Jump to content
Cold?
Local
Radar
Snow?

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


Recommended Posts

Posted
  • Location: Beccles, Suffolk.
  • Weather Preferences: Thunder, snow, heat, sunshine...
  • Location: Beccles, Suffolk.

    I've always thought that the synoptic pattern 'dictates' the NAO, and not the other way round? :cc_confused:

    Edited by Ed Stone
    • Like 1
    Link to post
    Share on other sites
    Spotted a post you think may be an issue? Please help the team by reporting it.
    • Replies 1.5k
    • Created
    • Last Reply

    Top Posters In This Topic

    Top Posters In This Topic

    Popular Posts

    As ever, a great summary of how things stand from Ed and I think for the last 19, or pretty much 20 days now we have all been waiting for this point in the month and beyond. It's been rather 'fun' wat

    The discussion of the OPI index trend is understandably causing a lot of interest and (cautious) excitement amongst members after Riccardo's very interesting opening post that started this thread. The

    The monitoring page of the “October pattern index†(OPI) is presented. The OPI index  represents a analytic synthesis of the October hemispheric pattern at 500 hPa, and shows from  previous studi

    Posted Images

    Posted
  • Location: Aylesbury, Buckinghamshire
  • Weather Preferences: Winter: Cold & Snowy, Summer: Just not hot
  • Location: Aylesbury, Buckinghamshire

    I would quite liked to have seen this resurrected. Nothing is 100% infallible otherwise forecasters wouldn't have a job. Plus, the hysteria/borderline obsession with this last year would be have been lessened.

    • Like 4
    Link to post
    Share on other sites
    Posted
  • Location: Truro, Cornwall
  • Weather Preferences: Winter - Heavy Snow Summer - Hot with Night time Thunderstorms
  • Location: Truro, Cornwall

    So, 70% of 83%, if my maths is right the OPI percent correlation with the CET average is 58.1% ? So in fact not much higher than 50/50, and that's to correlate with the CET, not necessarily to give a significantly below average month...........

     

    When you actually break down the correlations, the question starts to loom 'what was all the fuss about last year' ??? Perhaps less of a fault with the OPI, and more a fault of peoples tendency to over-excitement when any kind of predictor seems to be suggesting it'll be a cold winter ???

     

    T'was ever thus.........................

    Indeed if those stats are right then as you say its a pretty marginal thing really. Near 50% certainly does not equal a strong correlation, far from it. There are always variable factors at play as well that margins of error should be applied. But its not surprising it was hyped up so much especially by the hardcore winter enthusiasts as it gave something to pass the time until winter arrived but no year is ever the same as one the before or any other in fact. Clearly a lot dont possess much scientific insight into the matter. I would ditch something after 1 go. Always worth seeing how it goes over a number of times. Sometimes and often you can strike lucky with some things.

    Link to post
    Share on other sites
    Posted
  • Location: Surrey and SW France.
  • Location: Surrey and SW France.

    I did a bit of digging on this and found the group to be dedicated enthusiasts (much as we are) but seem to have a fondness for seeing patterns where none may exist.

     

    Note the mission statement of the Centro MeteoToscana.....

     

     

    CentroMeteoToscana is pleased to report that analysts CSCT are:

    Batstef, Burian, Cloover, Brand Fermana, Montel, Grandfather Rhone

    The CSCT, which collects the best analysts of the web at a national level, was created with the intention to propose a different approach to the Meteorology, characterized by people who do not just analyze the cards, but who go on to develop studies and theories keeping updated on latest developments in the teleconnection indices and impact circulatory macroscale.

    Being a member of the CSCT is therefore certainly an important award but also a task of undoubted responsibility, moral and intellectual.

     

    The first blurb in 2013....

     

    http://www.meteonetwork.it/cronaca-meteo/october-pattern-index-opi-un-nuovo-indice-altamente-predittivo-stagione-invernale

     

    ... and some of their other 'new' theories.

     

    http://www.centrometeotoscana.it/forum/index.php?topic=8875.0

     

    http://www.centrometeotoscana.it/forum/index.php?topic=6926.0

     

     

    • Like 2
    Link to post
    Share on other sites
    Posted
  • Location: Hadleigh, Suffolk
  • Weather Preferences: An Alpine climate - snowy winters and sunny summers
  • Location: Hadleigh, Suffolk

    Remember that the OPI is only correlated with the AO 83% of the time so we may have just gotten unlucky. Even then though, the AO only correlates with the CET about 70% of the time.

     

     

    That was my thought too, it was never touted as a guarantee....though a positive figure is almost always the death knell on the coming winter (as has been shown by the research). Maybe we should wait for another significantly negative value year to enable us to better make a judgement?

     

    I agree with you two guys. I do not agree with other posters who have made a tenuous link between OPI & AO correlation (83%) and then AO & CET correlation (70%) to arrive at their conclusion that it's the OPI that's probably wrong (70% of 83% = 58%). We all look for a negative AO to increase the UK's chance of a cold winter, but we also know a negative AO doesn't always deliver for our small island. So to say that because the OPI forecast a negative AO, but the negative AO didn't deliver a cold (CET) winter for the UK, so it's the OPI's fault, is a wrong conclusion.

