And we begin

[skierinvermont]

II have a question for you and the forum in general. If warmer air causes more snowfall and more snowfall generally results in colder temp.'s due to a higher albedo, might the Earth have a self-limiting mechanism to prevent runaway global warming at least for awhile?

Food for thought!

In late spring and summer and early fall when albedo matters most, snow cover has exhibited a strong negative trend. Advancing and retreating snow cover strongly enhanced the glacial cycles of the last 3MYA (positive not negative feedback).

[skierinvermont]

There has been a pretty persistent bubble of warm air at 70mb over central Asia, probably related to the positive snow cover anomalies. The last couple days it has been shrinking, and below normal air temperatures are beginning to dominate the northern stratosphere. Unless we see this trend reverse soon, I doubt that the AO will be strongly negative as it was in 09-10. However, I also doubt a strongly positive AO. There should be chances for the AO to go negative.

[Eastpa]

II have a question for you and the forum in general. If warmer air causes more snowfall and more snowfall generally results in colder temp.'s due to a higher albedo, might the Earth have a self-limiting mechanism to prevent runaway global warming at least for awhile?

Food for thought!

If its legitimate climate change, Mother Earth has a way shutting that whole thing down.

Sorry I couldn't resist, but the analogy applies.

[Jonger]

Uh, ok. You do know warmer air causes more snowfall and thin ice? Some "fightin" words.

Fightin or fighten.... I was intentionally spelling it wrong, so I should get a pass on the version I choose.

Jon

[das]

We'll test that theory this winter. I think we'll see a very negative AO. Solar max/min is overhyped.

Care to elaborate on why you think we'll see a "very negative" AO?

[CoastalWx]

I think there certainly is something to the solar max/min, but looking at the Ap and sunspot numbers are probably more telling as to what is bombarding the stratosphere. We've quieted down a little from this fall.

[Mallow]

Care to elaborate on why you think we'll see a "very negative" AO?

Because the only two winters that had a stronger SAI signal were 1976-1977 and 2009-2010. Two of the most -AO winters on record.

http://www.agu.org/pubs/crossref/2011/2011GL049626.shtml

[usedtobe]

Because the only two winters that had a stronger SAI signal were 1976-1977 and 2009-2010. Two of the most -AO winters on record.

http://www.agu.org/p...1GL049626.shtml

Mallow, if I remember correctly, the really high correlation period was only 10 year hardly enough to know how much stock to put in it. The longer period correlation were more in the .63 or .65 range. Not bad but when you scare it then it explains much less than half the variance. As a forecaster that's still not as high as you'd like to have much confidence. I lean towards the AO averaging in the negative territory but with very little confidence.

[Mallow]

Mallow, if I remember correctly, the really high correlation period was only 10 year hardly enough to know how much stock to put in it. The longer period correlation were more in the .63 or .65 range. Not bad but when you scare it then it explains much less than half the variance. As a forecaster that's still not as high as you'd like to have much confidence. I lean towards the AO averaging in the negative territory but with very little confidence.

True, but .63 to .65 for a 3-month lead time is pretty damn good. And most of the uncertainty comes in years where the SAI was somewhere in the middle. The more extreme SAI values are better correlated.

Also, the lower correlation is from the weekly data (before daily data was available). The higher correlation might actually be due to the fact that we're using daily data now, rather than just being a by-chance short-time-period thing.

[das]

Because the only two winters that had a stronger SAI signal were 1976-1977 and 2009-2010. Two of the most -AO winters on record.

http://www.agu.org/p...1GL049626.shtml

Thanks. I've meant to read Cohen's latest paper and this give me the impetus to do so. If anyone is interested, here's a link to the whole paper:

http://web.mit.edu/jlcohen/www/papers/CohenandJones_GRL11.pdf

[skierinvermont]

The whole mechanism for SAI ---> -AO is warming and wave propagation into the stratosphere. I'm not sure that we're really seeing that.

It's just like last year when everybody was using the strong correlation between the summer AO and winter AO to predict a winter -AO. But the usual mechanism wasn't there last year. Summer -AOs almost always have warm stratospheres persisting through late fall and into winter. But last fall was a major exception, which enabled those looking at the actual causation instead of just the correlation to correctly predict a +AO winter.

Look at the stratosphere today at 70mb. Not looking good to me.

[Bob Chill]

Because the only two winters that had a stronger SAI signal were 1976-1977 and 2009-2010. Two of the most -AO winters on record.

http://www.agu.org/p...1GL049626.shtml

I posted a breakdown of various data irt the state of oct snow cover and following winters in the MA sub-forum. Like Wes already said, pulling the top and bottom 10 out of a 40 year sample doesn't nearly cover a long enough period to draw conclusions but the does seem (at the very least) a bias towards a negative ao/nao combo when the NH snowcover is anomalous in Oct.

