Thursday, October 13, 2011

What Will Winter Bring?

The answer, of course, is how the heck do we really know? Ah, but that won't keep the weather-obsessed populace that is Minneapolis/St. Paul from wondering. And with that, we were wondering what your thoughts were regarding the upcoming winter. Please take the poll at the right.

What does the NWS say about this winter?
Given the talk of another La Nina forming -- and Accuweather's already infamous prediction of another brutal winter for these parts -- we were curious to know exactly how strong the correlation is between La Nina and a cold, snowy winter for Minnesota. We checked with the Twin Cities branch of the National Weather Service and this is what they told us:

It is true that there are signs of a La Nina returning for this winter out in the equatorial Pacific. About 70-80% of the time there is a La Nina, temperatures in MN for the Dec/Jan/Feb 3-month season are below normal (with about 60-70% of those winters experiencing above-normal snowfall). The Climate Prediction Center shows the historical correlation to the ENSO phase and temp/precip patterns across the US at this site.

Our own guess -- based on gut, not science -- is for average temperatures, slightly below normal snowfall and one to two short tropical getaways.

What are your gut thoughts on the winter ahead. Will it take a while to get going? Will it start with a vengeance? Will it feel like a St. Louis winter?

12 comments:

  1. I think the key is La Nina. There are hints of its return, so I certainly don't see this winter as being warm with minimum snow. I'm going to say typical. Not nearly as snowy as last winter, but we'll see plenty of flakes and feel plenty of cold.

    ReplyDelete
  2. wow we have 21 people who have guessed the snowfall amounts with only one comment??? interesting

    ReplyDelete
  3. My winter forecast...There will be heavy lake effect snow fall across the Great lakes including the Chicago and Detroit areas, extending into the Buffalo area. The area's south of there, say Indianapolis to Cincinnati will see the brunt of the cold. For the NE US I see a chance of a couple of good snowstorms , but it will be mostly mild and the snow will melt quickly, but on the up slope of the eastern Appalachians, the winter will be brutal.

    This leads me to Minnesota. The Trends seem to show that there will be several troughs that will effect us dropping some Alberta
    clippers our way as the season goes on 2-6" with each one. Yet on the other hand there are hint's that we could see strong ridging just to our west. One thing that seems certain, we will not see a Texas Panhandle storm affecting us.

    That leads me to what we saw last year. I think we will see several Colorado surface lows affecting this area just like they did last year, if my guess is right with the cold to the east and the warmth to the west we could be in the battle ground for heavy snow's.

    So hear goes my winter forecast, a heavy snowstorm around Nov 20th, but some of it will melt. Temps around average to slightly above average through the end of Jan. Total seasonal snowfall of 71" Almost the exact opposite of Accu Weather's forecast....wish me luck

    ReplyDelete
  4. Bill thanks for the new post,with the return of La Nina I believe this winter will be remembered more for its cold,rather then its snowfall,now dont get me wrong it will snow and it will be above average(but it wont be like last year when we were challenging the all time record towards the end of the season),my prediction at this stage kind of falls inline with Randyinchamplin's thinking,I believe we will have several bouts of clippers the key is where does the heaviest axis of snowfall line up(typically it runs just south of the metro and places around Mankato get hit pretty well),I'm calling for 77 inches that will be a good workout for my new snowblower.

    ReplyDelete
  5. Well as a statistician, I thought I would try a systematic approach to this: simply the average snowfall of the La Nina winters, plus a random correction taken from the statistical distribution that best approximates winter snowfall in the Twin Cities since records started: result? 62.1" is my prediction.
    I guess we'll see how pure statistical analysis can be useful (or most likely useless) in weather forecasting :)

    ReplyDelete
  6. Love it, Rigil! I wonder if you can somehow do the same thing for temperatures?

    ReplyDelete
  7. Bill,
    temperatures are a bit trickier as the 'official' NWS means are updated every 30 years or so. To keep things simple I used the current averages and looked at the discrepancies observed during La Nina winters + some random corrections. This is what I came up with (numbers are difference from averages):
    November: min +0.9; max 0.4
    December: min -0.5; max -1.1
    January: min -0.8; max -1.0
    February: min -3.8; max -2.9
    March: min -4.4; max -5.3 (!!!)
    April: min -0.4; max -2.2
    So it seems that the temperature effect of La Nina winters would peak in February/ March.
    Once again, we'll see what Mother Nature decides for this upcoming winter.

    ReplyDelete
  8. Great stuff, Rigil! Would you mind if I gathered your two comments and presented them in a fresh post so that more people can see it?

    ReplyDelete
  9. first flakes tuesday? GFS hinting at it.

    ReplyDelete
  10. I would not mind at all, Bill.
    It is your website after all not mine.
    Also for weather geeks out there I wanted to let you know that Thursday, October 27th at the U, there will be this lecture: "Probabilistic Weather Forecasting Using Ensemble Bayesian Model Averaging", which looks at ways to postprocess numerical model ensembles (GFS, etc..) to obtain better calibration. To a bayesian statistician with a passion for weather like mayself you cannot really ask for more... :)

    ReplyDelete
  11. Thanks, Rigil. Do you have any details on location, time, cost, etc., for that lecture? I couldn't find anything online.

    ReplyDelete
  12. @Rigil So if I'm interpreting your temperature assessment properly, you're saying that November is statistically likely to be between .4 and .9 degrees above normal? Is this a certain number of deviations worth? (or a certain percentage of the statistically likely outcomes?) And January would be between .8 and 1 degree below normal? And that April has the widest variation? Feel free to email me offline at mnforecaster@gmail.com.

    ReplyDelete