• Anderson, E. A., 1973: National Weather Service River Forecast System: Snow Accumulation and Ablation Model. NOAA Tech. Memo. NWS Hydro-17, 87 pp.

  • Barnett, T. P., , J. C. Adam, , and D. P. Lettenmaier, 2005: Potential impacts of a warming climate on water availability in snow-dominated regions. Nature, 438, 303309.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., and Coauthors, 2008: Human-induced changes in the hydrology of the western United States. Science,319, 1080–1083, doi:10.1126/science.1152538.

  • Brown, C., , and R. L. Wilby, 2012: An alternate approach to assessing climate risks. Eos, Trans. Amer. Geophys. Union,92, 401–402.

  • Bureau of Reclamation, 2011: SECURE Water Act section 9503(c): Reclamation Climate Change and Water 2011. Bureau of Reclamation Rep., 226 pp. [Available online at http://www.usbr.gov/climate/SECURE/docs/SECUREWaterReport.pdf.]

  • Bureau of Reclamation, cited 2012: Colorado River Basin Water Supply and Demand Study. [Available online at http://www.usbr.gov/lc/region/programs/crbstudy.html.]

  • Burnash, R. J. C., 1995: The NWS River Forecast System—Catchment modeling. in Computer Models of Watershed Hydrology, V. P. Singh, Ed., Water Resources, 311–366.

  • Burnash, R. J. C., , R. L. Ferral, , and R. A. McGuire, 1973: A generalized streamflow simulation system: Conceptual modeling for digital computers. Joint Federal and State River Forecast Center, U.S. National Weather Service, and California Department of Water Resources Tech. Rep., 204 pp.

  • Chambers, J. C., 2008: Climate change and the Great Basin. USDA Forest Service General Tech. Rep. RMRS-GTR-204, 4 pp.

  • Clow, D. W., 2010: Changes in the timing of snowmelt and streamflow in Colorado: A response to recent warming. J. Climate, 23, 22932306.

    • Search Google Scholar
    • Export Citation
  • Craig, R. K., 2010: “Stationarity is dead”—Long live transformation: Five principles for climate chang adaptation law. Harvard Environ. Law Rev., 34, 9–75.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , R. Knutti, , S. Solomon, , and A. S. Phillips, 2012: Communication of the role of natural variability in future North American climate. Nat. Climate Change, 2, 775779, doi:10.1038/nclimate1562.

    • Search Google Scholar
    • Export Citation
  • Farnsworth, R. K., , E. S. Thompson, , and E. L. Peck, 1982: Evaporation atlas for the contiguous 48 United States. NOAA Tech. Rep. NWS 33, 37 pp. [Available online at http://www.weather.gov/oh/hdsc/PMP_related_studies/TR33.pdf.]

  • Gangopadhyay, S., , and T. Pruitt, 2011: West-wide climate risk assessments: Bias-corrected and spatially downscaled surface water projections. Bureau of Reclamation Tech. Memo. 86-68210-2011-01, 138 pp.

  • Gillies, R. S., , S. Wang, , and M. Booth, 2012: Observational and synoptic analyses of the winter precipitation regime over Utah. J. Climate, 25, 46794698.

    • Search Google Scholar
    • Export Citation
  • Hamlet, A. F., , P. W. Mote, , M. P. Clark, , and D. P. Lettenmaier, 2005: Effects of temperature and precipitation variability on snowpack trends in the western United States. J. Climate, 18, 45454561.

    • Search Google Scholar
    • Export Citation
  • Harding, B. L., , A. W. Wood, , and J. R. Prairie, 2012: The implications of climate change scenario selection for future streamflow projection in the upper Colorado River basin. Hydrol. Earth Syst. Sci., 16, 39894007, doi:10.5194/hess-16-3989-2012.

    • Search Google Scholar
    • Export Citation
  • Hobbins, M., , A. W. Wood, , D. Streubel, , and K. Werner, 2012: what drives the variability of evaporative demand across the conterminous United States? J. Hydrometeor., 13, 11951214.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M., , D. Lettenmaier, , D. Cayan, , and B. Udall, 2009: Reconciling projections of Colorado River streamflow. Southwest Hydrol., 8, 2031.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , J. M. Melillo, , and T. C. Peterson, Eds., 2009: Global Climate Change Impacts in the United States. Cambridge University Press, 188 pp.

  • Knowles, N., , M. D. Dettinger, , and D. R. Cayan, 2006: Trends in snowfall versus rainfall in the western United States. J. Climate, 19, 45454559.

    • Search Google Scholar
    • Export Citation
  • Maurer, E. P., , L. Brekke, , T. Pruitt, , and P. B. Duffy, 2007: Fine-resolution climate projections enhance regional climate change impact studies. Eos, Trans. Amer. Geophys. Union, 88, 504.

    • Search Google Scholar
    • Export Citation
  • Mills, E., , E. Klotz, , T. Stonely, , M. Waters, , and L. Summers, 2003: Utah's M&I water conservation plan. Utah Division of Water Resources Rep., 44 pp. [Available online at http://www.conservewater.utah.gov/Final71403AACC.pdf.]

  • Milly, P. C. D., , K. A. Dunne, , and A. V. Vecchia, 2005: Global pattern of trends in streamflow and water availability in a changing climate. Nature, 438, 347350.

    • Search Google Scholar
    • Export Citation
  • Milly, P. C. D., , J. Betancourt, , M. Falkenmark, , R. M. Hersh, , Z. W. Kundzewicz, , D. P. Lettenmaier, , and R. J. Stouffer, 2008: Stationarity is dead: Whither water management? Science, 348, 573574.

    • Search Google Scholar
    • Export Citation
  • Mitchell, K. E., and Coauthors, 2004: The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. J. Geophys. Res., 109, D07S90, doi:10.1029/2003JD003823.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., 2006: Climate-driven variability and trends in mountain snowpack in western North America. J. Climate, 19, 62096220.

  • Nash, L. L., , and P. Gleick, 1991: The sensitivity of streamflow in the Colorado Basin to climatic changes. J. Hydrol., 125, 221241.

  • Ray, A. J., , J. J. Barsugli, , K. B. Averyt, , K. Wolter, , M. Hoerling, , N. Doesken, , B. Udall, , and R. S. Webb, 2008: Climate change in Colorado: A synthesis to support water resources management and adaptation Colorado Water Conservation Board Western Water Assessment Rep., 58 pp.

  • Rotstayn, L. D., , M. L. Roderick, , and G. D. Farquhar, 2006: A simple pan-evaporation model for analysis of climate simulations: Evaluation over Australia. Geophys. Res. Lett., 33, L17715, doi:10.1029/2006GL027114.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , M. P. Clark, , R. L. Armstrong, , D. A. McGinnis, , and R. L. Pulwarty, 1999: Characteristics of the western U.S. snowpack from Snowpack Telemetry (SNOTEL) data. Water Resour. Res., 35, 21452160.

    • Search Google Scholar
    • Export Citation
  • Stewart, T. S., , D. R. Cayan, , and M. D. Dettinger, 2005: Changes toward earlier streamflow timing across western North America. J. Climate, 18, 11361155.

    • Search Google Scholar
    • Export Citation
  • Wood, A. W., , L. R. Leung, , V. Sridhar, , and D. P. Lettenmaier, 2004: Hydrologic implications of dynamic and statistical approaches to downscaling climate model outputs. Climatic Change, 62, 189216.

    • Search Google Scholar
    • Export Citation
  • Woodbury, M., , M. Balso, , D. Yates, , and L. Kaatz, 2012: Joint Front Range Climate Change Vulnerability Study. Water Research Foundation Rep., 148 pp.

  • Woodhouse, C. A., , D. M. Meko, , G. M. MacDonald, , D. W. Stahle, , and E. R. Cook, 2010: A 1,200-year perspective of 21st century drought in southwestern North America, Proc. Natl. Acad. Sci. USA, 107, 21 28321 288, doi:10.1073/pnas.0911197107.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map of SLC water supply basins and delivery area. The Four Creeks are shaded in blue. The Provo, Weber, and Duchesne drainages are indicated by the purple, orange, and red shaded areas respectively. The stream-gauge location used for analyses in each basin was used to define the shaded basin area and is located at the lowest shaded point along the stream.