I sorted the 40 years using eurasian totals but it didn't really make that much of a difference vs the entire NH snow cover anomaly.

Here's the AO/NAO grid for the top 10 and bottom 10 snowcover years:

The averages definitely show a decent spread irt the state of the AO the following winter. 06-07 definitely shows that it can go the other way. I hope this year isn't the same because enso is eerily similar.

Full post is here:

http://www.americanwx.com/bb/index.php/topic/37592-breaking-down-october-nh-snowcover-data/

[Mallow]

The whole mechanism for SAI ---> -AO is warming and wave propagation into the stratosphere. I'm not sure that we're really seeing that.

That's the hypothesized mechanism, though it's certainly not known to be the case. Also, I don't know that you'd be seeing the signal there already?

I'm believing the correlation without actually suggesting I understand the cause.

[skierinvermont]

Can anybody point me to a source for the SAI where you can actually see the numbers in table format?

[skierinvermont]

Just looking at Oct snow, not SAI, 1971 is a good example of what I'm talking about. Lots of snow, but stratosphere remained cold, AO was positive. Somebody look at snow only would have predict the AO wrong, the person who was looking at the stratosphere would have been right.

Only one example of course, but will look further if I can find the SAI.

[okie333]

Mallow, if I remember correctly, the really high correlation period was only 10 year hardly enough to know how much stock to put in it. The longer period correlation were more in the .63 or .65 range. Not bad but when you scare it then it explains much less than half the variance. As a forecaster that's still not as high as you'd like to have much confidence. I lean towards the AO averaging in the negative territory but with very little confidence.

Read the paper again... it is the fact that daily values instead of weekly values are used that makes the correlation better... the correlation of the last 10-15 years is actually worse than the rest of the period when weekly values are used for all of them.

[usedtobe]

Read the paper again... it is the fact that daily values instead of weekly values are used that makes the correlation better... the correlation of the last 10-15 years is actually worse than the rest of the period when weekly values are used for all of them.

I just relooked and the 1998-2011 score is .85 correlation versus the 1975 to 2011 period which is .63 so I was wrong about the length of the shorter period but not about the longer time period having a lower correlation. In fact in the paper they state.

"The fact that we discovered a single predictive

index that explains close to 75% of the variance of

the winter AO (though the period is short and the degraded

SAI over a longer time period explains less of the AO

variance) is inconsistent with this thinking and demonstrates

that the AO, while thought to be unpredictable, may in fact

be one of the most easily predicted phenomenon known in

the climate system."

In fact the 20 odd year correlation explains around 72% of the variance. the longer period the with a correlation of .63 (actually a little below that) explains about 40% of the variance. That's still lower than you would like to use as a forecast tool. A longer time period should give you a better idea of the actual correlation and variance than the shorter time period unless you think the regime in different starting in 1998. Now it might be that they think the data is not as good before 1998. I think we'll find out how well their method works in the coming years.

[tacoman25]

Just looking at Oct snow, not SAI, 1971 is a good example of what I'm talking about. Lots of snow, but stratosphere remained cold, AO was positive. Somebody look at snow only would have predict the AO wrong, the person who was looking at the stratosphere would have been right.

Only one example of course, but will look further if I can find the SAI.

What is the correlation between the fall stratosphere temps and the following winter? As in, how strong is it?

I seem to remember some concern about a cold stratosphere in the fall of 2010, and that winter had some really strong blocking.

[skierinvermont]

What is the correlation between the fall stratosphere temps and the following winter? As in, how strong is it?

I seem to remember some concern about a cold stratosphere in the fall of 2010, and that winter had some really strong blocking.

I never ran a correlation last year but here are the statistics I came up with. A very strong correlation. I'd guess an r of .8 or .9. The most impressive stat if you combine it all is that the AO only had the opposite sign of what was indicated by this method in 1 out of 40 months (excluding near-neutral months). That means it was wrong only 2.5% of the time.

Here's my post from last year:

..

I found another stratospheric temperature data set that goes back to 1950. This time I focused on the area 50-90N 90E-150W (basically Russia and surrounding areas) where the correlation is strongest.

I used 70mb stratospheric temps from the NCEP reanalysis: http://www.esrl.noaa.../timeseries1.pl

Based on this I would give us at least a 75% chance of having a +AO in December+January. If the stratosphere remains near record cold for the next 2-3 weeks, our chances diminish even further. There has never been a significantly -AO January following a top 10 cold stratosphere.

November stratosphere

-The December AO was 1.09 higher following the 10 coldest than the 10 warmest.

-The January AO was 1.75 higher following the 10 coldest than the 10 warmest.

