Sunday, March 2, 2014

Is This Our Future?

A nearly snow free Kimball Junction, 2 March 2014
With an interesting February in the record books, I thought we would take a look back on the month and think about it in the context of the climate change that is anticipated in the coming century.

Based on observations from the Salt Lake City airport, February 2014 was very mild.  Only three days had minimum temperatures below average and only 6 had maximum temperatures below average.

Source: NWS
For the month, the average temperature was 42.2ºF, 7.9ºF above the average for 1981–2010.  In terms of precipitation, we received 1.57 inches, which is about 25% more than average.  In terms of snowfall, we got only 2.1 inches compared to an average of 10.7 inches.

And what about the snowpack?  Snow in the Salt Lake Valley was obliterated, as well as the snow at Mountain Dell, a popular area for cross country skiing.  The month was pretty much a disaster for snow at and below 7000 feet in the central Wasatch.  I've had to go higher for cross country skiing, but even the Basin Rec 5 K loop near Kimball Junction (6500–6800 feet) is barely hanging on.

Basin Rec 5 K look, Kimball Junction, UT, 2 March 2014
Granted, these aren't especially snowy areas (hence the sagebrush), but the melt out and tendency for rain at these elevations is more like one might expect in mid-to-late March rather than February.  

In contrast, above 8000 feet it's been a pretty good month, not just in the central Wasatch but the northern Utah mountains as a whole.  Snowpack snow-water equivalent has increased by at least several inches at most locations. 

To be sure, I don't want to suggest that the February weather was caused by global warming.  The warm and wet conditions largely reflects the persistent large-scale southwesterly flow and frequent tapping of moisture from the tropics and subtropics.

I do, however, wish to point out that this February is fairly close to what the climate models project will be average for our area by the end of the 21st century if we do little to mitigate future greenhouse gas emissions.  Under such high emission scenarios, the average temperature increase produced by the climate models run for the IPCC Fifth Assessment Report over northern Utah is about 10ºF.  During the months of Dec–Feb, the average change in precipitation produced by these models over northern Utah is about 15% (note: some models are wetter, some drier, so don't bank on a wetter future, but let's go with it for this discussion).

These numbers are roughly in line with the departures from average for February, so I look at last month as providing some illustration of the future average snow climate of the Wasatch if greenhouse gas concentrations reach outrageous levels and if the computer models are in the ballpark.  Think of it as a worst-case scenario.  As we saw in February, a greater fraction of precipitation will fall as rain instead of snow in the lower elevations.  In the upper elevations, at least during the heart of winter (Dec-Feb), most of the precipitation in an average year will still fall as snow (this won't be the case in an above average year), but it will be of the high density variety.  Of course, those altitudes will see some rain during warmer storms and warmer periods, which is rare in the current climate.

This sensitivity of snowfall to temperature and elevation in the central Wasatch is illustrated by the figure below, based on Jones (2010), showing an estimate of the amount of precipitation over northern Utah that currently falls as snow that would instead fall as rain for a temperature increase of 1, 2, 3, and 4ºC.  These are for the entire cool season, but illustrate that all elevations are not equal when it comes to the impacts of warming on snowfall.  The low elevations will see much larger declines in the fraction of precipitation that falls as snow than the upper elevations.  The snow in high elevation upper Little Cottonwood is going to become an increasingly precious commodity in the coming decades.  
Source: Steenburgh (2014), adapted from Jones (2010)
Of course, equating this February to the average at the end of the century given a high growth in greenhouse gas emissions is a bit of a worst case scenario, but I feel pretty confident that we will see decreasing reliability of snowfall in the lower elevations over the coming decades.  There will be a lot of weather variability, so these tends don't mean that every year will be bad, but more poor years will eventually emerge.  This will be the case even if winters become a bit wetter.  It is worth noting that declines in the fraction of precipitation that falls as snow have already been documented at lower elevations (below about 8000 feet) in the western U.S., with about half of the decline attributed to anthropogenic global warming (See A Look at Snowpack Trends).  


  1. The skate skiing ruts in the trails at the Olympic Park become deadly when the sun starts to go down and it's below freezing. I know this first-hand. :)

  2. Very interesting analysis. As some people know i have collected a huge amount of snowfall data from ski areas going back to the 1970's and there is zero trend in snowfall despite the acknowledged warming during the 1980's and 1990's.

    This February analysis demonstrates that even with a very large increase in temperature, snowfall is not affected at places with sufficiently high elevation like the Cottonwood ski areas. I believe a similar conclusion could be drawn for most Colorado ski areas plus a few other at high altitude in regional context like Mammoth and Big Sky.

    If 8F degrees above average can't move Utah's snow level above 8,000 feet, it's no surprise that the actual 1.5F or so increase since the 1970's has not changed snowfall to any discernable degree at nearly all western ski areas.

