A northern thermometer

Posted: December 26, 2011 by tchannon in climate

Figure 1

A contentious dataset is published by the Rutgers University Global Snow Lab, with some people arguing one way, and some the opposite way on what is happening. This tends to be based on selective data.

The best data they publish is weekly so lets use that.

Figure 1 shows the lot. I’ve also put a filtered line in there.


Figure 2

Figure 2 is clearer. I think it says not very much is going on but there might be change over the last few years.


Figure 3

Figure 3 shows Jan 2000 onwards.

This suggests more snow loss in summer but more snow cover in winter, hence two groups arguing in opposite directions.

I don’t know what is going on.

Snow lab data page is here

Data plotted is here (XLS)

  1. amcoz says:

    It’s pretty obvious to me: the iceman is busier in summer stockpiling for the winter.

  2. Brian H says:

    It’s a battle between the Snow Queen and the Sun King. The Sun King is on track to weaken, though, and the Snow Queen will Rule Us All!

  3. tallbloke says:

    From eyeball and memory there doesn’t seem to be any obvious correlation with ENSO. I wonder if it might be worth comparing the data with the Arctic Oscillation.

    One thing to notice is that overall, not much has changed. What is the average total snow cover Tim?

  4. Tenuc says:

    Thanks for the post Tim – I too am not sure what this means for recent years, but looking across the whole span of the graph it seems that more winter snow cover doesn’t lead to more summer snow cover, or vice versa.

    Seems to be some sort of balancing mechanism at work here, so small variations seen could be due to vagaries of change in the amount or water vapour and temperate zone temperature?

  5. George says:

    One thing I have noticed in the Continental US (according to NCDC here: http://www.ncdc.noaa.gov/oa/climate/research/cag3/na.html ) is that summer temperatures from 2000-2011 are trending up at 0.37 degF / Decade while winter temperatures are trending down at a whopping -2.90 degF / Decade.

    So summers are flat to slightly warmer while winters are much colder.

  6. George says:

    And if you go back to 1998, the winter trend is -3.27 degF / Decade. That’s just amazing. Last winter we had at least one day with snow all the way to the Gulf of Mexico. This year we have fairly little snow cover in the US so far but what snow we do have is oddly in the South. New Mexico and Texas have snow while Nebraska, the Dakotas, and Iowa are without snow cover.


  7. […] A northern thermometer Like this:LikeBe the first to like this post. […]

  8. Jim says:

    Who would have gussed, there is more snow cover in the winter?

  9. @George: That sort of trend is typical at high latitudes, even in the UK (CET – http://www.trevoole.co.uk/Questioning_Climate/_sgg/m2m1_1.htm) although less exaggerated. The rate of warming in summer is less than rate of warming in colder months. Possibly it is due to negative feedback which limits the upper temperatures?

  10. tchannon says:

    Jim has pointed out there is more snow cover in winter, wicked in a way since this involves one of the problems.

    What I’ve done with the data is very simple, the result is showing a change in area, fine, except this is not proportional. I’ve looked at various schemes to see if there is a better representation, nothing I found.

    The geometry of the polar snow area is complex, almost landlocked ocean with sea ice and the land with some mountainous areas.

    More snow in winter implies a change in variation through a year. I think a candidate cause might be change in precipitation.

    As ever much more data is needed.

  11. tchannon says:

    Tenuc, I’ve uploaded just the plotted data, see link at end of article.

    Mean area is about 27m sq km

  12. Tim, have you noticed the data drop-outs in the first few years? They are from summer periods and if not taken into account will tend to bias trends towards declining snow levels.

  13. tchannon says:

    Lot of kids drop out in early years, yes. 😉

    Will have no effect on what I have done, simply remove the common annual and the unimportant filtering is signal processing stuff where it will interpolate over, is inherent.

    OTOH if someone tries a linear trend it would need care. I suggest there is no rational evidence of a trend.

    Given the change in space technology over time the measurement (I assume is mostly eyeball) is likely to be questionable anyway.

  14. J Martin says:

    Isn’t this about tipping points.

    If one winter there is 10 cm of snow and it melts in the spring and the next winter there is 20 cm of snow and again it melts, there will be no trend. So as long as it doesn’t snow enough to overcome the spring thaw, there will be no trend.

    When a combination of sufficient increase in snow and a cooler spring summer combine to no longer thaw each winters excess snow, then a trend will start to show and I will want to move to somewhere warmer.

  15. tchannon says:

    There has been comment and email about linear trend.

    I deliberately kept this blog post non-contentious and without pushing strange stuff. This doesn’t mean I have not looked in detail at various times.

    The early missing data will mess up computing a linear trend using simple techniques.

    I have used the analysis software here to try and quantify a trend, essentially force it to look when it would otherwise ignore this as too minor. This could produce a straight line but gives a very long period curve. if I infill the missing data using modelling and produce a simple linear trend the result is very close to the “trend” of the found curve. The critical point is that linear trends on non-integer number of waves gives a fake slope.

    I am not happy about showing anything. In words, the wave has a minima at about year 2000. This is consistent with the stall in warming but should not be taken as provable.

    There are other possible items in there, most I am not mentioning without spending many hours trying to validate.

    I do mention the annual cycle seems to be more than doublets, suggesting modulation, no surprise.

    Spectra showing what is left after synchronous removal of the dominant 1 year, there is more in there.

  16. Joe's World says:


    Ever consider gases differentiate in temperatures under pressure?
    Our atmosphere is more like a focused lens?
    If there was no land cover, the temperatures would be colder.
    But the land absorbs more heat and stores it short term before releasing it.
    Winter time of solar heat is vastly less in time due to the angle changes of the tilting.
    While summer is vastly more time with less dense air.

  17. tchannon says:

    Joe’s W,
    No I haven’t, at least not in the context of polar conditions, wrong thread?

    I suspect there is sufficient noise/stirring to mix fairly fully, however, I have postulated the annual cycle for Mauna Loa CO2 comes from Arctic ice. Don’t think I’ve ever shown this stuff,
    Walk from Alaska to MLO and you would arrive with the gas.

  18. Tim, I think the CO2 and sea ice work is worth following through if you get a chance.