  • View in gallery

    The 30-yr mean monthly water delivery by major source. Mean annual delivery comprises 52% from Four Creeks; 35% from the Deer Creek/CUP; 8.5% from deep wells; and 4.5% from artesian and wells and springs.

  • View in gallery

    Schematic of potential impacts to water supply with shifting supply and demand. The solid blue and red lines represent smoothed 30-yr mean observed Four Creeks supply and SLC total system demand, respectively. The dashed blue line is a smoothed 30-yr average temperature +5°F simulation of Four Creeks supply, and the dashed red line is a hypothetical future demand scenario. Under current conditions, storage and groundwater are needed to make up the volume difference between late-summer (~July–October) supply and demand (the solid lines). Under possible future conditions, additional storage and groundwater may be needed to make up for the larger late-summer volume difference between future supply and demand (the dashed lines).

  • View in gallery

    Range of 112 CMIP3 BCSD-projected changes in annual temperature and precipitation (blue diamonds) in SLC combined watershed areas for a future 2035–64 vs 1981–2010 baseline climate (source: http://gdo-dcp.ucllnl.org). This study's sensitivity runs are also indicated: projected changes in temperature alone (red squares), precipitation alone (purple circles), and combined temperature and precipitation (green triangles).

  • View in gallery

    Sensitivities of April–October PET to its NLDAS drivers for the SLC area. The y axis represents PET responses (in mean daily millimeter depths) to perturbations in each of its six drivers (2-m temperature, 10-m wind speed, downward shortwave radiation, downward longwave radiation, atmospheric pressure, and specific humidity q) expressed as standard deviations from drivers' climatological mean values (1981–2010). The x axis represents the empirical standard deviations away from the long-term mean of the driver noted, with the remaining five drivers held constant at their climatological means. All curves cross at 7.59 mm day−1 at zero perturbation; this is the area-averaged observed long-term mean of PET. The values in the legend represent the response slope in millimeters per day per standard deviation of the indicated driver.

  • View in gallery

    Big Cottonwood Creek runoff sensitivity to temperature as indicated by 30-yr mean (water years 1981–2010) monthly runoff volumes forced by various temperature changes. Temperature changes (°F) are indicated in the legend, where “base” signifies base climate historical simulation with no temperature adjustment. Also shown is the observed mean monthly streamflow.

  • View in gallery

    Mean monthly flow sensitivities as percentage decrease in flow per degree Fahrenheit, for each of the Four Creeks individually and combined.

  • View in gallery

    Mean annual volume decrease under various projected temperature changes for the seven watersheds used by SLC. (PET is perturbed by same temperature increase.) Abbreviations are as follows: CK for City Creek; PC for Parleys Creek; BCW for Big Cottonwood Creek; LCW for Little Cottonwood Creek; and Duches for the Duchesne River.

  • View in gallery

    SS1: the driest single-year simulation for Four Creeks, with combined monthly volumes for 1992 and 1934 indexed from 1992 and reduced by temperature increases of 3° and 5°F. SS2: the paleoestimate is a 10% reduction of the 1934 record based on the draft local paleohydrologic reconstructions. Observations for 1981–2010 are the CBRFC naturalized flow average, and the base simulation is the CBRFC model base simulation for the 1981–2010 period with no temperature perturbation.

  • View in gallery

    SS1: Provo–Weber–Duchesne driest 5-yr drought average flows with +3° and +5°F forcing SS2 with estimated 5-yr paleodrought from draft tree-ring reconstruction.

  • View in gallery

    SS3: low year Deer Creek/CUP loss scenario. The blue line indicates supply assuming 1934 streamflow without temperature forcing and with maximum historical deep well use. The red line is SLC's observed average demand.

  • View in gallery

    Steady population growth demand cases (see Table 4 for case description). Note the D2 and D6 cases are based completely or in part on changes in per capita water consumption observed from 1988 to 2011. The D2 trend is not expected to continue indefinitely because of practical constraints, but further research is required to determine an inflection point in the trend.

  • View in gallery

    Fast population growth water demand cases (see Table 4 for case description). Note the D2 and D6 cases are based completely or in part on changes in per capita water consumption observed from 1988 to 2011. The D2 trend is not expected to continue indefinitely because of practical constraints, but further research is required to determine an inflection point in the trend.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 369 369 50
PDF Downloads 165 165 47

Planning for an Uncertain Future: Climate Change Sensitivity Assessment toward Adaptation Planning for Public Water Supply

View More View Less
  • 1 Western Water Assessment, Salt Lake City, Utah
  • | 2 National Center for Atmospheric Research, Boulder, Colorado
  • | 3 Physical Sciences Division, NOAA/Earth System Research Laboratory, Boulder, Colorado
  • | 4 Salt Lake City Department of Public Utilities, Salt Lake City, Utah
  • | 5 University of Utah, Salt Lake City, Utah
© Get Permissions
Full access

Abstract

Assessing climate change risk to municipal water supplies is often conducted by hydrologic modeling specific to local watersheds and infrastructure to ensure that outputs are compatible with existing planning frameworks and processes. This study leverages the modeling capacity of an operational National Weather Service River Forecast Center to explore the potential impacts of future climate-driven hydrologic changes on factors important to planning at the Salt Lake City Department of Public Utilities (SLC). Hydrologic modeling results for the study area align with prior research in showing that temperature changes alone will lead to earlier runoff and reduced runoff volume. The sensitivity of average annual flow to temperature varies significantly between watersheds, averaging −3.8% °F−1 and ranging from −1.8% to −6.5% flow reduction per degree Fahrenheit of warming. The largest flow reductions occur during the high water demand months of May–September. Precipitation drives hydrologic response more strongly than temperature, with each 1% precipitation change producing an average 1.9% runoff change of the same sign. This paper explores the consequences of climate change for the reliability of SLC's water supply system using scenarios that include hydrologic changes in average conditions, severe drought scenarios, and future water demand test cases. The most significant water management impacts will be earlier and reduced runoff volume, which threaten the system's ability to maintain adequate streamflow and storage to meet late-summer water demands.

Corresponding author address: Tim Bardsley, Western Water Assessment, 2242 W. North Temple, Salt Lake City, UT 84116. E-mail address: wwa.bardsley@gmail.com

This article is included in the Human Impact on Climate Extremes for Water Resources Infrastructure Design, Operations, and Risk Management special collection.

Abstract

Assessing climate change risk to municipal water supplies is often conducted by hydrologic modeling specific to local watersheds and infrastructure to ensure that outputs are compatible with existing planning frameworks and processes. This study leverages the modeling capacity of an operational National Weather Service River Forecast Center to explore the potential impacts of future climate-driven hydrologic changes on factors important to planning at the Salt Lake City Department of Public Utilities (SLC). Hydrologic modeling results for the study area align with prior research in showing that temperature changes alone will lead to earlier runoff and reduced runoff volume. The sensitivity of average annual flow to temperature varies significantly between watersheds, averaging −3.8% °F−1 and ranging from −1.8% to −6.5% flow reduction per degree Fahrenheit of warming. The largest flow reductions occur during the high water demand months of May–September. Precipitation drives hydrologic response more strongly than temperature, with each 1% precipitation change producing an average 1.9% runoff change of the same sign. This paper explores the consequences of climate change for the reliability of SLC's water supply system using scenarios that include hydrologic changes in average conditions, severe drought scenarios, and future water demand test cases. The most significant water management impacts will be earlier and reduced runoff volume, which threaten the system's ability to maintain adequate streamflow and storage to meet late-summer water demands.

