OCTOBER PATTERN INDEX (OPI) MONITORING WINTER SEASON 2014-2015

[Daniel*]

This is oh so exciting - my initial thoughts for winter 2014/2015 is for severe deep cold January into March period. It will be a slow burner, gaining in momentum as we go through the festive season.

OPI at a very healthy -2.70 let's hope all our dreams and wishes come true

In recent days, of unrest and excitement this is especially true!!

[bobbydog]

It normally updates about an hour after the 12z completes but today it is stuck in the 6z index.

Thanks mate

[SMU]

The only slight frustration I have is we don't have any reanalysed data pre 76, so years like 59,62,69 have no OPI index data to begin to see where the baseline is in terms of going from Average Negative AO months, to SUPER negative AO months ( months sub -2 index)

 

That means its difficult to produce a forecast as there is only 1 month sub -2 which is 2009 @ -3.4. When increasing the sample size to 4 would have meant some upper OPI years & maybe a lower one.

 

I still think the baseline to deliver these Super Months needs to be sub -2.

 

For the moment the 4 years above are the super Winters - with 'assumed values' below -2

 

Heres the analogues. Winters DJF 60,63,70,77

 

 

 

Now add in the Extreme East QBO you are left with

2009 & 1962.

 

 

You will see the Easterly QBO strengthens the Greenland blocking.

 

ENSO favours 2009.

 

 

For November, lots of low heights over the UK with a southerly tracking jet, however possibly not south enough in most samples for a cold CET, only westerly QBO months were cold in November.

 

I would go with November 2014 Mild ( possibly very mild) & wet.75%

 

25% with the positive height anomaly over Western / Northern Russia & the southerly jet getting far enough south to bring in the cold air.

 

What I expect is something at the outer edges of the average with a strongly mild month or really cold. The only way I see average is a even split of Very warm compensated by very cold weather....

 

S

Edited October 15, 2014 by Steve Murr

[Weather-history]

OPI for oct 85 @-1.9 would then refer to the following winter 85/86? If so Dec was mild with a cold end, Jan chilly but not exceptionally so and Feb 86 one the coldest of the 20th century.

Yes that's right although you could include November which was cold and wintry last week.

[BLAST FROM THE PAST]

I have already committed myself to saying a week or so ago that we will see an OPI supportive of HLB. A week later we are still well on track and I think we can continue to be feeling positive on this piece of the puzzle. I still like the 76/77 composite

BFTP

Edited October 15, 2014 by BLAST FROM THE PAST

[damianslaw]

The only slight frustration I have is we don't have any reanalysed data pre 76, so years like 59,62,69 have no OPI index data to begin to see where the baseline is in terms of going from Average Negative AO months, to SUPER negative AO months ( months sub -2 index)

 

That means its difficult to produce a forecast as there is only 1 month sub -2 which is 2009 @ -3.4. When increasing the sample size to 4 would have meant some upper OPI years & maybe a lower one.

 

I still think the baseline to deliver these Super Months needs to be sub -2.

 

For the moment the 4 years above are the super Winters - with 'assumed values' below -2

 

Heres the analogues. Winters DJF 60,63,70,77

 

Analogue 1.png

 

 

Now add in the Extreme East QBO you are left with

2009 & 1962.

 

Analogues 2.png

 

You will see the Easterly QBO strengthens the Greenland blocking.

 

ENSO favours 2009.

 

 

For November, lots of low heights over the UK with a southerly tracking jet, however possibly not south enough in most samples for a cold CET, only westerly QBO months were cold in November.

 

I would go with November 2014 Mild ( possibly very mild) & wet.75%

 

25% with the positive height anomaly over Western / Northern Russia & the southerly jet getting far enough south to bring in the cold air.

 

What I expect is something at the outer edges of the average with a strongly mild month or really cold. The only way I see average is a even split of Very warm compensated by very cold weather....

 

S

 

 

Yes a southerly tracking jet could be a strong theme for November - how far south will dictate whether we see mild westerlies/southwesterlies or colder northwesterlies. I hope far enough south to prevent an absolute soaking such as we saw in 2009 - no one would want a repeat of Nov 09.

 

There is a theory wet mild Novembers or at least 'wet' Novembers increase chances of cold winters. Nov 09 certainly fell into this category. I do like a cold foggy spell in November though.

[Fingers]

Update of -0.76 from the 00z. 5 days time until we can see the month end prediction with, hopefully, less swings and fluctuations through the final 10 days as actual data overrides forecasts. Let's hope for the usual drop by the 18z!

One quick question. We obviously don't take the data from the 18z pub run with the daily figure polling from 12z and the figure (generally, but not always) does trend lower through the day. Am I right in assuming that the 12z is considered the most accurate GFS run as the 00z is too progressive etc?

