On the face of it this morning's runs look promising enough, with GFS showing the deep low pressure tracking from NW of the UK to Scandinavia not quite managing to merge with further deep low pressure from the other side of the hemisphere, so preventing the development of a more organised polar vortex.

ECM does a better job than GFS, but of course it's output isn't used so we can only imagine what sort of values the ECM run would produce.

---

Given we're under the final ten day's of October and every day this month has been negative, surely the only question now is HOW negative will be the OPI be this October?

Yep - pretty sure I read that if we had neutral OPI values for the rest of the month we would still finish lower than -1.

---

My understanding is that the calculation is based on geopotential height anomalies across the Pole on the line from Canada to Siberia, i.e. horizontally on the projection below:

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=228&mode=12&carte=1

 

This is the 06z GFS for the very last day of October, and weakly positive anomalies are still clearly visible exactly along the Canada/Siberia Polar axis, which is what we need for the OPI to remain negative.

I know this is a hobby horse of mine, but the word "anomalies" is important. Weak low pressure across the Pole would still generate a red-coloured positive anomaly at the end of October, the time when you would normally expect the Polar vortex to be assembling. 

So don't assume that the sight of low pressure over the Pole would necessarily imply problems for maintaining a negative OPI.

---

Something interesting I've noticed regarding the entire list of OPI recorded years, is that the lower the figure is, the more frequent are the cold incursions during the Winter period, as well as a higher risk of a more sustained and intense cold spell during the Winter.  

And it seems to me that a positive OPI doesn't guarantee that there won't be a severe spell the following Winter, as is shown by the positive 1990 figure.

---

You can see how it happened here, Andy:
http://www.meteociel.fr/modeles/archives/archives.php?day=8&month=2&year=1991&hour=0&map=5&mode=0
A positive Arctic Oscillation with low pressure over the Pole, just as the OPI for Oct. 1990 suggested. But the Russian high marched westwards into Scandinavia underneath it.

The odd thing is that the Quasi Biennial Oscillation appeared to be in its positive phase at that time, which just goes to show, it's a funny old game, eh?

---

You can see how it happened here, Andy:
http://www.meteociel.fr/modeles/archives/archives.php?day=8&month=2&year=1991&hour=0&map=5&mode=0
A positive Arctic Oscillation with low pressure over the Pole, just as the OPI for Oct. 1990 suggested. But the Russian high marched westwards into Scandinavia underneath it.

The odd thing is that the Quasi Biennial Oscillation appeared to be in its positive phase at that time, which just goes to show, it's a funny old game, eh?

I'm fairly sure it was mentioned on here last winter that the QBO was in a positive phase throughout that season. Indeed, this was sited by a few people as being one of the reasons why last winter was so poor from the point of view of those seeking cold weather.

---

This looks a strong indicator for the AO. I can't help wondering whether some refinement may improve it further. the chances of the most indicative period being exactly 1 month long and exactly matching with October must be quite slim, perhaps there is some mileage in looking at whether the relationship with AO improves further by including the last week of September, or the First week of November, or excluding the first week of October, etc... if you get my drift.

However seeing as the figure looks like being negative this year, I am hoping it is accurate again this year as I am hoping for a bit of snow!

---

Must admit to being pretty skeptical about this

 

Seems to be lots of excitement about 1 piece of a massively complicated jigsaw.

 

Even accounting for the fact this isn't a 100% correlation for the AO (i.e you may fall in the x% where it's just wrong), picking the AO as a dominant factor in getting cold air to the UK just seems daft*

*Appreciate that this isn't what the research/monitoring of this index is about, just seems to have been adopted by some to that end.

 

As last year showed we still know very little about what makes the weather tick in seasonal terms and if a bunch of extremely clever people with extremely powerful computers can be shown up then I'm not going to be pinning my cold hopes on this particular method of looking at one factor influencing our weather

 

I guess as a method for assessing how one small part of the puzzle may pan out, it's interesting but given how the puzzle works in terms of UK cold (i.e the last piece needed can end up being the one that changes the picture) then it's a very small step on a long, long journey. 

Still cynicism aside I hope it means at least 1 snowfall imby as I want to build a snowman with my daughter

---

Part of the reason for the excitement is the very firm correlation between a strongly negative OPI and cold winter months in the UK when the Quasi-Biennial Oscillation is negative (easterly), as it will be this winter.

The problem, of course, is that by adding the QBO filter to the results you reduce even further the already tiny numbers of years of data on which to base conclusions. But it does at least acknowledge another piece to the jigsaw.

