North American snow models miss the mark – observed trend opposite of the predictions

While some other bloggers and journalists insist that recent winter snows are proof of global warming effects, they miss the fact that models have been predicting less snow in the norther hemisphere. See this  2005 peer reviewed paper:

Frei, A. and G. Gong, 2005. Decadal to Century Scale Trends in North American Snow Extent in Coupled Atmosphere-Ocean General Circulation Models. Geophysical Research Letters, 32:L18502, doi: 10.1029/2005GL023394.

It says exactly the opposite of what some are saying now. – Anthony


Guest post by Steven Goddard

A 2005 Columbia University study titled “WILL CLIMATE CHANGE AFFECT SNOW COVER OVER NORTH AMERICA?” ran nine climate models used by the IPCC, and all nine predicted that North American winter snow cover would decline significantly, starting in about 1990.

In this study, current and future decadal trends in winter North American SCE (NA-SCE) are investigated, using nine general circulation models (GCMs) of the global atmosphere-ocean system participating in the upcoming Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC-AR4)…

all nine models exhibit a clear and statistically significant decreasing trend in 21st century NA-SCE

Some of the models predicted a significant decline in winter snow cover between 1990 and 2010.

Climate Model predictions of Snow Cover Decline

As we know, winter snow cover has actually increased about 5% since it bottomed in 1989, and is now close to a record maximum.

Below is another interesting graph.  It shows the number of top 100 snow extent weeks by decade.  I took the top 100 weekly snow extents (out of 2227) from the Rutgers record and sorted them by decade.  The past decade has been at least as snowy as the 1970s.

The past decade has had the most weeks in the top 100 since 1966.

NASA Earth Observatory Images

Above are images from NASA showing snow extent from 2001 to 2004.  Below is an image from 2010, showing snow cover in all 48 states.

NOAA Image  – February 12, 2010


UPDATE: Here is a new graph of north American winter trend produced by Steve at the request of commenters:

So far, the climate models have the wrong polarity on their predictions of winter snow cover changes.


newest oldest most voted
Notify of
James Crisp

Is that snow in Mexico? I’m no expert but I’m guessing that is quite rare?

It’s good to see Steven posting again. I missed his posts while he was gone.

adrian smits

There is the most powerful repudiation of agw. You can actually see how cold it is getting!

Lazarus Long

So has anyone ever taken historic climate data, plugged it into these “models”, and compared them to, you know, reality?

DeWitt Payne

Once again it is shown that climate models have no skill at regional forecasting. This sort of thing isn’t science, it’s pure speculation. Reputable journals should stop publishing these papers.

Henry chance

I have several conclusions when models are wrong. They in this case seem to leave out variables and make certain assumptions. It is a shame they are using models and ignoring actual temperatures and precipitation.
Test the models and find they are wrong, then we don’t accept their conclusions.


Will they revisit the data and post the new findings? Unlikely. Has anybody asked the authors, “what say you?”

hippie longstocking

I am shocked and bewildered! Computer models created to generate a specific output using questionable logic and programming didn’t match reality? I am quite surprised. It truly is worse than we thought…


Fortunately the ink was dry and the predictions were recorded. The words that the snow was exactly as ‘we’ said turn out to be the current convenient advantage of hindsight mixed with the hope of short memory of the public.
Thank you Steven Goddard for the clear as always post.


A proof that they change what they say along the way.
It should be more, it shold be less.

P Gosselin

Oh another hockey stick that proves AGW!
Don’t you know? More snow means it’s getting warmer. Yes snow way down south in Florida and Houston means – WARMER!

Charles. U. Farley

Just another case of being wrong on so many of their “predictions”.
Maybe theyd benefit from a gypsy costume and a crystal ball to aid them?
Cross my palm with carbon….

Phillip Bratby

It’s what you expect when your models have an in-built bias and no validation.