    The real issue here is that computer models predicting the 7-10F increase in temps by 2100 can't explain the essentially flat temperatures of the past 15 years. While it's reasonable to believe the long term temperature trend is up, the degree of increase is likely grossly exaggerated in the computer projections.

    Combine the actual effect of temperature rise upon western ski areas with the questionable credibilty of the computer projections, and one should conclude that predictions of the demise of North American skiing in publications such as the Porter Fox book The Future of Snow are speculative at best. The weather analysis of this February demonstrates that many western North American ski areas are situated in locations that are not particularly sensitive to rising temperatures in terms of snowfall. If the worst case temperature scenarios came to pass, there would be many areas of the world with far more serious problems than out ski areas.

    1. Tony:

      A couple of thoughts with regards to your comments. Much depends on altitude and location, how far in the future you are looking, and how high greenhouse gas concentrations go. Changes in snowfall stem from two factors. One is changes in cool-season precipitation (are winters getting wetter or drier), the other is changes in the fraction of precipitation that falls as snow. Changes in snowpack depend on these two factors, plus changes in the snow energy balance at the ground, which are dependent on temperature changes, cloud cover changes, the impacts of dust on the absorption of solar radiation, etc. The more you bore into the problem, the more multifaceted it becomes, making blanket generalizations difficult.

      So, the way I look at it is that low altitude ski areas with warm climates like Snoqualmie Pass, WA are perhaps the most vulnerable to climate change as they sit right where an increase of a couple of degrees will likely greatly shift the fraction of precipitation that falls as snow (this is already happening at low elevations in the west as discussed at An increase in wintertime precipitation may help offset this for a while, but eventually temperature change will win. At a place like Snoqualmie, it is probably the early season ski season (Nov-Dec) that is most vulnerable. It is already marginal temperature-wise during that period and they will probably find it harder and harder to operate those months over the next couple of decades.

      High elevation resorts in the west aren't as vulnerable to these changes because their climates are colder and thus it takes more warming to have an impact on snowfall and snowpack *during the ski season* (that's an important caveat when you are talking, for example, about high alpine areas).

      However, aren't as vulnerable doesn't mean invulnerable. For mid-mountain Snowbird (9500 ft), here are estimates based on the current climate of how much precipitation that currently falls as snow would instead fall as rain for each degree of warming:

      1ºC: 3%
      2ºC: 7%
      3ºC: 13%
      4ºC: 19%

      The numbers are larger at the base of Snowbird (and larger still at the base of Snowbasin, Park City, Deer Valley, etc.). Further, they are for the entire cool season, which means losses would be smaller during the heart of winter (Dec-Feb) but larger during the shoulder months, so operations for Thanksgiving and December are more vulnerable, but important for revenue.

      One thing I've skipped in this discussion are storm track and precipitation trends. That's a bit more of a wild card. Most computer models call for a drier southwest and a wetter northwest, with Utah and Colorado in the transition zone. If the climate in northern Colorado and Utah were to become wetter, it might offset some of the snow fraction losses, at least during the ski season.


  3. To be fair, I'm actively looking for the most likely places where snowfall amounts might change with the rise in temps since the 1970's. The only clear cut example is the Whistler base. The Whistler Alpine set its snowfall record in 1999, but the Whistler base recorded only 65% as much snowfall as in the prior record year of 1974.

    I think Snoqualmie Pass is the next most vulnerable location (I see we share that opinion). I have compared the snowfall from the late 1960's or early 1970's up to 1992 vs. the 1993 and later seasons in the Pacific Northwest with the following ratios of later period to earlier period snowfall.
    Whistler Alpine 115%
    Mt. Bachelor 111%
    Crater Lake 106%
    Mt. Rainier 106%
    Snoqualmie 102%
    Whistler Base 81%

    So overall PNW ski areas have not yet been adversely affected by the rise in temps since the 1970's. But the high-to-low order of the percentages above corresponds almost exactly to the high-to-low order of altitude within regional context.

    I also have some detail info from the Central Sierra Snow Lab, which is west of Donner Summit at relatively low elevation for the region at 6,883 feet. Once again the average snowfalls before and after 1992 are essentially identical. But the ratio of total water to total snow was was 13.3% before 1987 and 14.3% after 1987.

    So a few signs are there in actual data from ski areas, but they are subtle and you have to look pretty hard to find them.

    The real issue remains that prediction of temperatures rises as much as 4C are not very credible in view of the past 15 years actual data. Even the IPCC had to back off the high end projections in their 2013 report vs. the prior report in 2007. And if the models can't predict global temperatures accurately, why would we think they could predict "a drier southwest and a wetter northwest, with Utah and Colorado in the transition zone?" It is well known that clouds and water vapor are the weak links in the computer models, and precipitation forecasts are highly dependent upon both.