National Park Service U. S. Department of the Interior Natural Resource Program Center

Figure 19.

Temperature increases are more pronounced during the cool season (Hamlet and Lettenmaier 2007). In the northern U.S. Rocky Mountains, annual rates of increase are roughly 2–3 times that of the global average (Vose et al. 2005; Bonfils et al. 2008; Pederson et al. 2010; Hall and Fagre 2003), a pattern that is evident at northern latitudes and higher elevation sites throughout the West (Diaz and Eischeid 2007). In addition, nighttime minimum temperatures are increasing faster than are daytime maximums, resulting in a decreased diurnal temperature range (Pederson et al. submitted). This has implications for species like the mountain pine beetle (Dendroctonus ponderosae), whose population dynamics are governed by minimum temperatures (Carroll et al. 2004). Mean regional spring and summer temperatures were 0.87ºC higher in 1987 to 2003 than in 1970 to 1986, and were the warmest since 1895 (Westerling et al. 2006). Bonfils et al. (2008) and Barnett et al. (2008) found that the recent warming observed in mountainous areas across the West cannot be entirely explained by natural forcings (e.g., solar, volcanic, and ocean-atmosphere interactions); a major portion is attributable to human-influenced changes in greenhouse gas and aerosol concentrations.

Trends in precipitation for the study area are far less clear. Instrumental data from the last century show modest increases in precipitation for much of the northwestern United States (fig. 20; Mote et al. 1999, 2003, 2005), but no trends in parts of the southern Rockies (Ray et al. 2008). Natural variability in precipitation is evident in the instrumental record for all of the climate regions, and long-term drought conditions during the last century impacted large areas. Though 20^th century droughts had substantial socioeconomic and ecosystem impacts, there is ample evidence they were not as severe in duration and magnitude as a number of drought events that occurred during the last millennium (Cook et al. 2007, 2004; Meko et al. 2007). For example, two droughts in the 1930s and 1950s impacted much of the study area. The 1930s drought was more widespread and pronounced in the northern and central climate regions while the 1950s drought was centered more on the south-central and southwestern United States (Cook et al. 2007; Gray et al. 2004; Fye et al. 2003). Research suggests that climatic conditions that influenced the nature and location of these droughts are likely linked to low-frequency oscillations in ocean-atmosphere interactions (McCabe et al. 2004; Gray et al. 2007; Hidalgo 2004; Graumlich et al. 2003), with evidence for substantial surface feedbacks during the 1930s “Dust Bowl” drought (Cook et al. 2009).

Surface hydrology

Generally speaking, snowmelt and peak runoff has tended to occur earlier since 1950, and river flows in many locations are decreasing during late summer (fig. 21; Pederson et al. 2010, 2009; Mote 2006; Barnett et al. 2008; Stewart et al. 2004; McCabe and Clark 2005). Recent impacts on snowpack and surface hydrology are strongly associated with more precipitation falling as rain than snow, and warming temperatures driving earlier snowmelt and snowmelt-driven runoff (fig. 21; Pederson et al. 2010, 2009; Mote 2006; Barnett et al. 2008; Stewart et al. 2004; McCabe and Clark 2005), leading to reduced surface storage of moisture and increasingly low baseflows during the dry summer months (Luce and Holden 2009). Examination of paleoclimate data and instrumental records (e.g., stream gages, snow course records, valley meteorological stations) suggests, however, that the total amount of cool season precipitation received across a particular region is more strongly associated with natural multidecadal, decadal, interannual, and intra-annual variability in ocean-atmosphere conditions (e.g., PDO, AMO, and ENSO).

Modest increases in precipitation have occurred in parts of the central and northern study area, but modest declines have taken place in parts of the southern Rockies, and significant declines in snowpack are evident throughout much of the Northwest. This decline is especially prevalent in the northern U.S. Rocky Mountains and parts of the Upper Columbia Basin (Hamlet and Lettenmaier 2007; Pederson et al. 2004, 2010; Selkowitz et al. 2002; Mote et al. 2005, 2008). Warmer temperatures and declining snowpack have, in turn, contributed to significant declines in the region’s glaciers. In Glacier National Park, glaciers have decreased in area by more than 60% since 1900 (Hall and Fagre 2003; Key et al. 2002), and only 26 of the 150 glaciers and snow and ice fields present in 1910 remain (Pederson et al 2010). Changes in glacier mass and area are emblematic of changes in surface hydrology across the West, with the recent substantial declines related to increases in greenhouse gases and aerosols (Barnett et al. 2008; Bonfils et al. 2008; Pierce et al. 2008; Hidalgo et al. 2009). During the last century, drought conditions have been increasing in the central and southern parts of the study area and decreasing in the northern parts (Andreadis and Lettenmaier 2006). Drought conditions are expected to increase for much of the study area over the coming decades (Hoerling and Eischeid 2007).

