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

[chionomaniac]

What happened to the updates of the OPI values? It has to be updated since yesterday morning.

Evidently there is no update if the GFS forecast is an outlier. They were emailed by a member of the Netweather forum on this side of the Atlantic and this was the response given:

 

Since The index is calculated both with consolidated and forecasted charts, if the forecasted pattern is out of a specific range (too wide variability) those runs are not used and you see only the last one we used. 

OPI team

[Crocodile23]

 

Evidently there is no update if the GFS forecast is an outlier. They were emailed by a member of the Netweather forum on this side of the Atlantic and this was the response given:

 

Since The index is calculated both with consolidated and forecasted charts, if the forecasted pattern is out of a specific range (too wide variability) those runs are not used and you see only the last one we used. 

OPI team

 

 

Hmm that's odd. Seems somewhat like cherry picking of data, but i guess most probably isn't and their method and procedure says this should be done. The method may have for example some limits that accepts or not the run and with these limits the r=0.91 of the correlation has been calculated.

 

 

I should add that i guess this should not matter for the final value of OPI in 31th of October. Right? Because that should be based on actual data right?

[chionomaniac]

Hmm that's odd. Seems somewhat like cherry picking of data, but i guess most probably isn't and their method and procedure says this should be done. The method may have for example some limits that accepts or not the run and with these limits the r=0.91 of the correlation has been calculated.

 

 

I should add that i guess this should not matter for the final value of OPI in 31th of October. Right? Because that should be based on actual data right?

I think that it may have something to do with the Asian forecast anomaly. The fact is that they can ignore any forecast data they like, but they cannot cherry pick the consolidated actual data. As someone pointed out over here, why would you cherry pick data anyway, as it is a correlation for a winter that hasn't occurred yet. As ever, we wait for the paper....

[Crocodile23]

I think that it may have something to do with the Asian forecast anomaly. The fact is that they can ignore any forecast data they like, but they cannot cherry pick the consolidated actual data.

 

Exactly and this is what matters after all. The in-between values that we get now is just for fun.

 

As someone pointed out over here, why would you cherry pick data anyway, as it is a correlation for a winter that hasn't occurred yet. As ever, we wait for the paper....

 

I'm looking forward to seeing it. Although i have not yet got an answer about the approximate time of publication, even though i asked 4-5 times from last year

BTW what your stratospheric analysis of now signals for the winter's AO? Does it concur about a generally negative AO? Or this can't be predicted from so early(stratospherically speaking)?

[chionomaniac]

Exactly and this is what matters after all. The in-between values that we get now is just for fun.

 

 

I'm looking forward to seeing it. Although i have not yet got an answer about the approximate time of publication, even though i asked 4-5 times from last year

BTW what your stratospheric analysis of now signals for the winter's AO? Does it concur about a generally negative AO? Or this can't be predicted from so early(stratospherically speaking)?

Using analogues prior to October's snow gain, one would expect a warmer than average stratosphere leading to a negative AO. Blocking over the north Atlantic/Greenland could be a feature Dec/January. If anything the first two weeks of October have cemented these thoughts- and that is without even considering the OPI. But November is the big month when we really get an idea of how the stratosphere will influence early winter. Never make a winter forecast too early without taking the November stratosphere into consideration is my motto!

 

Surprised there isn't a dedicated strat thread on this side of the Atlantic!

[Spwild47]

Correct

quote name="Crocodile23" post="3088754" timestamp="1413653495"]Hmm that's odd. Seems somewhat like cherry picking of data, but i guess most probably isn't and their method and procedure says this should be done. The method may have for example some limits that accepts or not the run and with these limits the r=0.91 of the correlation has been calculated.

 

 

I should add that i guess this should not matter for the final value of OPI in 31th of October. Right? Because that should be based on actual data right?

[Spwild47]

October Eurasian snowcover anomalies looking robust ...especially south (of 60N).

[frosty]

This animated view of the isobaric winds at 10hpa could be very telling in

that surely this would relate to a stronger BDC pattern and a greater EP flux

movement northward.

