 2001. Study Plan to Evaluate Effects of Aspinall Unit Operations to Benefit Habitat and Recovery of Endangered Fishes in the Gunnison and Colorado Rivers Aspinall Study Plan, by Aspinall Unit Study Plan ad hoc Committee, May 2011. Boyer, J.M. and A. Cutler. 2004. “Gunnison River / Aspinall Unit Temperature Study – Phase II, Final Report.” Upper Colorado River Endangered Fish Recovery Program. Project #107. USBR Contract #01-FC-40-5340. http://www.coloradoriverrecovery.org/documents‐publications/technical‐ reports/isf/tempphase2.pdf Chapra, S.C. and G. Pelletier. 2003. QUAL2K: A Modeling framework for simulating river and stream water quality (beta version): documentation and users manual. Civil and Environmental Engineering Dept., Tufts University, Medford, MA. Kaeding, L.R. & Osmundson, D.B. Environ Biol Fish (1988) 22: 287. https://doi.org/10.1007/BF00004894. McAda, C.W. 2003. Flow recommendations to benefit endangered fishes in the Colorado and Gunnison Rivers. Recovery Program Project No. 54. US Fish and Wildlife Service, July 2003, Revised October 2003. http://www.coloradoriverrecovery.org/documents‐publications/technical‐ reports/isf/GunnCoflowrec.pdf Muth, R.T., L.W. Crist, K.E. LaGory, J.W. Hayse, K.R. Bestgen, T.P. Ryan, J.K. Lyons, and R.A. Valdez. 2000. Flow and temperature recommendations for endangered fishes in the Green River downstream of . Final Report to Upper Colorado River Endangered Fish Recovery Program, Denver, Colorado. http://www.coloradoriverrecovery.org/documents‐ publications/technical‐reports/isf/flaminggorgeflowrecs.pdf Osmundson, D.B. 1999. Longitudinal Variation in Fish Community Structure and Water Temperature in the Upper Colorado River. Final Report, Recovery Action Plan, Project # 48-A. U.S. Fish and Wildlife Service. US Fish and Wildlife Service, Gunnison River Basin Programmatic Biological Opinion, 2009 http://www.coloradoriverrecovery.org/documents-publications/section-7-consultation/GUPBO.pdf

14

Hydrologic Appendix

20 21 Water Years of Temperature Data (C) at Gunnison River Below Crystal Dam (1997‐2017) 15 2007

10 Celsius

5

0

2000 ‐5 Oct Nov Dec Jan Feb Mar Apr May Jun

Figure A-1. Daily mean observations of water temperatures in the Gunnison River below Crystal Dam from water year 1997 to 2017. Each year is plotted in a different color. Data points for 2000 (blue) include unusually cold temperatures observed in October-December, while data points for 1007 (yellow) include unusually warm temperatures observed in March.

22 Water Years of Temperature Data (Co) at

25 Gunnison River above the North Fork (1997‐2017) 2007 20

15

10 Celsi

5

0

‐5 Oct Nov Dec Jan Feb Mar Apr May Jun

Figure A-2. Daily mean observations of water temperatures in the Gunnison River above the North Fork Gunnison River Confluence from water year 1997 to 2017. Each year is plotted in a different color. Data points for 2007 (green) include unusually warm temperatures observed in March, April and May.

15