Figure 21.

Ocean-atmosphere interactions are important drivers of interannual to multidecadal variability in temperature and precipitation (Pederson et al. 2010, 2009; Gray et al. 2007, 2004, 2003; McCabe et al. 2004; Hidalgo 2004; Cayan et al. 1998; Dettinger et al. 2000), but their impacts vary greatly across latitudinal, elevational, and longitudinal gradients. The El Niño–Southern Oscillation and Pacific Decadal Oscillation measures of high and low frequency variability in sea surface temperatures (SSTs) are, respectively, the predominant sources of interannual and interdecadal climate variability for much of the study area (Mantua et al. 2002, 1997). ENSO variations are commonly referred to as El Niño (the warm phase) or La Niña (the cool phase). An El Niño event is characterized by warmer than average sea surface temperatures in the central and eastern equatorial Pacific Ocean, reduced strength of the easterly trade winds in the tropical Pacific, and an eastward shift in the region of intense tropical rainfall (Bjerknes 1969; Cane and Zebiak 1985; Graham and White 1988; Horel and Wallace 1981; Philander 1990; Chang and Battisti 1998). A La Niña event is characterized by the opposite: cooler than average sea surface temperatures, stronger than normal easterly trade winds, and a westward shift in the region of intense tropical rainfall. Warm and cool phases typically alternate on 2 to 7 year cycles (Bjerknes 1969; Cane and Zebiak 1985; Graham and White 1988; Horel and Wallace 1981; Philander 1990; Chang and Battisti 1998).

The PDO exhibits alternate cool and warm phases with a spatial pattern similar to that of ENSO, but these phases typically last for 20 to 30 years (Mantua et al. 1997). Several switches occurred between warm and cool PDO phases during the 20^th century and the magnitude of PDO phases increased in the latter half (McCabe et al. 2004; Mantua et al. 2002, 1997). The Atlantic Multidecadal Oscillation (AMO), representing low frequency (50–80 yr) oscillations in North Atlantic SSTs, has been linked to multidecadal variability in temperature and precipitation in the western United States through complex interactions with the PDO and ENSO, but the magnitude of the AMO influence is debated (Kerr 2000; McCabe et al. 2004; Enfield et al. 2001).

Changes in ENSO and PDO impact precipitation differently across the West, with winter precipitation in the Upper Columbia Basin and the northern Rockies being negatively correlated with warm conditions in the equatorial Pacific (i.e., during El Niños). Parts of the central and southern Rockies tend to be wetter than average during El Niños and dryer during La Niñas (Mote et al. 2005). Likewise, the Northwest generally receives more winter precipitation during the cool-phase PDO, and the Southwest often receives more precipitation during the warm phase. Further evidence for spatial heterogeneity in ENSO and PDO impacts can be seen in Greater Yellowstone, where windward aspects and the high mountains and plateaus to the west tend to follow more of an Upper Columbia Basin and northern Rockies pattern, while locations to the leeward and east often behave like the central and southern Rockies (Gray et al. 2004).

The PDO has two phases: warm (positive index value) and cool (negative index value). Figure 1 shows the sea surface temperature anomalies associated with the warm phase of the PDO and ENSO, both of which favor anomalously warm sea surface temperatures near the equator and along the coast of North America, and anomalously cool sea surface temperatures in the central north Pacific. The cool phases for PDO and ENSO (not shown) have the opposite patterns: cool along the equator and the coast of North America and warm in the central north Pacific. Each PDO phase typically lasted for 20 to 30 years during the 20^th century, and studies indicate that the PDO was in a cool phase from approximately 1890 to 1925 and 1945 to 1977 (Mantua 1997, 2002). Warm phase PDO regimes existed from 1925 to 19456 and from 1977 to at least 1998. Pacific climate changes in the late 1990s, in many respects, suggested another reversal from warm to cool phase and possibly back to warm.

Figure 1. Warm phase PDO and ENSO. The spatial pattern of anomalies in sea surface temperature (shading, °C) and sea level pressure (contours) associated with the warm phase for 1900–1992. Note that the main center of action for the PDO (left) is in the north Pacific, and for the ENSO (right) in the equatorial Pacific. Contour interval is 1 millibar, with additional contours drawn for +0.25 and 0.5 mb. Positive contours are dashed; negative contours are solid. Source: Climate Impacts Group, University of Washington.
Natural variation in the strength of PDO and ENSO events impacts climate regions in different ways. In the Northwest and parts of the central Rockies, warm-phase PDO and El Niño winters tend to be warmer and drier than average, with below normal snowpack and streamflow, whereas La Niña winters tend to be cooler and wetter than average, with above normal snowpack and streamflow (Graumlich et al. 2003; Cayan et al. 1998). In the southern Rockies and the Southwest, warm-phase PDO and El Niño winters tend to be wetter than average, with above normal snowpack and streamflow, and La Niña winters tend to be drier than average, with below normal snowpack and streamflow (Cayan et al. 1998; Swetnam and Betancourt 1998; Dettinger and Ghil 1998; Mote 2006).