Corresponding author address: Tim Bardsley, Western Water Assessment, 2242 W. North Temple, Salt Lake City, UT 84116. E-mail address: wwa.bardsley@gmail.com

This article is included in the Human Impact on Climate Extremes for Water Resources Infrastructure Design, Operations, and Risk Management special collection.

1. Introduction

Many studies indicate that warming temperatures over the next several decades will lead to general decreases in runoff across the western United States, along with a shift toward earlier timing of runoff, regardless of any decline in annual precipitation (Nash and Gleick 1991; Barnett et al. 2008; Gangopadhyay and Pruitt 2011; Bureau of Reclamation 2012; Woodbury et al. 2012) (for a complete bibliography, see Bureau of Reclamation 2012). Such changes could pose significant challenges to municipal water managers in the region, who already face pressures from population growth and land-use change.

The Intermountain West is particularly vulnerable to climate-induced hydrologic impacts because of its dependence on the accumulation and storage of snow in mountain watersheds, which serves as a massive natural reservoir. Across the region, the percentage of total annual runoff generated by snowmelt from April to July generally ranges from 50% to 80% (Serreze et al. 1999; Stewart et al. 2005). In the watersheds supplying Salt Lake City, this figure varies around a mean of 72%. Further, including minor snowmelt in March and August and snowmelt's contribution to base flow in other months, the total snow cycle contribution to surface water supplies feeding the Salt Lake City Department of Public Utilities (SLC) likely exceeds 80%. However, temperature increases are already leading to more winter precipitation falling as rain instead of snow across the western United States (Gillies et al. 2012; Knowles et al. 2006). If this trend continues, the region will see further decreases in snowpack, earlier snowmelt, and a shifting to earlier runoff (Clow 2010; Barnett et al. 2005; Barnett et al. 2008). Prior studies project significant impacts on water providers' abilities to meet summer water demands with current water storage infrastructure (Ray et al. 2008; Chambers 2008; Karl et al. 2009; Woodbury et al. 2012).

Moreover, although traditional water management has long relied on an assumption of stationarity, future climate change may result in hydrologic regimes not well represented by historically observed records (Milly et al. 2008; Woodhouse et al. 2010), rendering this assumption no longer defensible for engineering, planning and management applications (Craig 2010). Improved awareness of decadal or longer-term variations or trends in observed and projected climate changes has thus led water managers to evaluate numerous approaches to gauge the variability and risk of a changing climate (e.g., Bureau of Reclamation 2011; Bureau of Reclamation 2012).

This paper reports quantitative analyses of hydrologic sensitivities to climate variability and possible future climate change impacts that could affect the ability of SLC to continue to provide sufficient high-quality water to its customers. To plan for resilient and sustainable long-term water supplies, SLC's management has recognized the need to understand better the range of potential impacts of climate change or variability with regard to runoff timing, volume, and severe drought in their specific water supply basins. This paper describes the investigation of these potential impacts by a collaborative effort among utility staff, a federal agency, and universities that was facilitated by the Western Water Assessment (WWA), a National Oceanic and Atmospheric Administration (NOAA)-funded Regional Integrated Sciences and Assessments program. To facilitate use of the NOAA Colorado Basin River Forecast Center's operational hydrologic modeling framework and to simplify communication with water managers in the United States, results are reported in imperial units.

The primary objective of this study is to inform water management and long-range planning decisions through a partial bottom-up assessment (Brown and Wilby 2012) of SLC system sensitivities to potential vulnerabilities in water supplies given climate operational adaptation options and measures. SLC's water management decisions depend on observed and forecast snowpack, runoff patterns, and short- and long-term climate. In particular, SLC must decide when to use various sources of water, especially surface water streams and man-made reservoirs, to meet demand. For instance, SLC must balance the costs and benefits of water supply from the Parleys Reservoir System, which is solely in SLC's water rights control, with supplies from the Deer Creek System, where a more complex water rights relationship among several entities exists. For long-term planning, SLC's water management decisions must focus on whether new sources of water and/or water storage should be developed to meet demand. Therefore, SLC's objective through this work is to understand hydrologic sensitivities, which in turn will help SLC be more strategic in its next phases of climate adaptation planning. This assessment addresses these objectives through the following analyses: 1) hydrologic modeling of the SLC watersheds; 2) simulating runoff sensitivities for a range of changes in temperature, potential evapotranspiration (PET), and precipitation; 3) evaluating extremely low future water supply and drought scenarios; and 4) formulating and assessing test cases for future water demand.

2. Background

2.1. Salt Lake City's water supply system

SLC delivers treated water to over 349 000 people in and beyond the municipal boundary of Utah's largest city. SLC's water supplies come from four primary sources: canyons in the Wasatch Mountains immediately east of Salt Lake City, federally funded projects diverting water from the east side of the Wasatch Mountains and western Uinta Mountains, local springs, and local deep wells (Figure 1).

Figure 1.
Figure 1.

Map of SLC water supply basins and delivery area. The Four Creeks are shaded in blue. The Provo, Weber, and Duchesne drainages are indicated by the purple, orange, and red shaded areas respectively. The stream-gauge location used for analyses in each basin was used to define the shaded basin area and is located at the lowest shaded point along the stream.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

SLC's highest-quality and most reliable waters flow from canyon streams in the Wasatch Mountains and are fed by gravity to treatment facilities and the distribution system, with approximately 50%–60% of SLC's current water supply provided by four Wasatch streams: City, Parleys, Big Cottonwood, and Little Cottonwood Creeks (collectively called the “Four Creeks”; Figure 2). Because SLC owns the majority of the water rights in the Four Creeks, it maintains a greater level of direct control over the management of both the water resources and their watersheds, relative to its other surface water sources that are part of the Deer Creek system (described later), including significant watershed protections to maintain high water quality. The high water quality of the Four Creeks not only results in lower capital and operating costs for water treatment but also helps SLC meets federal and state drinking water requirements. Reservoir storage for water supply in the Four Creeks watershed is relatively small, at approximately 21 000 acre-feet (21 KAF). That storage, held in several small lakes almost entirely on Parleys Creek, is equivalent to the mean annual flow in the creek, or approximately 20% of average annual system demand. Opportunities to increase storage capacity are limited. SLC supplements water supplies from the Four Creeks with local springs and deep wells as needed to meet summertime demands.

Figure 2.
Figure 2.

The 30-yr mean monthly water delivery by major source. Mean annual delivery comprises 52% from Four Creeks; 35% from the Deer Creek/CUP; 8.5% from deep wells; and 4.5% from artesian and wells and springs.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Through various exchanges, SLC has also secured water supplies from two federal water projects: Deer Creek Reservoir on the Provo River and the Jordanelle Dam [a component of the Bonneville Unit of the Central Utah Project (CUP)], whose source waters include the Provo, Weber, and Duchesne Rivers through diversions. Water supplies from the CUP and Deer Creek are referred to collectively as the “PWD.” After collection, this water is diverted into and conveyed through the Provo River system to SLC's system (Figure 1). The average annual naturalized flow on the Provo is 135 KAF, while the combined average annual flow volume on the PWD rivers is 453 KAF. The combined storage capacity of Deer Creek and Jordanelle reservoirs on the Provo is 473 KAF, with additional storage on the Weber and Duchesne systems. SLC rights on the PWD system represent a fraction of these average flows.

The watersheds used in these analyses were defined as the drainages upstream of long-term gauges near headwater locations in order to avoid including human impairments of flows. These source watersheds are also shared with numerous other communities in northern Utah and they collectively serve over one million people, industry, and agriculture. The Four Creek watersheds, while highly variable, tend to be smaller, steeper, and—in the case of Big and Little Cottonwood Creeks—wetter than the comparatively large PWD basins that share a headwater ridgeline in the western Uinta Mountains (Table 1 and Figure 1).

Table 1.

Watershed characteristics.

Table 1.