High latitude blocking could be a strong feature going forward into late

Autumn and winter.

http://earth.nullschool.net/#current/wind/isobaric/10hPa/orthographic=0.64,81.72,512

Click on EARTH in the left hand corner to see the relevant information

[CoastalWx]

This animated view of the isobaric winds at 10hpa could be very telling in

that surely this would relate to a stronger BDC pattern and a greater EP flux

movement northward.

High latitude blocking could be a strong feature going forward into late

Autumn and winter.

http://earth.nullschool.net/#current/wind/isobaric/10hPa/orthographic=0.64,81.72,512

Click on EARTH in the left hand corner to see the relevant information

 

Pardon me asking, but would this be at all due to the appearance of a weak + height anomaly over extreme NE Siberia? I note we don't have a clear massive strat PV to keep the BDC weaker.

[griteater]

Latest OPI based on the 00z GFS is in at -3.04.  Given what we've seen to date in Oct, along with the projected pattern through the end of the month, it looks quite possible that the OPI will end up at -2.5 or lower.  If that's the case, it would easily be the 2nd lowest reading on record back to 1976.  Only one that would be lower is Oct 2009 at ~ -3.4

[Wow]

-2.75 today.  Definitely looking like we'll be between -2 and -3.

[Spwild47]

I read through the netweather post and someone pointed out that elypticization is a simple formula...((Asia gpt + America gpt)-(Arctic gpt))/10

So in the post above for Elypticization you get ((2.56+2.81)-2.61)) / 10 = .28

As for Axes it is the angle between the current axis vs. Some ideal angle of 34%? Between the oceans, and is modified by some factor that apparently gets exponentially larger as you approach the ideal angle..

So at this point the negative OPI is driven almost entirely by the axis vs. Gpt anomaly...

[hosj III]

Last update has it around -2.38. It has really stabilized in the past few days and it will likely end up between -2.1 and -2.5, which would predict the third lowest DJF AO on record. (Behind 1976-1977 and 2009-2010). It will be interesting to see if it is accurate this winter or if it will bust again like last winter. I'm not really sure anyone here believes this on face value, but it's becoming clear that the AO should be predominantly negative this winter.

[griteater]

Last update has it around -2.38. It has really stabilized in the past few days and it will likely end up between -2.1 and -2.5, which would predict the third lowest DJF AO on record. (Behind 1976-1977 and 2009-2010). It will be interesting to see if it is accurate this winter or if it will bust again like last winter. I'm not really sure anyone here believes this on face value, but it's becoming clear that the AO should be predominantly negative this winter.

 

hosj III - the second lowest OPI was actually in Oct 1985, near -2.0.  Take a look at that OPI chart again - the Oct 1976 value that was sub -2 was the actual AO for that winter, but the Oct 1976 OPI was around -1.8.  So this October's OPI could very well be the 2nd lowest since 1976.

[GaWx]

Last update has it around -2.38. It has really stabilized in the past few days and it will likely end up between -2.1 and -2.5, which would predict the third lowest DJF AO on record. (Behind 1976-1977 and 2009-2010). It will be interesting to see if it is accurate this winter or if it will bust again like last winter. I'm not really sure anyone here believes this on face value, but it's becoming clear that the AO should be predominantly negative this winter.

 

 I've been busy monitoring Eurasian SCE/SAI. So, I haven't given the OPI sig. attention. To be perfectly honest, I admit that I don't fully understand how the OPI is calculated and how it differs from just taking the  average AO for Oct. I'm assuming it may weigh certain parts of the month differently from others. Can someone please explain the major differences from just averaging the AO in a nutshell/simple manner? If so, I may gain interest in this in addition to continuing to closely follow the SCE/SAI as well as the simple averaging of Oct. AO days. If it is too complicated, I'll probably not be as interested.

 

In layman's terms, wtf is the OPI?

[hosj III]

 I've been busy monitoring Eurasian SCE/SAI. So, I haven't given the OPI sig. attention. To be perfectly honest, I admit that I don't fully understand how the OPI is calculated and how it differs from just taking the  average AO for Oct. I'm assuming it may weigh certain parts of the month differently from others. Can someone please explain the major differences from just averaging the AO in a nutshell/simple manner? If so, I may gain interest in this in addition to continuing to closely follow the SCE/SAI as well as the simple averaging of Oct. AO days. If it is too complicated, I'll probably not be as interested.