Figure 2. Multivariate ENSO index, 1950–2009. Positive (red) values indicate an El Niño event; negative (blue) values a La Niña event (Wolter and Timlin 1998, 1993).

Simulations of 21^st century climate suggest a northward movement of the storm tracks that influence precipitation over much of the western United States (Yin 2005; Lorenz and DeWeaver 2007), which could reduce precipitation for large parts of the study area. McAfee and Russell (2008) show that a strengthening of the Northern Annual Mode (an index of sea level pressure poleward of 20ºN), which results in a poleward displacement of the Pacific Northwest storm track, increased west to east flow, reduced spring precipitation west of the Rockies, and increased spring precipitation east of the Rockies (McAfee and Russell 2008). This shift in the storm track is expected to persist well into the future and may reduce the length of the cool season, when circulation patterns provide the bulk of precipitation for large areas of the central and northern parts of the study area (McAfee and Russell 2008). If this becomes a more permanent shift in the storm-track position, it could lead to a longer duration of warm-season conditions (i.e., predominately warm and dry) for the Upper Columbia Basin, northern Rockies, and parts of the central Rockies. Changes in storm-track position and circulation patterns will be superimposed on a background of natural variability. Thus, various combinations of ENSO, PDO, and broad-scale trends could lead to local impacts that vary greatly in their magnitude over time.

Recent changes such as warming temperatures and associated declines in snowpack and surface-hydrology are already influencing ecosystem dynamics. Examples observed in the last century include: earlier spring blooming and leaf-out, forest infilling at and near the treeline, and increased severity of disturbances such as wildfire and insect outbreaks, all of which are likely to continue with additional warming. Spring blooming of a number of plant species has occurred earlier throughout much of the western United States, in some cases by as much as several weeks (Cayan et al. 2001; Schwartz and Reiter 2000; Schwartz et al. 2006). In the northern Rockies, increased density of trees at or near treeline has been observed at some sites (Butler et al. 2009). This “infill” phenomenon is not uncommon in the West and is predicted to continue where minimum temperatures rise, snowpack in high-snowfall areas decreases, and moisture is not limiting (Graumlich et al. 2005; Lloyd and Graumlich 1997; Rochefort et al. 1994; Millar et al. 2009). While evidence for infill is widespread, upslope movement in treeline position is much more variable, and research suggests that it will be characterized by a high degree of spatial heterogeneity in relation to other variables that control treeline position, e.g., aspect, soils, and micro-topography (Lloyd and Graumlich 1997; Graumlich et al. 2005; Bunn et al. 2005, 2007).

Changing climate conditions are also influencing disturbance processes that regulate ecosystem dynamics. Warming temperatures, earlier snowmelt, and increased evapotranspiration are increasing moisture stress on forest species and making them more susceptible to insect attack. An increase in the extent, intensity, and synchronicity of mountain pine beetle attacks in the western United States and Canada has been linked to forests stressed by drought, which makes trees less able to resist infestations (Nordhaus 2009; Hicke and Jenkins 2008; Romme et al. 2006; Logan et al. 2003; Carroll et al.2004; Breshears et al. 2005). Warming temperatures have also influenced bark beetle population dynamics though reduced winter kill and by facilitating reproduction and dispersal (Carroll et al. 2004; Black et al. 2010). In some cases, past forest management (e.g., factors related to structural characteristics of host stands) may also facilitate beetle infestation (Nordhaus 2009; Black et al. 2010). The rate of fire disturbance is also increasing across the West, particularly in the northern Rockies (fig. 22, Westerling 2008).



Figure 22. Annual number of forest fires >1000 acres (total column height) in the northern Rockies (black area) and other western states. Source: Westerling 2008; reprinted with permission.

The extent of the western United States burned in wildfires each year is strongly linked to interannual climate variability (Littell et al. 2008, 2009a; Morgan et al. 2008; Higuera et al. 2010). Changes in surface hydrology associated with reduced snowpack, earlier spring runoff and peak flows, diminished summer flows, and a lengthening fire season have all been linked to increased frequency of large fires, with the most evident impacts at mid-elevation forests in the northern Rockies since the mid-1980s (Westerling et al. 2006, fig. 22).

