Negative AO Decembers and what it may imply for winter

[Bob Chill]

A bunch of you probably remember the December +AO post I made last year and the subsequent discussion. For those who don't remember, the short version is a big anomalous +AO in Dec is tough to break down and it screws up the first have of winter with ease.

I'm sure most on here know what the different phases of the AO look like but I figured it would be best to post anom maps so those who only know the AO as a # can see what it looks like in the atmosphere.

Negative AO:

Positive AO:

Pretty easy to see why we like the negative version. In the neg map the highest anoms are over GL. This is most likely because I used just a couple years in the composite and they also had a neg NAO. The AO and NAO also share domain space so they can look similar at times. The easiest way to think about it is when heights around the pole are low all the cold air is locked up with it. When the heights are high it displaces the cold air southward.

Temps maps bring the point home even better.

Neg AO:

Pos AO:

The question that was raised earlier this year is whether an anomalous negative AO in December implies a continuance of the pattern in the following months just like with a positive AO. The short answer is yes with exceptions of course.

Here's a list of the all the -1.0 or lower AO Decembers since 1950:

There is definitely a correlation. The sample size is small but there are still some reasonable conclusions. One thing that is easy to see is a strong -AO in December does in fact imply a pretty good chance of a -AO January. The correlation drops off each subsequent month.

Inside the #'s:

15 of 18 Januaries had a negative AO (83%)

10 of 18 Februaries had a negative AO (56%)

8 of 18 Marches had a negative AO (44%)

Looking at the averages of the set also implies that it's more common for the pattern to relax through the period than to reload. Individual year exceptions of course but overall the -AO pattern behaves very similar to the +AO pattern we looked at last year.

The larger the anomaly the larger the correlations. I pulled just the -AO Decembers with the index coming in @ -1.75 or lower. Here's what the data looks like:

Inside the #'s:

All 10 Januaries featured a negative AO.

8 of 10 Februaries featured a negative AO

6 of 10 Marches featured a negative AO.

This really enforces the importance of the index in Dec and what it implies further down the road. IF (and it's a big IF) we get a December AO reading below -1.5 or lower then it really looks good for us going forward. At least in temperature department.

The monthly Dec AO readings don't give as full of a picture as the dailies. The various years behaved differently. Sometimes the AO is negative pretty much through the entire period like 2009, other times it can flip and stick positive, and then there are times when it's up and down equally.

I graphed the AO dailies for a couple select years to show various behavior.

There are 3 years on the -AO December list that were also in my top 10 snow cover list. 1976, 2002, and 2009 are on both lists. They have also been tossed around a lot as analogs. Not as much now with ENSO crapping out on us but they still apply IMO. I still like 1976 as an analog this year. It wasn't a good snow year but it sure was cold. Sometimes too much of a good thing is a bad thing. Big negative AO/NAO combos are notoriously dry (and cold) in the MA. If I lived in NC I would be rooting for that. I like 76-77 because of the state of ENSO is very similar and the new "snow cover index" is very similar.

Here's 1976-77:

02-03 is a great example of a productive snow year when the indexes were favorable. Not too negative and not very volatile. Just a good steady winter of snow and cold.

So what happens this year? I'm not really sure. Way too many mixed signals with just about everything. Guidance is pointing towards December starting off with a pretty big -AO but it seems the models want to relax it trough the first 10 days of the month. I would have to think the odds at least favor having a -AO December but how strong is pretty much impossible to predict.

[WxUSAF]

Nice work, Bob. I'll take my chances with the -AO and hope the Pac lines up in a good way for us, at least at times. Hard to think we strike out with a AO and NAO that both could average below 0 for the winter. It's not 11-12 anymore.

[usedtobe]

Bob, Thanks. I may use that in my article tomorrow since we're heading towards Dec.

[Bob Chill]

If anyone has the snowfall data for the 18 years on the list it would be a nice addition to the post.

[Deck Pic]

nice work

[usedtobe]

If anyone has the snowfall data for the 18 years on the list it would be a nice addition to the post.

Bob,

I'll do something with that in the next day or two.

Wes

[usedtobe]

Bob,

Here are the DCA snow for the neg AO Dec seasons. The average is higher than normal but that's more because of the really big years than that the AO is the magic bullet. Makes Matt's seasonal outlook look good.

[gymengineer]

Bob,

Here are the DCA snow for the neg AO Dec seasons. The average is higher than normal but that's more because of the really big years than that the AO is the magic bullet. Makes Matt's seasonal outlook look good.

For everyone hoping to avoid a repeat of last season, though, there are also "only" 5 real stinker years in there (<10" at DCA) out of 18. And a couple of even those terrible DCA seasons were not as terrible NW of town ('00/'01 and '96/'97). So, the weenie hope is there for an above average season and the expectation is there for a season significantly better than the last one.

[WinterWxLuvr]

Bob,

Here are the DCA snow for the neg AO Dec seasons. The average is higher than normal but that's more because of the really big years than that the AO is the magic bullet. Makes Matt's seasonal outlook look good.

A median of 14 isn't bad. Probably translates to 25+ in most areas. 5 of 19 over 30 isn't bad either.

[PhineasC]

Bob, you have been putting on a clinic lately. Great work and good news, it would seem.

[Bob Chill]

I went ahead and did the full data pull including ENSO. I sorted it by snowfall. We all know that weak/mod ninos are always our best bet here. 1995-96 was just a Nina anomaly. I wasn't living in MD for that year but from everything I've read it behaved more like a Nino than a Nina. I guess it just goes to show how expected results can be trumped by other factors. That applies a lot with LR forecasting.

One thing of note is that the right combo of indexes trumps enso. Of the 8 years of 15"+, 4 of them featured cool neutral enso. Weak warm under perform on this list but the sample size is really small and it's based on -AO Decembers so I doubt it means much in the grand scheme.

I'll always find 02-03 interesting. It was just a year when everything kept going right. The NAO/AO went negative at the right times but were never anomalous for long periods of time in Jan and Feb. It was just a productive stj and the PNA certainly did its part.

The PNA does seem to be pretty important. It's common sense of course but it's worth noting that when the AO and NAO aren't doing it on their own, you need to get a favorable PNA ridge if you want good snow.

It's also worth noting that the PNA has basically been negative for 2 straight months with just 2 positive blips. It's a lock that the PNA will come in negative in Nov so that will make it 7 straight months where the monthly PNA index has come in negative. It's unusual but not unprecedented to have runs like this. 1989 had an 8 month stretch of negative monthlies. I haven't look very hard yet but scanning through the PNA monthlies show most streaks snap and hold the other way for a bit. Always exceptions but I would hedge that our -PNA streak is going to end probably in Dec and maybe we can get a really nice pattern going despite enso crapping out.

Guidance still pointing at a pretty good -AO for the first half of Dec. Since we can't root for a Nino anymore we gotta go all in for a tanking Dec AO.

[sauss06]

Bob, great write up. Easy to understand too for newbies!

[Ginx snewx]

Fantastic stuff Bob, thanks.

[Wonderdog]

Bob, it's very apparent that you put a lot of effort into this. Thank you.