P Gosselin

All that extra mointure going up there in the air like that, and spillin into Florida, Houston, Rome and hell even Baghdad. Damn it must be getting really hot out there! So Put on your aluminium foil suits
…or frrrrrrryyyyyyyy in the hot sizzling snnnnnnoooooowwwww…


O/T but I have noticed in the IJIS site with sea ice extent, that there has been a significant jump upward over the past couple weeks. Shockingly, there has been no press announcements about it by the warmists, even though, when it was below 2007 levels in January, they were all over it.

Jim F

How to back up your argument.
This site is fun:
“…A complete list of things caused by global warming….”


Hmm… I find it interesting… those pushing the AGW agenda claim that the extra snow is an expected outcome supporting AGW due to increased water vapour in the air. However, this study clearly shows that the AGW predictions for the US was clearly LESS snow. So far, then, reality is falsifying the predictions, meaning the models have clear problems. If an AGW supporter would still like to claim that the extra winter snow is an expected outcome, can they please point to scientific studies showing that was clearly the expectation for the US? You cannot predict one thing based on your AGW science, then when the outcome is the opposite, then claim that *that* scenario is also consistent. Such makes AGW unfalsifiable and hence unscientific.

P Gosselin

Seriously, like in my last posts, it looks like Arctic sea ice is going to hit the 14M mark. I didn’t expect it this year.
Oh! but it’s paper thin – don’t you know! The last polar bear is likely gonna drown this summer!
Sorry, but I just watched 3 min of that nutjob Rahmstorf, and this is what happens when you do. The guy just has a way of spreading his madness.

Bob Tatz

Do you know if there are any plots from the data on:
that show the daily “Area Covered By Snow” data by year?
The coverage for US seems to be running 10-15% higher consistently this year (which agrees with your post). The data is accessible… I just don’t want to reinvent the wheel if a plot is available.

John R. Judge

When the pundits and politicians say that the current snow storms are not “proof”, they are correct. When they say that they are not “evidence”, as many of them do, they are blowing smoke. Rising graphs may have ups and downs. They don’t have 100+ year records.


Re: the last map/image…
Snowcover on all 48 states, but also; All of Canada, except one tiny green spot at Vancouver. (Which is, of course, normal for them).
Silly, I know, but I can’t resist poking fun at the IOC and their ilk.


Yeah, but the IPCC has always been a snow job.


You`re using ths data upside down.

Thanks, Steven Goddard, for keeping the spotlight of public attention focused on errors in the computer models that were predicting less snow.
Decades of filth and abuse of science by the National Academy of Sciences and all the research agencies whose budgets require NAS review (NASA, DOE, NOAA, etc) are concealed beneath the climategate iceberg, so you can expect negative comments from those with a vested interest in the flow of federal reserch funds.
Hang in there Steve!
Oliver K. Manuel
Emeritus Professor of
Nuclear & Space Sciences
Former NASA PI for Apollo

William Sears

It was Langmuir who said that it is a characteristic of pathological science to give ad hoc answers, invented on the spur of the moment, to any and all criticisms. See Number Watch for a listing of all Langmuir’s conclusions.


surprise, surprise


If the alarmists can come up with a theory, one that makes sense, that says global warming SHOULD result in more snow cover, that’s fine. But here’s my problem. If you can come up with a theory that says global warming should result in MORE snow cover (as they now have) AND a theory that says that global warming should result in LESS snow cover (which they have in the past), then whatever happens to snow cover, it can be blamed on global warming. A similar paradox exists for hurricanes (one theory says that global warming should result in more and more severe hurricanes, and another says global warming (at least that of the anthropogenic persuasion) should result in less and less severe hurricanes) and Antarctic ice cover. And I’m sure, for other climate variables, if the facts contradicted what the AGW theory says, they could just come up with another theory to explain why the data says something different from what the previous theory said, but still be consistent with the notion of AGW.
The problem here is, no matter what happens with snow, ice cover, glaciers, rainfall, hurricanes, air temperatures, ocean currents, ocean temperatures, sea levels, and anything else that has anything to do with climate, SOMEBODY is going to come up with a theory, one that makes sense on some level, to explain how it is caused by anthropogenic global warming. And so, no matter what happens, AGW can NEVER be disproven, unless, maybe, everything stayed exactly the same for about 30 years straight. Of course, that will never happen, but even if it did, I’m sure some alarmist would come up with a theory about how a complete lack of climate change was evidence of climate change.
So, I guess what I’m saying is, I challenge the global warming alarmists to specify any and all circumstances, with respect to levels and/or changes in any set of climate variables, over any time span, that they would accept as evidence that either a) global warming is not occurring, or b) global warming is not caused by mankind. If they cannot specify any such set of circumstances, then we know that it is impossible to ever change their minds, and we can restrict this discussion to one of REAL science.