Given the limited storage capacity in the Four Creeks watersheds, runoff timing is crucial to system operations. Under normal conditions, snowmelt runoff from the Four Creeks begins to diminish in midsummer, so SLC uses other sources to meet late-summer demand, much of which includes outdoor irrigation. Current management of the watershed makes the relationship between the shape of the Four Creeks streamflow supply curve and the system water demand curve in any given year a critical factor in SLC's decision making. Of primary concern is the potential for future earlier runoff and expanded demand with rising temperatures leading to a greater volume to be made up by other sources (illustrated in Figure 3 as areas formed when the red curves are above the blue curves). Thus, an analysis of how the interaction of these supply and demand curves may shift because of changes in climate is essential to assessing risk to long-term supply reliability.

Figure 3.
Figure 3.

Schematic of potential impacts to water supply with shifting supply and demand. The solid blue and red lines represent smoothed 30-yr mean observed Four Creeks supply and SLC total system demand, respectively. The dashed blue line is a smoothed 30-yr average temperature +5°F simulation of Four Creeks supply, and the dashed red line is a hypothetical future demand scenario. Under current conditions, storage and groundwater are needed to make up the volume difference between late-summer (~July–October) supply and demand (the solid lines). Under possible future conditions, additional storage and groundwater may be needed to make up for the larger late-summer volume difference between future supply and demand (the dashed lines).

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Possible changes in water demand from SLC customers are an important part of such an assessment. Water use in Salt Lake City is driven by a complex combination of human, economic, and climatologic factors. Salt Lake City has a semiarid climate with four distinct seasons, generally confining outdoor water use in SLC's service area to April–October. Therefore, outdoor water use is defined as the April–October water use in excess of the indoor use. Indoor water use is estimated by assuming that the rate of total use from November through February in any given year is representative of that year's indoor use rate for every month. Both indoor and outdoor per capita water use have decreased with time, reflecting successful water conservation efforts. Outdoor per capita use has decreased at a more rapid rate than indoor per capita use. Outdoor watering, which is currently 45% of the use of the annual water supply, is significantly correlated to monthly or seasonal temperature, precipitation, and PET and is thus the variable most likely to be affected by future climate change. Total annual water use in the SLC delivery area has been fairly constant over the past 30 years despite steady population growth. SLC reduction in per capita water use has far exceeded the state of Utah's published 2050 water conservation plan (Mills et al. 2003) and the more aggressive unpublished governor's 2013 update, nearly accomplishing the plans' 2050 and 2025 goals, respectively, by 2010.

2.2. Observed and projected climate changes

Temperature, precipitation, and evapotranspiration are primary climate drivers of hydrology, and therefore critical inputs to the Colorado Basin River Forecast Center (CBRFC) hydrology model used in the analyses described in this paper. Climate data show that temperatures in Utah and the Intermountain West have increased faster in recent decades compared to continental U.S. and global averages; from 1895 to 2012, the average annual temperature across Utah increased at 0.21°F decade−1 (http://www.ncdc.noaa.gov/cag/). Long-term observations from two weather stations located in the mountains of the Four Creeks watershed (Brighton and Mountain Dell) indicate significant increases in average annual temperature from 1941 to 2009. We also identified significant upward trends in observed average annual, spring, summer, and fall temperatures at Salt Lake City from 1881 to 2010 (http://data.giss.nasa.gov/gistemp/station_data/), finding that the steepest rate of change occurs in summer (see Table 2).

Table 2.

Observed trends, calculated using the Mann–Kendall test.

Table 2.

Climate projections indicate that these temperature trends are likely to continue. Mean annual temperature projections for the northern Utah mountain domain, comprising 60 ⅛° grid cells centered on the Four Creeks and PWD watersheds, were obtained from phase 3 of the World Climate Research Programme (WCRP) Coupled Model Intercomparison Project (CMIP3) climate projections (http://gdo-dcp.ucllnl.org/downscaled_cmip3_projections/; see Maurer et al. 2007). These data were derived from 16 global climate models (GCMs) run under three emissions scenarios (A1b, A2, and B1), resulting in a total of 112 runs due to multiple runs with some of the GCMs. The GCM output was statistically downscaled using the monthly bias-correction/spatial disaggregation (BCSD; Wood et al. 2004) approach. All runs show increasing temperatures through the twenty-first century. Model results indicate temperature increases in the range 1°–6°F by 2035–64, as compared to a 1981–2010 base period (Figure 4 and Table 3).

Figure 4.
Figure 4.

Range of 112 CMIP3 BCSD-projected changes in annual temperature and precipitation (blue diamonds) in SLC combined watershed areas for a future 2035–64 vs 1981–2010 baseline climate (source: http://gdo-dcp.ucllnl.org). This study's sensitivity runs are also indicated: projected changes in temperature alone (red squares), precipitation alone (purple circles), and combined temperature and precipitation (green triangles).

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Table 3.

Summary of 112 CMIP3 BCSD climate projections. Changes are for 2035–64 compared to a 1981–2010 base period for 60 ⅛° grid cells centered on SLC water supply basins.

Table 3.

Both the observed precipitation records for Salt Lake City and climate model projections of precipitation for the region tell a less-certain story than for temperature. Trend analyses for Salt Lake City Airport's observed precipitation records from 1928 to 2010 show increases in annual, winter, spring, and fall precipitation, but only the spring and fall changes are statistically significant (Table 2). These precipitation trends are consistent in sign with the projected ensemble mean from the CMIP3 BCSD dataset described above, indicating increases in all seasons except summer. The range of the 112 monthly projections shows large uncertainties in both the sign and magnitude of future precipitation trends: 68 out of the 112 projections (60.7%) indicate increasing annual precipitation, with a mean change of +2.1%. Overall, however, the range is −17.8% to +21.5% by 2035–64 (base period is 1981–2010; Figure 4 and Table 3). The uncertainty in the direction of change aligns with prior findings that northern Utah is located on a boundary between zonally oriented regions of greater precipitation projection consistency: that is, the drier U.S. Southwest and the wetter U.S. northern tier (Milly et al. 2005). Uncertainty in the projections also arises from the difficulty that even downscaled outputs exhibit accurately reproducing precipitation patterns across mountainous areas and from the weakness of the precipitation change signal at regional scales in current ensembles of climate projections (Deser et al. 2012; Harding et al. 2012).

Limited information is available for projections of future nontemperature drivers of PET: that is, specific humidity, wind speed, shortwave and longwave radiation, and atmospheric pressure. April–October estimates of PET in the SLC delivery area from a physically based PET (the PenPan equation of Rotstayn et al. 2006). The same method described later to drive the CBRFC hydrology model does not reveal a statistically significant trend in PET over the period 1981–2010 corresponding to the CBRFC hydrology model calibration period.

Changes in runoff timing and volume are expected because of increases in temperature alone. Although multiple studies have documented decreasing snowpack and earlier runoff in the western United States, the largest runoff changes appear in warmer coastal mountains, with little change in the higher elevations and colder climates of the Intermountain West (Knowles et al. 2006; Mote 2006). Gillies et al. (Gillies et al. 2012) evaluated changes in the precipitation regime in Utah, using observation-based gridded precipitation and temperature data, combined with a rain–snow threshold, and found a significant increase in precipitation paired with a decrease in snowfall, resulting in a 0.045% yr−1 decreasing trend in the wintertime snow/precipitation ratio for the period of 1950–2003. Among the gauged data available in the SLC area, the U.S. Geological Survey (USGS) stream gauge for the Weber River near Oakley, Utah, provides the longest continuous headwater gauge data available (1905–2010). Although gauge data show decreasing flows and a tendency toward earlier runoff, only flow decreases from October through February are significant (Table 2).

3. Methods

The sensitivity assessment of potential climate change impacts to SLC's water resources comprises four major components detailed below: 1) hydrologic modeling of the SLC watersheds; 2) a watershed-specific analysis of sensitivity to changes in temperature, PET, and precipitation for the seven primary supply watersheds; 3) tests of scenarios of extremely low future water supply and droughts using results from the watershed sensitivity analyses in item 2 above; and 4) water demand analyses, which allowed for the generation and evaluation of future demand test cases.