-The December AO was .63 higher following cooler than average than warmer than average

-The January AO was 1.71 higher following cooler than average than warmer than average

November + December stratosphere

-The December AO was 1.82 higher following the 10 coldest than the 10 warmest

-The January AO was 2.81 higher following the 10 coldest than the 10 warmest

-The December AO was .72 higher following cooler than average than warmer than average

-The January AO was 1.41 higher following cooler than average than warmer than average

- 8/10 coldest had +AO Dec, only 1 below -.1

- 8/10 coldest had +AO Jan, none below -.2

-9/10 warmest had -AO Dec, none above +.2

-9/10 warmest had -AO Jan, none above +.3

[uncle W]

having a negative AO doesn't guarantee a snowy winter...The storm track so far is encouraging...If it was winter blizzard warnings would be issued for many areas with the next storm...

[Mallow]

I never ran a correlation last year but here are the statistics I came up with. A very strong correlation. I'd guess an r of .8 or .9. The most impressive stat if you combine it all is that the AO only had the opposite sign of what was indicated by this method in 1 out of 40 months (excluding near-neutral months). That means it was wrong only 2.5% of the time.

That is interesting.

But you're basing this on November's overall mean stratospheric temperature anomalies, right? Isn't it possible we start out a little cool but then warm up significantly as the month goes on?

EDIT: For example, the first two days of the month in 1976 showed a mostly cold stratosphere, as well.

[skierinvermont]

That is interesting.

But you're basing this on November's overall mean stratospheric temperature anomalies, right? Isn't it possible we start out a little cool but then warm up significantly as the month goes on?

EDIT: For example, the first two days of the month in 1976 showed a mostly cold stratosphere, as well.

Yes there are one or two examples I've found where November Asia 70mb started cold and averaged cold but ended the month warm, and the AO that DJ was negative. The strongest lead time is probably at 3-4 weeks on a weekly basis. Anything longer than that and the correlation starts to weaken but still remains quite strong until a 2-2.5 month lead time.

Also the 1 in 40 stat I gave you was the mean Nov+Dec 70mb predicting the Dec and Jan AO. If you only use November it gets it wrong something like ~3 times, its neutral like ~5 times, and gets it right the remaining ~32/40. And just looking at the data casually, late November is a bit better than early November. So there is still time for the stratosphere to warm before the end of the month.

Even October asian 70mb temps have a decent correlation to the winter AO.. the stratosphere tends to stay in a similar state unless something interrupts it. Heck, even August does (which is why a -AO summer often is followed by a -AO winter).

[Mallow]

In other words, I still trust that the SAI is going to come through for us, and that it portends a good -AO winter.

[das]

Here is the anigif for the entirety of October. Gains have continued into the first week of November but the correlation shown in research is only for October so I did not show beyond the end of the month.

[skierinvermont]

Stratospheric temps over Asia continuing to drop.. starting to look like a classic pre- +AO stratosphere.

[uncle W]

Stratospheric temps over Asia continuing to drop.. starting to look like a classic pre- +AO stratosphere.

the AO is way positive today...up sharply from last month...

[WXeastern]

Stratospheric temps over Asia continuing to drop.. starting to look like a classic pre- +AO stratosphere.

What does current Asian stratospheric temps have to do with winter Arctic Oscillation values?

[skierinvermont]

What does current Asian stratospheric temps have to do with winter Arctic Oscillation values?

See stats on previous page. This chart sums it up pretty well too:

[WXeastern]

Since the topic of this thread is directly related to October AO values..here is the following met winter average anomaly temps from all October AO values since 1950 that averaged out to at least -0.5 ...The mean December AO value from these is -0.343

The October input years are 1960 1966 1968 1974 1976 1979 1980 1981 1993 1997 2002 2003 2004 2006 2009

Average anomaly winter temps with a previous October AO value of at least -1.00

And since this October went -1.51 here is the average anomaly winter temps from the only two years that had October AO values of at least -1.40

I find it very hard to argue against a cooler than normal winter in the eastern US!

[skierinvermont]

That's not really that strong of a correlation. Several of the years you mentioned were not cold, it just averages slightly cold in the means (-1F). 06-07 was notoriously warm. The correlation I presented on the previous page is much larger.

Also the mechanism of transmission from October to winter is likely the stratosphere. This is evidenced by the fact that the years you mentioned have a warm stratosphere in November on average. However, this years stratosphere is dissimilar to the years you have mentioned. It is cold.

This is the mean stratospheric temperatures from the years you mentioned. It is the opposite of what we are witnessing this year which will likely be an exception to your list, such as 2006-2007 was. The whole mechanism of transmission from October AO values, and October snow cover, is the stratosphere. When that mechanism of transmission fails, as it is this year, the correlation breaks down, as it did in 2006-2007.