In 2002 a MET office spokeman predicted UK children would in future see snow as a once a lifetime event.
Hence 2009 and 2010 UK major snowfalls have had the right effect on AGW beliefs 😉 .
And the next lot is due to fall Sunday.

Mr B

The 1970 data is interesting in comparison to 2000; considering that 1970 was the great global cooling scare. I wonder if the sky was falling, media wise, with each snow storm back then?

I believed you the first time, Steve!


Some of IA is scheduled to get up to 4″ of additional warming today and over the weekend. This season’s been fun, setting the historic snow coverage record in IA and all, but I think we’re all getting a little tired of it.


I wish readers here would forward stories like this to PBS and ask WHY the NEWSHOUR is so one-sided on AGW.
The PBS Ombudsman is here:
In the run-up to Copenhagen, the PBS NewsHour stated “”this huge team of scientists from all over the globe issued these unanimous warnings about the really extreme danger to the planet.”


And I look out of my office window in rural Hampshire and see snow and hail failing from the sky. So glad it is a thing of the past.

Physics Major

So anyone who claims that increased snow extent is consistent with CO2 induced global warming is what? A liar maybe?


Modelling is pseudo-scientists child games. Not matured men and women replace thinking and empirical testing by playing, THIS IS PLAYING, it is like day-dreaming, it´s onanistic. It reveals an astounding decline in the american culture. This is your private and yours only apocalypse, your long awaited armageddon. But you can reborn out from the ashes like a phoenix, you can remake yourselves by radically pulling out all those evil weeds growing everywhere in your society.

Michael Jankowski

Gavin readily admits “regional” forecasting has little skill…but it is supposed to have skill at continental scales and higher. This is “continental.” 😉

Excellent stuff. Warmists sometimes try to say that “models actually expected Antarctic ice to grow”, while looking into IPCC, exact opposite is found.

John F. Hultquist

The underlying assumption of the models would seem to be an inverse relationship between ppm CO2 and snow cover. Your first chart’s orange line, if inverted, would seem to track the expected rise in concentration of this naughty gas in the atmosphere. Because I think this assumption is wrong I am not surprised that the models are out of whack.
Areas that are snow covered or very cold are not experiencing plant growth. No plant growth means a lack of photosynthesis and the processing of carbon dioxide. Earlier springs should lead to reduction of atmospheric carbon dioxide and later springs should lead to increases from home heating and industrial processes. It seems to me the direction of causation is more direct in this sense than in the struggle to show CO2 impacts the onset of spring.

This meme that our models predicted more snowfall is such BS spin. I also looked at a few articles to see published predictions. In a Nature 2005 review paper in the abstract they state
“In a warmer world, less winter precipitation falls as snow and the melting of winter snow occurs earlier in spring”
The bulk of the paper is creating alarm about how the lack of snowfall reduces water supply. These guys are like day-traders just riding whatever trend is there, and then blaming it on CO2. THey have a model to cover all things. Its Snake oil. Pathetic science.

Chad Woodburn

If increases in snow can serve as evidence and proof for AGW, then so can increases in the extent of glaciers (since the snow is the major source of glacial volume). That way, whether the glaciers grow or shrink, AGW alarmists can use them as proof of global warming.
This is why we must get the alarmists to explicitly state the conditions under which their theory can be falsified. And if they were doing real science, they would volunteer and include those conditions in their own studies. I say “real science” because real science involves testing hypotheses, and the tests by definition must differentiate the significance of the potential outcomes.