 

In layman's terms, wtf is the OPI?

The exact equation hasen't been revealed, but it appears to calculate a number based on the 500mb height anomalies in the northern hemisphere. 

I hope these charts from the first page of this thread explain it:

 

 

 

 

It seems like using the OPI for DJF AO rather than the simple October AO is more accurate because it includes the Aleutian low/+AO that both correlate with a -AO in DJF. Other than that, it seems the north-central siberian ridge is the most important feature of a -OPI October.

[GaWx]

hosj,

 I appreciate your reply! This is very interesting though obviously more nebulous to follow than the straight AO and the Eurasian SCE/SAI being that I'm a hard stats kind of guy. If that 0.91 correlation is accurate and not fudged in any way or backward fitted, it would be about the largest correlation of any I've ever seen with wx indices!

[hosj III]

hosj,

 I appreciate your reply! This is very interesting though obviously more nebulous to follow than the straight AO and the Eurasian SCE/SAI being that I'm a hard stats kind of guy. If that 0.91 correlation is accurate and not fudged in any way or backward fitted, it would be about the largest correlation of any I've ever seen with wx indices!

The main challenge to the credibility of this index is its failure last year. The predicted +1.6 verified as +.18.

[Quixotic1]

-2.12 the final number?

[WidreMann]

The main challenge to the credibility of this index is its failure last year. The predicted +1.6 verified as +.18.

It got the sign right, and, as others have analyzed, the polar pattern was more like a +AO, but the +AO was off-center because of the strongly -EPO and that resulted in the index being lower than it otherwise would have been, but the pattern still made sense.

[GaWx]

It got the sign right, and, as others have analyzed, the polar pattern was more like a +AO, but the +AO was off-center because of the strongly -EPO and that resulted in the index being lower than it otherwise would have been, but the pattern still made sense.

 

Also, the mean DJF AO going back to 1950 is ~-0.30 as opposed to 0. So, a +0.18 for DJF is more like a +0.48 vs. the mean DJF AO. I wonder if this is taken into account by the OPI #.

[hosj III]

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

 

Final value and winter outlook up.

[Ginx snewx]

concur, ran a composite of all October's with AO days LESS than -2 in Oct. 15/20 winters in ORH were at or above climo snow with 11 out of those 15 significantly or well above normal. The years following winters composite 5H pattern.

their final looks exactly like this

[Absolute Humidity]

OPI -AO projection of -2.18 matches well with SAI, of which Cohen says Oct 2014 is the highest since its inception. 

[40/70 Benchmark]

OPI -AO projection of -2.18 matches well with SAI, of which Cohen says Oct 2014 is the highest since its inception.

I thought this season's SAI, like the SCE, was second to 1976?

[Absolute Humidity]

I thought this season's SAI, like the SCE, was second to 1976?

From Dr Cohen's blog:

"The snow advance index (SAI; which is more strongly correlated with the AO than the SCE) is the highest value since daily snow cover has been available (1997)."

https://www.aer.com/science-research/climate-weather/arctic-oscillation

However, you too are probably correct as well since in 1976 there was no SAI due to lack of daily values but if there were, it probably would beat oct 2014.

[Bob Chill]

I wonder how the OPI guys feel this winter is going based on their index. 

[hosj III]

I wonder how the OPI guys feel this winter is going based on their index. 

 

We can get a -6 AO February.

[Absolute Humidity]

I wonder how the OPI guys feel this winter is going based on their index.

Well, it was nagative five out of the last six days so it's a start.

[donsutherland1]

For purposes of keeping score on the state of the AO, the following averages will be needed during the remainder of meteorological winter to achieve a neutral or negative December-February average:

 

DJF Average of 0.000: -1.037 (somewhat unlikely according to 1950-2014 data, but still possible)

DJF Average of -0.500: -2.191 (very unlikely)

DJF Average of -1.000: -3.344 (extremely unlikely)

 

For another perspective, if the AO is negative for the remainder of meteorological winter, 54% of days during the DJF period would have a negative AO and 46% would have a positive AO.