3.1. Watershed modeling

The colocation of WWA personnel (the lead author) at CBRFC, a regional operational center supplying short- and seasonal-range model-based streamflow forecasts to SLC water operations and management, facilitated the application of existing calibrated hydrology models for the SLC system. CBRFC's modeling environment includes the Sacramento Soil Moisture Accounting Model (SACSMA) coupled with the Snow-17 temperature index snow model (Burnash 1995; Burnash et al. 1973; Anderson 1973). These models (referred to in aggregate as “the CBRFC model”) were chosen because of their existing calibrations for the watersheds of interest available through the CBRFC. The CBRFC model was run within the National Weather Service (NWS) Community Hydrologic Prediction System (CHPS), which is driven by three climatological forcings: mean areal temperature, precipitation, and PET, which are specified for two to three elevation zones in the drainage area of each forecast point. In addition, CBRFC maintains a database of daily naturalized flows developed using all available records impacting forecast points.

A weakness of using the current operational CBRFC model in climate sensitivity analyses is that the PET input to the Sacramento model is a static annual cycle of monthly average values (“static PET”). This PET climatology was derived for each watershed elevation zone from atlases of observed pan evaporation and modeled PET across the period 1956–70 (Farnsworth et al. 1982). Not only does this period not coincide with the operational CBRFC calibration period but also, by its nature, the static PET exhibits no variability on interannual or weather-related scales. CBRFC has also investigated the incorporation of temporally dynamic (daily, weather-scale varying) PET input. Dynamic PET is a physically based estimate driven by temperature, specific humidity, wind speed, shortwave and longwave radiation, and atmospheric pressure (in order of sensitivity; see Figure 5) derived from ⅛° gridded meteorological forcings from the North American Land Data Assimilation Systems (NLDAS; Mitchell et al. 2004). Thus, while it is conceptually more suitable for climate sensitivity analyses, this is constrained to the extent that these physical drivers can be projected with confidence. Changing from the standard climatologic PET forcing to this dynamic PET input did not appreciably change the performance statistics of historical simulations after model recalibration: high correlations of monthly flows were achieved in both cases (e.g., 0.93 and 0.94 for Big Cottonwood with static and dynamic PET, respectively). Nonetheless, in this study we use dynamic PET inputs in which the future PET is sensitive to changes in temperature only, due to lack of confidence in future changes in the other drivers.

Figure 5.
Figure 5.

Sensitivities of April–October PET to its NLDAS drivers for the SLC area. The y axis represents PET responses (in mean daily millimeter depths) to perturbations in each of its six drivers (2-m temperature, 10-m wind speed, downward shortwave radiation, downward longwave radiation, atmospheric pressure, and specific humidity q) expressed as standard deviations from drivers' climatological mean values (1981–2010). The x axis represents the empirical standard deviations away from the long-term mean of the driver noted, with the remaining five drivers held constant at their climatological means. All curves cross at 7.59 mm day−1 at zero perturbation; this is the area-averaged observed long-term mean of PET. The values in the legend represent the response slope in millimeters per day per standard deviation of the indicated driver.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

3.2. Watershed sensitivity analyses

Climate sensitivity scenarios were created by applying uniform (throughout the year) temperature shifts and precipitation scalings to the historical period forcings for water years (WY) 1981–2010. Given the large range in GCM-projected precipitation for the region, both positive and negative precipitation perturbations were run. We conducted seven temperature runs with unperturbed precipitation, six independent precipitation runs with unperturbed temperature, and six combined precipitation–temperature runs. The dynamic PET inputs for each temperature forcing sensitivity run reflected the consistent temperature-driven changes in PET. Figure 4 shows this study's sensitivity run forcing changes compared to midcentury CMIP3 BCSD ensemble results, illustrating that the scenarios evaluated span the projected ranges, at least for annual average changes.

The temperature and precipitation sensitivity cases displayed in Figure 4 were chosen to evaluate a range of potential changes to SLC's watersheds. CMIP3 BCSD results summarized in Table 3 indicate that future summers may be significantly drier and warmer than other seasons, and this seasonality of projected climate changes are expected to influence runoff changes. The modeling resources available, however, limited this study to using seasonally uniform temperature and precipitation change scenarios. These nonetheless provide useful, basic insights into water supply sensitivity. Even were it possible to analyze ensembles of projections, recent assessments of the influence of climate model internal variability on GCM output urge caution in deriving specific GCM-based climate scenarios for regional water supply analyses (Deser et al. 2012; Harding et al. 2012).

3.3. Future water supply scenarios

Scenario development for climate change impacts to SLC's system focused on potential drought, given its significant impacts on water supply reliability in the Salt Lake region. Drought scenarios that cause the greatest stress on the SLC's system vary based on duration and watershed. Limited storage capacity on the Four Creeks makes a single extremely dry water year (such as 1934, the lowest annual streamflow volume on record) the greatest concern with respect to availability of those water resources. In contrast, the drought scenario of greatest concern on the PWD system is a multiyear drought (such as the drought of 2000–04, the lowest 5-yr observed total runoff volume) because the system has multiannual storage capacity. A concurrent 1934 drought on Four Creeks and 2000–04 drought on the PWD would severely test SLC's system—a scenario that has been used for planning purposes—is included in supply scenarios 1 (SS1) and 2 (SS2) below.

The climate–streamflow sensitivity analyses results were used to evaluate potential impacts of extreme drought exacerbated by future climate change. Possible future water supply scenarios were developed in consultation with SLC to represent a range of potential future system challenges. Given the greater uncertainty surrounding possible future changes in precipitation, the scenarios included impacts of temperature increases without changes in precipitation. The results we present come from the evaluation of these scenarios for potential changes in annual volume and timing relative to typical demand. Future analyses will use a system planning model (currently in development) to further quantify potential impacts of these and additional scenarios to system reliability. We summarize three scenarios below.

The first supply scenario, SS1, represents the driest years of the observed record combined with warming. The 1934 streamflow is not available from the 1981–2010 calibration records of the CBRFC model, so we estimate it by scaling monthly flows from the lowest year in the calibration period (1992) to the 1934 observed volume. This monthly ratio was then used to estimate 1934 temperature-perturbed streamflow from 1992 simulations.

The second scenario, SS2, incorporates information on paleodroughts derived from tree-ring records. Preliminary reconstructions of annual streamflow for the watersheds within SLC's supply area indicate that both the 1934 and 2000–04 droughts were extreme events in the context of the past 600 years. The reconstructed and observed values for 1934 both fall within the lowest (first) percentile of these annual values, and the mean reconstructed and observed values for 2000–04 both fall within the second percentile of all reconstructed running 5-yr means (M. Bekker 2013, personal communication).

The third scenario, SS3, represents the loss of Deer Creek Reservoir from the system. In addition to the SLC share of Deer Creek Reservoir storage, the reservoir also delivers CUP water, and its loss would represent a significant loss of supply, particularly for late-season demands. The reservoir is vulnerable to higher temperatures—for example, a warmer climate combined with low reservoir levels may increase water temperatures leading to higher probability of algae blooms, which, in the extreme, could necessitate taking Deer Creek Reservoir offline.

3.4. Future water demand test cases

Two population growth assumptions were selected to drive all future water demand scenarios: steady growth at the observed rate (+0.86% yr−1) and faster growth at the rate of the fastest five consecutive years of observed growth (+1.91% yr−1). Seasonal variability in observed outdoor water use is correlated with mean monthly or April–October PET, temperature, and precipitation, with seasonal April–October r2 values of 0.49, 0.50, and 0.57, respectively. Eight future water demand cases (D1–D8 in Table 4) were developed based on different assumptions about future per capita water use. The first five cases, D1–D5, combine indoor and outdoor and incorporate a range of per capita water conservation scenarios. Cases D6–D8 calculate outdoor use separately and incorporate the observed relationship between outdoor water use and temperature and the observed 1970–2011 trend in April–October temperatures (+2.8°F by 2050; Table 2). We chose to use the temperature relationship in future demand scenarios, rather than PET or precipitation, because of greater confidence in magnitude and direction of its future trends. The 8 future water demand cases were coupled with the two population growth estimates, generating a total of 16 potential water-use test cases further described in Table 4.