I'm not concerned about people with "extremely powerful computers" having been shown up, even if they are "extremely clever". We and they know that their models go off the rails after a week. They lack both the data and the computer power for the result to be any different. By running them out to 6 months they must surely realize they're just trying to find a way to pass the time. An ensemble of 1,000 runs is 1,000 wrong answers. I can't see how any trend displayed by those model runs is down to anything but chance.

But this index is offering something different imo. It's asking whether a certain set of conditions at a certain period of the autumn sets in train a mechanism by which a specific set-up is highly likely to occur at some point in the subsequent months, regardless of other inputs. And for the UK there's speculation that if one of those other inputs is an easterly QBO, then the two ought, on the basis of limited evidence, to combine very nicely indeed if you're after one very cold month.

---

Few potential long-range predictors sound more promising than one involving a particular sequence of events unfolding through time as a consequence of one other, with a starting point two months in advance of the target season.

For a truly reliable indicator, however, we would need to know of and fully understand every step along the way. At this stage, I'm not sure how far beyond 'B often happens after A when C is present, by some mechanism' with regards to the OPI.

 

Another promising predictor using a sequence of events is that of snow advance across Asia during October coupled with anomalously low ice cover in the bordering areas of the Arctic Ocean during that same period, but it currently lacks in scientific verification.

 

Perhaps one day these predictors, if successful enough, will be used to weight model forecasts in favour of the patterns being signaled. At least then the right target might be aimed for, even if a bulls-eye remains highly improbable, as will always be the case due to the nature of chaos theory.

---

My understanding is that the Snow Advance Index has been subsumed into the October Pattern index. The advance southwards of Eurasian snow in October is reckoned to be determined by the factors which are analysed to calculate the OPI. With this in mind Prof. Cohen, the creator of the SAI, is collaborating with the people behind the OPI on a joint paper to be peer-reviewed.

Regarding the mechanism for A leading to B and C, I suspect that this is an aspect of the study which is subject to peer review and consequently not revealed to us mortals at this stage. But something along the lines of conditions which lead to warm air advecting northwards and waves travelling upwards this month might do the trick. But I'll admit I'm way out of my depth here.

---

OPI forecast to be -2.28 after 06z

http://app.til.it/opi/

---

This mornings OPI forecast stands at -2.57 with less than a week to go...

http://app.til.it/opi/

---

-2.38 at the moment. Definitely going to be negative this year. Should be an interesting test for the index.

---

Seems to have settled at around -2.3. If we keep this to the end of the month will be the second most negative OPI since 1976 with only 2009 more negative!

---

Seems to have settled at around -2.3. If we keep this to the end of the month will be the second most negative OPI since 1976 with only 2009 more negative!

It would be entertaining to see the cold winter signal from this manifesting as a dramatic pattern shift on the scale of December 2009 

The way the pattern appears to be evolving as we head into November, it's not out of the question that we could find ourselves in a similar situation to November 2009, my wettest recorded month with 243 mm. Given that that brought flooding in what was a fairly dry year up to that point, a repeat in this soggy old year would be disastrous. I remember the major player back then was a strong block over western Russia, causing frontal systems to become slow moving over the UK. How about a Scandi block this time around? 

 

In short, things seem encouraging for longer term cold, but concerning for shorter term flooding.

---

Thats an interesting point James.

Doing some crossrefencing from the OPI figures it is interesting to take a relatively snowfree location like mine in South Dorset and see how it fares in winters following the most negative OPI readings. After all we all know the usual suspects can get snow even in generally mild winters.

There is a very good correlation for snow in South Dorset from sub-1 OPI years. Which is encouraging. Sometimes we don't even have to be sub minus 1. The Blizzard of 78 in the southwest came off the back of a -0.95 reading for Oct 77.

Looks like we could be homing in on something below minus 2 which as Brian says will be a good test of the index.

I know it may sound a bit'in my own backyard' but when you own backyard has very little in the way of snow generally then if an index can pick up on years when my own often snowless back yard could get a decent fall or two, then it could be on to something.

Here's hoping

I just really hope I get lucky this time around when I travel to Newcastle in December, I really want to be there when they get days on end of snow - settling snow!!

2009 I went home 4 days before the snows arrived. 

2010 I arrived on the day if the thaw!!

 

 

 

If anything it is drifting slightly more negative, now at -2.45. It looks certain to settle between -2 and -2.5.

---

If this upcoming frigid winter could just hang on until January I think those of us watching the annual CET climb towards an all time record would be grateful. A cold December, while interesting, would ruin the attempt.

---