Clem

Edited October 16, 2014 by Fingers

[bluearmy]

The gfs 00z runs isn't very good over the polar field for high anomalies through week 2. Gem looks a lot more promising at day 10 which provides some comfort in respect of where the OPI may end up.

Edited October 16, 2014 by bluearmy

[Blessed Weather]

Oh dear! The ECM/GFS 8-10 Day Mean 500 Height Anomaly output from the 12z runs yesterday have lost the positive anomaly over the Pole out at 23rd-25th Oct. Both models showing this. I've been checking this chart since quite early in the month and this is the first output to lose the anomaly.

 

Edited October 16, 2014 by Blessed Weather

[karyo]

Let's hope that both models have overeacted to a signal and the transition to lower heights across the polar region is slower so that it gives us valuable time for the final OPI. It would be so disappointing if the tables turn so fast considering how well we've done so far!

 

Karyo

[BLAST FROM THE PAST]

One would like to think that the last week of October if becoming positive wouldn't overturn 3 weeks of negative?

BFTP

[Duane S.]

One would like to think that the last week of October if becoming positive wouldn't overturn 3 weeks of negative?

BFTP

 

There would surely have to be a VERY anomalously positive pressure over the NH during the final 7 - 10 days to drive up the end score to anywhere near positve I would have thought.

 

I'm sure we would have gladly banked a final OPI figure of, say, -1 on the 1st of October but even that would seem a tad disappointing after the potential this month of a much lower figure and the potential winter goodies that would bring. The correlation to -AO winter months is greatly improved when < -1.5 (I think I read in this thread?), so this would be a nice to acheive if we can. If the 00z runs become the norm I suspect we are looking at between -0.5 and -1.0 end score.

 

Within 4 or 5 days the wilder run-run swings should start to abate I would imagine

Edited October 16, 2014 by s4lancia

[bluearmy]

We shouldn't forget that we only have solid data for half of the month plus a semi reliable forecast for the next week and uncertain for the last three days which drive the current OPI number.

From my perspective, a decently negative OPI will not carry much weight if we end the month with a raging p/v. That would seem unlikely but is certainly not discountable.

I still remain confused as to whether the daily opi number trends towards the final figure on the daily graph or whether it has to be derived once all the 31 days numbers are available?

[johnholmes]

Oh dear! The ECM/GFS 8-10 Day Mean 500 Height Anomaly output from the 12z runs yesterday have lost the positive anomaly over the Pole out at 23rd-25th Oct. Both models showing this. I've been checking this chart since quite early in the month and this is the first output to lose the anomaly.

 

ECM-GFS 8-10 Mean 500 Hgt Anomaly on 12z 15Oct2014.gif

 

don't lose heart with just one output, and even then if the NOAA 6-10 and 8-14 shows different again don't lose heart. As with ANY model output it is pretty useless just taking one output you have to watch them over several days for any consistent signals of the status quo persisting or what looks like a genuine change in pattern. Something that is a necessity with the 500mb anomaly charts.

That being the case, my comments on my daily file on the anomaly charts has this comment which may be of interest, written yesterday evening, updated after seeing the morning output from ec-gfs

 

Noaa

6-10

It is a similar chart to its last few, re trough having become more rounded, small +ve near azores, some indication of slight ridgining/+ve heights gt lakes and the continuation of ridging slightly and +ve heights w of the marked trough s of Alaska. Something that I commented on in net wx HAD been absent in the first part of October unlike much of last October. ?? Is it this responsible for the decreasing –ve OPI values=I am am not sure but it may be having some effect.

The ec-gfs has also shown something along these lines over the past few runs

8-14 and as is often the case it continues the ‘smoothing’ out of troughs/ridges with its treatment of the far w of the chart, alaska and west

Edited October 16, 2014 by johnholmes

[Interitus]

I still remain confused as to whether the daily opi number trends towards the final figure on the daily graph or whether it has to be derived once all the 31 days numbers are available?

 

It will (should!) trend towards a final figure. Regarding the heights, they haven't generally been as big a contributor to the index as the angle, averaging +0.136 since the 9th, range -0.23 to 0.43. The angle contribution has averaged -2.082 range -4.14 to 0.

[LomondSnowstorm]

Worth remembering too that the outer ranges of the OPI were driving a much more positive value on the 6Z a few days ago which didn't materialise, and we have all but banked a negative OPI. Given the +288 ECM ens that knocker posted on the MT had a positive anomaly over the pole even after in fairly underwhelming ECM ens suite out to day 10 I think the 00Z value is towards the upper end of the values we could reasonably expect to verify.

[radiohead]

We shouldn't forget that we only have solid data for half of the month plus a semi reliable forecast for the next week and uncertain for the last three days which drive the current OPI number.

From my perspective, a decently negative OPI will not carry much weight if we end the month with a raging p/v. That would seem unlikely but is certainly not discountable.