Over the course of the 20^th century, the instrumental record in the northern Rockies showed a significant increase in average seasonal, annual, minimum, and maximum temperatures (figs. 23, 24; Loehman and Anderson 2009; Pederson et al. 2010, submitted). Regional average annual temperatures increased 1–2C (2–4°F) from 1900 to 2000 (Pederson et al. 2010). Seasonal and annual minimum temperatures are generally increasing much faster than maximum temperatures (Pederson et al. 2010, submitted). In particular, summer and winter seasonal average minimum temperatures are increasing significantly faster than the season’s average maximum temperatures, causing a pronounced reduction in the seasonal diurnal temperature range (Pederson et al. 2010). The magnitude of minimum temperature increases also appears seasonally variable: in areas with mid-elevation snow telemetry (SNOTEL) sites, Pederson et al. (submitted) estimated minimum temperature increases since 1983 of 3.8 ± 1.72˚C (6.8 ± 3.10°F) in winter, 2.5 ± 1.23˚C (4.5 ± 2.21°F) in spring , and 3.5 ± 0.73˚C (6.3 ± 1.31°F)annually (fig. 24). The magnitude of changes varies locally, but there are few exceptions to this general warming trend.

Temperature trends within the northern Rockies generally track Northern Hemisphere trends across temporal scales (fig. 23). This similarity between regional and continental trends suggests that large-scale climate forcings such as greenhouse gases, sea surface temperature patterns, volcanic activity, and solar variability also drive regional temperatures (Pederson et al. 2010).

Throughout the West, high interannual, annual, and decadal variability in precipitation exceeds any century-long trends (Ray et al. 2008). General patterns throughout the latter part of the 20^th century indicate that areas within the northern Rockies experienced modest but statistically insignificant decreases in annual precipitation (Mote et al. 2005; Knowles et al. 2006). Although few statistically significant trends are evident in regional 20^th century precipitation, rising temperatures throughout the West have led to an increasing proportion of precipitation falling as rain rather than snow (Knowles et al. 2006). Winter temperatures well below 0C make the northern Rockies less sensitive than other western regions where small temperature increases in temperature are affecting the number of freezing days (Knowles et al. 2006).

Like most of the western United States, the snow water equivalent (SWE) of winter snowpack largely controls surface runoff and hydrology in the northern Rockies for the water year (e.g., Pierce et al. 2008;Barnett et al. 2008, Stewart et al. 2005; Pederson et al. submitted). Studies have demonstrated a statistically significant decrease in winter snowpack SWE across the region during the second half of the 20^th century (Barnett et al. 2008; Pederson et al. submitted).

The warm phase of the PDO was associated with reduced streamflow and snowpack in the northern Rockies during the 20^th century, (Fagre et al. 2003), and the cool phase with increased streamflow and snowpack (Pederson et al. submitted). One ecological response to these shifts has been changes in the distribution of mountain hemlock (Tsuga mertensiana). At high elevations, where mountain hemlock growth is limited by snowpack-free days, a warm-phase PDO often results in decreased snowpack and increased mountain hemlock growth (Peterson and Peterson 2001). At low elevation sites where moisture is limiting, a warm-phase PDO commonly leads to less moisture and consequently decreased mountain hemlock growth and establishment (Fagre et al. 2003).

Pederson et al. (submitted) summarize how variation in Pacific SSTs, atmospheric circulation, and surface feedbacks influence climate conditions, snowpack, and streamflow in the northern Rockies. Winters with high snowpack tend to be associated with the cool phase of the PDO, a weakened Aleutian Low, and low pressure centered poleward of 45°N across western North America (fig. 25). During years of high snowpack, mid-latitude cyclones tend to track from the Gulf of Alaska southeast through the Pacific Northwest and into the northern Rockies. The relatively persistent low-pressure anomaly centered over western North America is also conducive to more frequent Arctic-air outbreaks, resulting in colder winter temperatures.

The ENSO is an important driver of snowpack and streamflow at interannual scales, and the influence of related tropical Pacific atmospheric circulation anomalies persists well into the spring. Changes in spring (MAM) temperatures and precipitation are associated with changes in regional atmospheric circulation, and strongly influence the timing of streamflow in the northern Rockies (fig.25). Springtime geopotential heights over western North America influence the amount and more importantly the timing of snowmelt and streamflow across the northern Rockies. Specifically, high pressure anomalies centered over western North America are associated with higher spring temperatures and consequently an increasing number of snow-free days and earlier arrival of snow melt-out and peak streamflow. Atmospheric circulation changes in March and April can, in turn, initiate surface feedbacks that contribute to surface temperature and hydrograph anomalies (fig. 25). Hence, warming temperatures in the northern Rockies lead to earlier snowmelt and runoff and associated decreasing snowpack and streamflow, but these patterns can be partially attributed to seasonally-dependant teleconnections and atmospheric circulation patterns, as well as to surface-albedo feedbacks that interact with broad-scale controls on snowpack and runoff (Pederson et al. submitted).