Table 4.

Water demand test case descriptions and summary. D1–D5 combine indoor and outdoor water use and use past 10-yr averages as starting points and are multiplied by population trend. D6–D8 treat indoor and outdoor water use separately and apply temperature relationships to outdoor use. D1–D8 are then multiplied by two population growth trends to yield a total of 16 test cases plotted in Figures 12 and 13.

Table 4.

4. Results

4.1. Streamflow sensitivity to warming

Modeled streamflow impacts from increased temperatures indicated a reduction in annual volume and a shift to earlier runoff at higher temperatures in all watersheds of interest to SLC. Higher-elevation watersheds were less sensitive to changes in temperature and precipitation than lower-elevation watersheds.

Figure 6 shows simulated changes to the mean monthly hydrograph with each temperature forcing on Big Cottonwood Creek, which is the largest of the Four Creeks and representative of their modeled responses to increasing temperatures. Mean sensitivity of annual flow volume among all seven watersheds and across all temperature perturbations was −3.8% °F−1, and these sensitivities tended to be linear across multiple temperature perturbations. Table 5 shows that, across the Four Creeks, sensitivity varied significantly. In all Four Creeks, temperature increases reduce streamflow in the high water demand months of June–October. Monthly mean streamflow sensitivities to temperature changes are plotted for Four Creeks in Figure 7 The PWD system water supply is less impacted by changes in timing of runoff due to significant reservoir storage (results not shown).

Figure 6.
Figure 6.

Big Cottonwood Creek runoff sensitivity to temperature as indicated by 30-yr mean (water years 1981–2010) monthly runoff volumes forced by various temperature changes. Temperature changes (°F) are indicated in the legend, where “base” signifies base climate historical simulation with no temperature adjustment. Also shown is the observed mean monthly streamflow.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Table 5.

Mean watershed sensitivities, expressed as percentage change in annual flow volume ΔQ per °F change in temperature Δ°F or ΔQ per percentage change in precipitation ΔP. Combined averages are weighted by average flow contribution.

Table 5.
Figure 7.
Figure 7.

Mean monthly flow sensitivities as percentage decrease in flow per degree Fahrenheit, for each of the Four Creeks individually and combined.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Even without changes in precipitation, warming of any magnitude results in decreases in flow. Most of the reduction in flow can be attributed to increased losses due to actual evapotranspiration (ET) during periods when soil moisture is available. ET increases largely because of shortening of the snow-covered season and consequent lengthening of the growing season, as well as increased temperatures. Most ET losses occur during the wetter portion of the season when runoff is higher. ET in the CBRFC model parameterization is constrained by PET and moisture availability. Model projections for midcentury warming are generally in the range 2°–6°F, which would correspond to decreases in annual volume ranging from −4% to −37% (Figure 8). The Four Creeks in particular indicate a strong relationship between watershed elevation and sensitivity to temperature increases that is not as clearly apparent in the much larger PWD drainages (Table 1).

Figure 8.
Figure 8.

Mean annual volume decrease under various projected temperature changes for the seven watersheds used by SLC. (PET is perturbed by same temperature increase.) Abbreviations are as follows: CK for City Creek; PC for Parleys Creek; BCW for Big Cottonwood Creek; LCW for Little Cottonwood Creek; and Duches for the Duchesne River.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

These results (indicating diminished flows with warming) agree qualitatively with other regional studies such as the West-Wide Climate Risk Assessment (Gangopadhyay and Pruitt 2011) and the Joint Front Range Climate Change Vulnerability Study (JFRCCVS; Woodbury et al. 2012). However, direct comparisons are difficult since most studies, besides portions of the JFRCCVS, evaluated the impacts of combined changes in precipitation and temperature and allowed for interannually and seasonally varying future climate changes.

4.2. Precipitation sensitivities

Precipitation has a greater impact on runoff volume than temperature, and the partitioning of precipitation between rain and snow has large impacts on the timing and runoff efficiency of a basin. Thus increased precipitation may offset temperature-induced volume changes. On average, annual runoff volume in SLC's watersheds varied approximately 1.9% per percent change in precipitation, similar to runoff elasticities to precipitation found elsewhere in the southwestern United States (Hoerling et al. 2009; Woodbury et al. 2012). For example, from an annual average perspective, the precipitation increase needed to offset flow impacts from a 5°F increase in temperature is in excess of 10% for a lower-elevation watershed and less than 10% for a higher-elevation watershed. As noted earlier, the CMIP3 BCSD-projected annual precipitation changes midcentury vary significantly, implying that future precipitation trends remain an area of significant uncertainty and that precipitation changes could either mitigate temperature effects on annual flow volume or greatly exacerbate them. Projections have better agreement on the likelihood of increasing winter and decreasing summer precipitation (Table 3). When combined with increasing temperatures, these would cause additional increases in winter flows and further reduction in summer flows, thereby degrading the ability of direct runoff to meet summer demands and increasing SLC's reliance on groundwater and reservoir storage.

4.3. PET sensitivities

The static PET approach in the current operational CBRFC model would be expected to underestimate streamflow reductions under warming and ET increases because of decreases in the snow-cover season. We confirm that the CBRFC model runs for SLC's watersheds using the dynamic PET described earlier indicate significant reductions in flows as compared to the static PET operational model. For instance, the mean annual flow reduction is doubly severe on Big Cottonwood Creek with the dynamic PET driver (comparison not shown). Whereas the relative flow decrease varies from basin to basin, the temperature-driven flow sensitivity using dynamic PET is generally about twice that of the static PET simulations.

4.4. Runoff timing

Our model results indicate that warming leads to significant changes in the timing of runoff in all of SLC's surface water supply sources. Precipitation changes alone, uniformly applied, have only minor impacts to runoff timing (Hamlet et al. 2005; Clow 2010). Earlier runoff, absent reductions in volume, results in a larger gap between natural runoff and late-summer season demands, necessitating the use of other water sources to make up the shortfall.

The centroid of runoff volume—defined as the day when half of the water year volume has passed a gauge location—provides a good indicator for changes in streamflow timing. Modeled warming-driven changes in the centroid for Big Cottonwood Creek are presented in Table 6. On average, the sensitivity of timing for Big Cottonwood Creek is 3 days earlier per degree Fahrenheit of warming. Our findings are in line with both the JFRCCVS, which projected 0–18 days earlier runoff timing for 2040, depending on the basin and climate scenario (Woodbury et al. 2012), and Stewart et al. (Stewart et al. 2005), who observed 1948–2002 trends in Utah in the range of 0–3.6 days earlier per decade.

Table 6.

Projected changes in the runoff centroid timing (days) for various temperature increases (°F) in absence of precipitation changes for Big Cottonwood Creek.

Table 6.

4.5. Future water supply scenarios

Scenarios of extreme drought exacerbated by warming (SS1) indicate the potential for experiencing future flows significantly below the observed single-year and multiyear droughts of record. Single-year, warming-induced flow reductions, estimated from the lowest year in the calibration period (1992) for the Four Creeks, are approximately −3.6% °F−1. Multiyear flow–temperature sensitivity on the PWD, based on the 2000–04 drought, which was more critical on this system, was −4.5% °F−1. Thus, independent of possible changes in the overall precipitation regime, the potential exists for midcentury (2035–64) 1-yr droughts on the Four Creeks of approximately 4%–22% below existing record low flows (Figure 9). The estimated drought from preliminary tree-ring records (SS2) would be roughly equivalent to a 3°F induced increase to the 1-yr drought on the Four Creeks or 2°F for the 5-yr drought on the PWD (Figure 10). An event of this magnitude on both the PWD and the Four Creeks system simultaneously would place extraordinary stress on SLC's system. Estimating potential water shortages for this scenario is difficult, however, because of uncertainty in management and allocations on the PWD system, which are outside of SLC's direct control.