I still remain confused as to whether the daily opi number trends towards the final figure on the daily graph or whether it has to be derived once all the 31 days numbers are available?

 

I'm not really sure what use the early figures really are. The daily figure seems to be largely based on a GFS 10 day forecast, weve seen how wildly that swings from run to run even now so I'm not sure what value old GFS 10 day forecasts from earlier in the month are to us now. The wild swings to me indicate the normal flucations we see in the GFS output once you get to day 10, which is well into FI. I'm really not sure what adding up those past figures based on old GFS runs which would obviously contain a lot of "wrong" FI forecasts is of any value now apart from to show how the GFS output is unreliable at that range and liable to the kinds of swings such as what we now see this morning.

 

Not to mention these daily figures all seem to hinge on which GFS run the site last updated with on that day. Yesterday, it didn't update beyond the 06Z run, the 12Z could have changed the daily figure dramatically, but for whatever reason, it wasn't updated so now yesterdays figure becomes based on the 06Z. This wouldn't matter so much if there wasn't so much intra-run variance. It just makes the whole thing even more confusing and questionable to me...

 

 

Ok now I really will stop looking at the OPI until October 31st

Edited October 16, 2014 by radiohead

[Fingers]

We shouldn't forget that we only have solid data for half of the month plus a semi reliable forecast for the next week and uncertain for the last three days which drive the current OPI number.

From my perspective, a decently negative OPI will not carry much weight if we end the month with a raging p/v. That would seem unlikely but is certainly not discountable.

I still remain confused as to whether the daily opi number trends towards the final figure on the daily graph or whether it has to be derived once all the 31 days numbers are available?

Slightly controversial. That's the problem, we haven't got solid data due to the forecast element of the opi figure as it reads. Ideally you need a published daily month to date rolling figure (the trend) + a 10 day forecast figure. From this mornings 00z we could, emphasise could, have a positive opi for the month so far with the forecast 10 day period bringing it back to a negative -0.76 figure. Very unlikely but until we get very close to month end we can only go by our observations and hope the vortex remains subdued and delivers -1.5 as the trend appears to be negative.

[Yarmy]

The volatility is so large that it would seem to make more sense to use a 5-day forecast instead of 10, since there will only be relatively minor changes (usually) in that output.

 

FWIW, T240 on the 06z has weak heights over the pole again.

 

Edited October 16, 2014 by Yarmy

[LomondSnowstorm]

The volatility is so large that it would seem to make more sense to use a 5-day forecast instead of 10, since there will only be relatively minor changes (usually) in that output.

 

FWIW, T240 on the 06z has weak heights over the pole again.

 

 

Yep, far better looking output from the 6Z (for once!). I'd be surprised if the OPI value wasn't lower on this run. For comparison here's what the 00Z had at +240:

Vortex slap bang over the pole. 

Edited October 16, 2014 by LomondSnowstorm

[Fingers]

You almost want to extract the forecast element until the 27th when the GFS is in the semi reliable time frame and you've had 26 days of data. Last year though looked pretty reliable from the 21st to be fair. Let's just keep the daily Opi negative for now!

[beng]

The volatility is so large that it would seem to make more sense to use a 5-day forecast instead of 10, since there will only be relatively minor changes (usually) in that output.

 

 

Yes I was thinking exactly the same thing - or perhaps down the line they could issue both - a 5 day which would be more reliable and then the 10 day as well for the masochists. 

[mountain shadow]

 

FWIW, T240 on the 06z has weak heights over the pole again.

 

 

I would class a 1030mb high pressure over the pole at the end of October as pretty strong, not weak.

[KyleHenry]

Looking at the running OPI in comparison to Solar cycles 19 to 23

OPI for 1962/63 unfortunately not available guessing at a possible -4 figure.

1977 -1.75 .

1978. -0.95 both very cold and snowy UK.

1984 -1.8.

1985. -1.9

1986 -1.3 Very cold and snowy UK

1994-1996 +0.7, -0.66, -0.75

Still largely cold/snowy of UK.

2009-2011 speak for themselves.

All above in coincidence with Solar minimum.

All of the high positive OPI 1988, 1992, 2006,2013 (all snowless) all occurring in solar maximum.

Riccardo, I hope will be able to comment on relationship between both if any, in his out coming paper.

Comparison between both I used with only NW Europe region. ( UK) in mind.

When the final OPI for October 2014 comes in,A better anomaly year estimate can be made.

As we stand coming off double peak Solar max on the downward trend. East based QBO, and more or less weak El Niño.

2017-2022 next Solar Min I see some seriously extreme Winters inbound. OPI for those future Octobers will be very interesting.

Edited October 16, 2014 by KyleHenry

[karyo]

Yes, the 6z is a relief to see with high pressure back in the Arctic! Hopefully, the OPI when it updates will be much lower than -1.