Frank K.

Here’s a link to the original paper:
The money quote (please see Fig. 1 in the paper):
“Mean simulated NA-SCE values range between
0.55 and 0.95, compared to observed values of ~0.75
(for perspective, Greenland represents ~0.10 of North
America) (Figure 1). Only two models have mean values
within ±0.05 of observed values; only two models overestimate
NA-SCE; and all other models underestimate NASCE.
Neither model that overestimates NA-SCE accounts
for fractional snow coverage in grid boxes. There is no
apparent relationship between model spatial resolution and
mean NA-SCE (not shown).”
“Figure 1 indicates that models disagree with each
other, and with observations, on the timing of decadal scale
NA-SCE variations. None of the model simulations capture
the observed temporal pattern or anything similar
. While
some models at least approximately capture the magnitude
of decadal scale variations in NA-SCE, others show little or
no decadal scale variability.”

If we get anymore global warming we’ll end up with a snowball Earth. Save the whales!

B. Smith

Must be all the dirty coal particulates and pollutants reflecting energy back into space!
Thank the Almighty for all those sooty coal-burning Indian and Chinese industries!
Ever since I saw the NASA sat pics clearly showing heavy banding of CO2 (especially in the northern hemisphere) around the globe, as opposed to a blanket layering as I understood was being claimed, I’ve wondered of the implications. I am not an atmospheric scientist so forgive me if the following questions seem too basic.
1) Do the various climate models use an evenly-distributed blanket of CO2 for their calculations, or do at least some of them use the uneven concentrations clearly shown by the sat photos?
2) What are the PPM’s of the areas of concentration as opposed to areas that are clearly devoid of excess CO2? It defies logic to accept that the concentrated bands of CO2 would have the same parts per million of CO2 as areas that show no concentration of CO2.
3) Intuitively, these CO2 concentrations could have some effect on weather patterns in those areas of concentration. These bands of concentrated CO2 should also positively affect the land-based flora of the massed regions as it’s “settled science” that plants love CO2. Am I off on my thinking here?
Any help from the cognoscenti is truly appreciated!


James Crisp (07:36:51) :
I would be rare in the lowlands, but there are mountains down there, even some glaciers. I’m sure the glaciers will be gone bt 2035.

Jeff in Virginia Beach

Correct me if I’m wrong, but as I remember it they were not only predicting less snow due to AGW, they were counting on that in the models as one of the drivers for the high ends of their temperature predictions for the year 2100.
Less snow = lower albedo = faster warming making even less snow and melting poles, which makes for even faster warming- a vicious circle or, if you prefer, a snowball effect.
So, at what point does the failed prediction of diminishing snow coverage alter the predicted temp range for the year 2100?


This is EXACTLY the kind of argument that is needed and that is why all the old graphs etc must be saved so there is no doubt about it. This posting would be a tremendous embarrassment to those who said just last week that this recent USA and Europe and Asia etc.. snow was proof of global warming. They said EXACTLY THE OPPOSITE.

John Diffenthal

The reality is that if anyone normal had build a model or series of models for a client organisation which was so dismal in its forecast accuracy, the client (and with good reason) would probably ask for its money back.
Maybe the way to attack this is through a taxpayers alliance!


But Global Warming predicts everything. How dare you question God sir, err I mean the holy Global Warming theory!


James Crisp (07:36:51) :
Is that snow in Mexico? I’m no expert but I’m guessing that is quite rare?

In the mountains, that’s normal. There’s even one small ski resort in Mexico – Monterreal.
Further south, in the mountains of Equador, one (or rather, two) can even have a snowball fight over the equator.
“The snows of Kilimanjaro” also are located just some 330 km from the equator.

Steve Goddard

The graph in the article shows winter snow extent for the entire Northern Hemisphere. Here is one showing North America only, same upwards trend – contrary to the climate models.