Figure 9.
Figure 9.

SS1: the driest single-year simulation for Four Creeks, with combined monthly volumes for 1992 and 1934 indexed from 1992 and reduced by temperature increases of 3° and 5°F. SS2: the paleoestimate is a 10% reduction of the 1934 record based on the draft local paleohydrologic reconstructions. Observations for 1981–2010 are the CBRFC naturalized flow average, and the base simulation is the CBRFC model base simulation for the 1981–2010 period with no temperature perturbation.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Figure 10.
Figure 10.

SS1: Provo–Weber–Duchesne driest 5-yr drought average flows with +3° and +5°F forcing SS2 with estimated 5-yr paleodrought from draft tree-ring reconstruction.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

SLC losing the Deer Creek and CUP supplies (SS3) would result in significant supply deficits in late summer (and potentially winter), likely leading to impacts to water service and disruption. If the loss occurred during an extreme drought condition on the Four Creeks, this would result in deficits during all months except April and May (Figure 11). Deficits would far exceed water available from local storage and SLC's current groundwater sources, which have maximum historic annual withdrawals of 16 KAF. Note that the SS3 scenarios did not account for warming-induced flow reductions, which would increase the deficit between available supply and demand. A summary of key findings from future water supply scenarios and streamflow sensitivities, along with potential water supply impacts as estimated with assistance from SLC, can be found in Table 7.

Figure 11.
Figure 11.

SS3: low year Deer Creek/CUP loss scenario. The blue line indicates supply assuming 1934 streamflow without temperature forcing and with maximum historical deep well use. The red line is SLC's observed average demand.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Table 7.

Key sensitivity and water supply scenario findings (percentage runoff changes are to annual flow).

Table 7.

4.6. Water demand test cases

Figures 12 and 13 display the results of the water demand test cases, which are also summarized in Table 4. Future demand is dominated by assumptions of population growth and water conservation rates, but climate also plays an important role. Water demand sensitivity to temperature appears to be on the order of 3 KAF °F−1. Current rates of reduction in per capita water-use reduction (test case D2) cannot be reasonably expected to continue indefinitely into the future, since the current trend intercepts zero in approximately 70 years. Determining how long this rate can be continued without damaging landscapes requires additional research.

Figure 12.
Figure 12.

Steady population growth demand cases (see Table 4 for case description). Note the D2 and D6 cases are based completely or in part on changes in per capita water consumption observed from 1988 to 2011. The D2 trend is not expected to continue indefinitely because of practical constraints, but further research is required to determine an inflection point in the trend.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

Figure 13.
Figure 13.

Fast population growth water demand cases (see Table 4 for case description). Note the D2 and D6 cases are based completely or in part on changes in per capita water consumption observed from 1988 to 2011. The D2 trend is not expected to continue indefinitely because of practical constraints, but further research is required to determine an inflection point in the trend.

Citation: Earth Interactions 17, 23; 10.1175/2012EI000501.1

5. Discussion

5.1. Uncertainties and limitations in sensitivity analyses

The analyses presented above that consider the impact of temperature changes on PET (and therefore on water demand timing) rely on a number of simplifying assumptions about variability due to and stationarity in the other PET drivers. First, although temperature is the dominant driver of interannual variability in PET for SLC, specific humidity, wind speed, and solar radiation are also influential (Figure 5) (Hobbins et al. 2012). In fact, during the summer months crucial to water demand, wind speed and specific humidity dominate PET variability. Reliable projections of humidity and wind speed changes are not available from the CMIP3 archives (for details, see Maurer et al. 2007), however, and such influences on this study's results remain an uncertainty.

Second, recent temperature increases in the SLC delivery area have not increased estimated PET, even considering all drivers as described earlier. Observations from 1960 to 2001 indicate that, although the frost-free season in SLC has lengthened by 1.2 days yr−1, there is as yet no observed lengthening of the outdoor watering season (Table 2). This may be because climate-driven impacts to the observed length of the outdoor watering season are significantly impacted by and difficult to untangle from SLC's water conservation efforts.

Finally, future timing of runoff will be influenced by changes in land use, albedo, and sublimation rates, which are not considered in these analyses or explicitly in the operational CBRFC model. We also acknowledge the limitations inherent in our modeling framework: although the CBRFC model has been calibrated to the past 30 years of observations, its ability to accurately simulate streamflow for climate conditions significantly different from those occurring in the calibration period is unknown.

5.2. Adaptive strategies

SLC's ultimate goal in developing these sensitivity analyses is to understand hydrologic sensitivities relevant to the next phases of climate-related adaptation planning efforts for their specific water resource management system. Some general water resource management strategies, such as watershed protection and water conservation, are already employed to protect and optimize SLC's water resources and are likely to provide additional buffers against future supply and demand mismatches if continued into the future. Because the sensitivity analyses suggest that increasing temperatures will lead to decreasing water supplies, especially during SLC's traditional high demand times, it is prudent for SLC to consider additional management strategies to ensure system reliability in an uncertain future.

The sensitivity analyses have also raised questions for SLC about identifying and implementing other future adaptive strategies that could become necessary to meet future demand in climate change scenarios over time, including changes that may be more costly and complex. These “high impact” strategies may include reservoir expansion, additional groundwater resources, aquifer storage and recovery, wastewater reuse, and the purchase or development of additional water sources. Additional iterative research and strategic focus on a range of potential climate scenario futures will be the subject of ongoing work by SLC and its partners.

6. Conclusions and next steps

The effort described herein has advanced the understanding of potential vulnerability of climate change on water supplies in northern Utah to climate variation and change, demand change, and given operational adaptation options. The seven watersheds evaluated for sensitivities to changes in precipitation and temperature represent important water supplies far beyond those serving Salt Lake City. Results from these analyses align with other regional studies indicating a shift to earlier runoff with warming and the tendency to decrease total runoff volume even absent changes in precipitation. Tree-ring records and analyses of temperature-induced reductions in flow indicate the possibility of short and long-term droughts well beyond any severity in the observed record. We believe that such altered streamflow regimes should lead water managers to reconsider their drought planning, long-term sustainable yields, and storage requirements in order to meet late-season demands in an uncertain future. SLC and other neighboring and regional water managers will need to continue to study potential impacts of climate variability and change to their water supply infrastructure.

This work lays a foundation for follow-on efforts by the study partners toward the development of applied and actionable science to support long-range planning and infrastructure design needs and adaptive responses. New tools and products under development are intended to add sophistication and detail to these analyses of SLC's system. For example, an expanded range of sensitivity and scenario analyses using CMIP projections and dynamically downscaled high-resolution regional climate model projections will better represent hydroclimatic sensitivities given the extreme topographic relief of the region's watersheds, and a water system planning model in development will aid in testing a range of future runoff, demand, and operational scenarios to help plan for uncertainty. There is also a continuing need to evaluate climate sensitivities to many parameters not adequately addressed in this study or others, including changes in land use, evapotranspiration, and water demand. Finally, this sensitivity assessment focused solely on municipal water supply and demand, yet SLC and many other utilities concerned with climate change impacts are also interested in understanding possible ramifications for storm water runoff and water quality, among other parameters.

Acknowledgments

The authors gratefully acknowledge technical and office support provided by the NWS CBRFC, especially Craig Peterson, John Lhotak, Michelle Stokes, and Kevin Werner. We are also grateful to our two anonymous reviewers and internal reviews from Eric Gordon and Jeff Lukas (WWA) and graphics assistance from Ami Nacu-Schmidt (WWA). Jeff Lukas and Matthew Bekker (Brigham Young University) provided draft tree-ring chronologies. This material is partially based upon work completed by the CI-WATER project supported by the National Science Foundation under Grant 1135483. Partial funding for this effort came from the Western Water Assessment RISA program based at the University of Colorado, Boulder, under NOAA Grant NA10OAR4310214.

References

  • Anderson, E. A., 1973: National Weather Service River Forecast System: Snow Accumulation and Ablation Model. NOAA Tech. Memo. NWS Hydro-17, 87 pp.

  • Barnett, T. P., , J. C. Adam, , and D. P. Lettenmaier, 2005: Potential impacts of a warming climate on water availability in snow-dominated regions. Nature, 438, 303309.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., and Coauthors, 2008: Human-induced changes in the hydrology of the western United States. Science,319, 1080–1083, doi:10.1126/science.1152538.

  • Brown, C., , and R. L. Wilby, 2012: An alternate approach to assessing climate risks. Eos, Trans. Amer. Geophys. Union,92, 401–402.

  • Bureau of Reclamation, 2011: SECURE Water Act section 9503(c): Reclamation Climate Change and Water 2011. Bureau of Reclamation Rep., 226 pp. [Available online at http://www.usbr.gov/climate/SECURE/docs/SECUREWaterReport.pdf.]

  • Bureau of Reclamation, cited 2012: Colorado River Basin Water Supply and Demand Study. [Available online at http://www.usbr.gov/lc/region/programs/crbstudy.html.]

  • Burnash, R. J. C., 1995: The NWS River Forecast System—Catchment modeling. in Computer Models of Watershed Hydrology, V. P. Singh, Ed., Water Resources, 311–366.

  • Burnash, R. J. C., , R. L. Ferral, , and R. A. McGuire, 1973: A generalized streamflow simulation system: Conceptual modeling for digital computers. Joint Federal and State River Forecast Center, U.S. National Weather Service, and California Department of Water Resources Tech. Rep., 204 pp.

  • Chambers, J. C., 2008: Climate change and the Great Basin. USDA Forest Service General Tech. Rep. RMRS-GTR-204, 4 pp.

  • Clow, D. W., 2010: Changes in the timing of snowmelt and streamflow in Colorado: A response to recent warming. J. Climate, 23, 22932306.

    • Search Google Scholar
    • Export Citation
  • Craig, R. K., 2010: “Stationarity is dead”—Long live transformation: Five principles for climate chang adaptation law. Harvard Environ. Law Rev., 34, 9–75.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , R. Knutti, , S. Solomon, , and A. S. Phillips, 2012: Communication of the role of natural variability in future North American climate. Nat. Climate Change, 2, 775779, doi:10.1038/nclimate1562.

    • Search Google Scholar
    • Export Citation
  • Farnsworth, R. K., , E. S. Thompson, , and E. L. Peck, 1982: Evaporation atlas for the contiguous 48 United States. NOAA Tech. Rep. NWS 33, 37 pp. [Available online at http://www.weather.gov/oh/hdsc/PMP_related_studies/TR33.pdf.]

  • Gangopadhyay, S., , and T. Pruitt, 2011: West-wide climate risk assessments: Bias-corrected and spatially downscaled surface water projections. Bureau of Reclamation Tech. Memo. 86-68210-2011-01, 138 pp.

  • Gillies, R. S., , S. Wang, , and M. Booth, 2012: Observational and synoptic analyses of the winter precipitation regime over Utah. J. Climate, 25, 46794698.

    • Search Google Scholar
    • Export Citation
  • Hamlet, A. F., , P. W. Mote, , M. P. Clark, , and D. P. Lettenmaier, 2005: Effects of temperature and precipitation variability on snowpack trends in the western United States. J. Climate, 18, 45454561.

    • Search Google Scholar
    • Export Citation
  • Harding, B. L., , A. W. Wood, , and J. R. Prairie, 2012: The implications of climate change scenario selection for future streamflow projection in the upper Colorado River basin. Hydrol. Earth Syst. Sci., 16, 39894007, doi:10.5194/hess-16-3989-2012.

    • Search Google Scholar
    • Export Citation
  • Hobbins, M., , A. W. Wood, , D. Streubel, , and K. Werner, 2012: what drives the variability of evaporative demand across the conterminous United States? J. Hydrometeor., 13, 11951214.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M., , D. Lettenmaier, , D. Cayan, , and B. Udall, 2009: Reconciling projections of Colorado River streamflow. Southwest Hydrol., 8, 2031.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , J. M. Melillo, , and T. C. Peterson, Eds., 2009: Global Climate Change Impacts in the United States. Cambridge University Press, 188 pp.

  • Knowles, N., , M. D. Dettinger, , and D. R. Cayan, 2006: Trends in snowfall versus rainfall in the western United States. J. Climate, 19, 45454559.

    • Search Google Scholar
    • Export Citation
  • Maurer, E. P., , L. Brekke, , T. Pruitt, , and P. B. Duffy, 2007: Fine-resolution climate projections enhance regional climate change impact studies. Eos, Trans. Amer. Geophys. Union, 88, 504.

    • Search Google Scholar
    • Export Citation
  • Mills, E., , E. Klotz, , T. Stonely, , M. Waters, , and L. Summers, 2003: Utah's M&I water conservation plan. Utah Division of Water Resources Rep., 44 pp. [Available online at http://www.conservewater.utah.gov/Final71403AACC.pdf.]

  • Milly, P. C. D., , K. A. Dunne, , and A. V. Vecchia, 2005: Global pattern of trends in streamflow and water availability in a changing climate. Nature, 438, 347350.

    • Search Google Scholar
    • Export Citation
  • Milly, P. C. D., , J. Betancourt, , M. Falkenmark, , R. M. Hersh, , Z. W. Kundzewicz, , D. P. Lettenmaier, , and R. J. Stouffer, 2008: Stationarity is dead: Whither water management? Science, 348, 573574.

    • Search Google Scholar
    • Export Citation
  • Mitchell, K. E., and Coauthors, 2004: The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. J. Geophys. Res., 109, D07S90, doi:10.1029/2003JD003823.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., 2006: Climate-driven variability and trends in mountain snowpack in western North America. J. Climate, 19, 62096220.

  • Nash, L. L., , and P. Gleick, 1991: The sensitivity of streamflow in the Colorado Basin to climatic changes. J. Hydrol., 125, 221241.

  • Ray, A. J., , J. J. Barsugli, , K. B. Averyt, , K. Wolter, , M. Hoerling, , N. Doesken, , B. Udall, , and R. S. Webb, 2008: Climate change in Colorado: A synthesis to support water resources management and adaptation Colorado Water Conservation Board Western Water Assessment Rep., 58 pp.

  • Rotstayn, L. D., , M. L. Roderick, , and G. D. Farquhar, 2006: A simple pan-evaporation model for analysis of climate simulations: Evaluation over Australia. Geophys. Res. Lett., 33, L17715, doi:10.1029/2006GL027114.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , M. P. Clark, , R. L. Armstrong, , D. A. McGinnis, , and R. L. Pulwarty, 1999: Characteristics of the western U.S. snowpack from Snowpack Telemetry (SNOTEL) data. Water Resour. Res., 35, 21452160.

    • Search Google Scholar
    • Export Citation
  • Stewart, T. S., , D. R. Cayan, , and M. D. Dettinger, 2005: Changes toward earlier streamflow timing across western North America. J. Climate, 18, 11361155.

    • Search Google Scholar
    • Export Citation
  • Wood, A. W., , L. R. Leung, , V. Sridhar, , and D. P. Lettenmaier, 2004: Hydrologic implications of dynamic and statistical approaches to downscaling climate model outputs. Climatic Change, 62, 189216.

    • Search Google Scholar
    • Export Citation
  • Woodbury, M., , M. Balso, , D. Yates, , and L. Kaatz, 2012: Joint Front Range Climate Change Vulnerability Study. Water Research Foundation Rep., 148 pp.

  • Woodhouse, C. A., , D. M. Meko, , G. M. MacDonald, , D. W. Stahle, , and E. R. Cook, 2010: A 1,200-year perspective of 21st century drought in southwestern North America, Proc. Natl. Acad. Sci. USA, 107, 21 28321 288, doi:10.1073/pnas.0911197107.

    • Search Google Scholar
    • Export Citation
Save