     

    However, I note Nouska's comments with interest, and agree with Nick L that it would have been good to see another year's OPI with less excitement. The OPI may be a false dawn, but it's the incorrect associations and maths I disagree with.

    Link to post
    Share on other sites
    Posted
  • Location: Huddersfield, 145m ASL
  • Weather Preferences: Lots of snow, lots of hot sun
  • Location: Huddersfield, 145m ASL

    No-one's saying the OPI was wrong, just that the fact the OPI was predicting a negative AO, (which as Pete pointed out above is actually just saying that pressure in the Arctic will be above average), in no way justified the somewhat over-excited conclusions people drew from it's index figure last October. In other words, although the Index figure suggested winter 14/15 was more likely to have had a higher than average proportion of negative AO phases, the actual correlation of that probability to it occuring was in fact only somewhere between 55% and 60%, so not the high level of correlation the reaction to last year's OPI figure suggested was the case.

     

    Nothing wrong the OPI as such, the error was in the interpretation of it's output................

     

    And just to reiterate, the AO, and the NAO, are not weather phenomena in themselves, they do not cause weather patterns, they are simply methods of describing pressure phases in those areas, so when the AO is negative it simply means pressure in the Arctic is higher than average, and the reason that we get excited by that is because, in winter, favourably placed high pressure in that region often results in cold weather for the UK, (the fabled Greenland high for example).

    • Like 5
    Link to post
    Share on other sites
    Posted
  • Location: Hadleigh, Suffolk
  • Weather Preferences: An Alpine climate - snowy winters and sunny summers
  • Location: Hadleigh, Suffolk

    No-one's saying the OPI was wrong, just that the fact the OPI was predicting a negative AO, (which as Pete pointed out above is actually just saying that pressure in the Arctic will be above average), in no way justified the somewhat over-excited conclusions people drew from it's index figure last October. In other words, although the Index figure suggested winter 14/15 was more likely to have had a higher than average proportion of negative AO phases, the actual correlation of that probability to it occuring was in fact only somewhere between 55% and 60%, so not the high level of correlation the reaction to last year's OPI figure suggested was the case.

     

    Nothing wrong the OPI as such, the error was in the interpretation of it's output................

     

    And just to reiterate, the AO, and the NAO, are not weather phenomena in themselves, they do not cause weather patterns, they are simply methods of describing pressure phases in those areas, so when the AO is negative it simply means pressure in the Arctic is higher than average, and the reason that we get excited by that is because, in winter, favourably placed high pressure in that region often results in cold weather for the UK, (the fabled Greenland high for example).

     

    No problems with that. Thank you for your explanation.

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: Leeds/Bradford border, 185 metres above sea level, around 600 feet
  • Location: Leeds/Bradford border, 185 metres above sea level, around 600 feet

     

    I agree with you two guys. I do not agree with other posters who have made a tenuous link between OPI & AO correlation (83%) and then AO & CET correlation (70%) to arrive at their conclusion that it's the OPI that's probably wrong (70% of 83% = 58%). We all look for a negative AO to increase the UK's chance of a cold winter, but we also know a negative AO doesn't always deliver for our small island. So to say that because the OPI forecast a negative AO, but the negative AO didn't deliver a cold (CET) winter for the UK, so it's the OPI's fault, is a wrong conclusion.

     

    However, I note Nouska's comments with interest, and agree with Nick L that it would have been good to see another year's OPI with less excitement. The OPI may be a false dawn, but it's the incorrect associations and maths I disagree with.

    My suspicion is that low sample sizes inflate how often it correlates with the AO which was the issue the last 2 winters.

    The OPI may well have merit but at a more reasonable correlation.

    • Like 2
    Link to post
    Share on other sites
    Posted
  • Location: Chandlers Ford, south of Winchester.
  • Location: Chandlers Ford, south of Winchester.

    Has anyone got the link for 2015-2016 OPI figure? Riccardo?

    This may not have delivered last year but the correlation on previous years is pretty good and I would like to watch it develop, though maybe not as intensely as last year!

    • Like 2
    Link to post
    Share on other sites
    Posted
  • Location: Wroclaw, Poland
  • Location: Wroclaw, Poland

    AO predictions based on OPI Index are getting worse since 2012. There was amazing correlation 0.97 for 2000-2012 period, now it's only 0.73 for 2000-2014. It can be simple to explain If we had "calibarition period" before 2012.

     

    opi2014.png

    Link to post
    Share on other sites
    Posted
  • Location: Broadmayne, West Dorset
  • Weather Preferences: Snowfall in particular but most aspects of weather, hate hot and humid.
  • Location: Broadmayne, West Dorset

    I was sad to see the OPI get a bit of a kicking after having created quite a bit of excitement last Autumn.

     

    Unfortunately in our got to have it now world everyone wants the finished article straight away. I always viewed it as a work in progress

    and hopefully the Riccardo and his chums can hone the theory somewhat. The comment about sample size are valid and so it will only be as we go forward that we will see if the OPI can improve it's performance.

    • Like 2
    Link to post
    Share on other sites
    Posted
  • Location: Near King's Lynn 13.68m ASL
  • Weather Preferences: Hoar Frost, Snow, Misty Autumn mornings
  • Location: Near King's Lynn 13.68m ASL

    The real problem is that not a single one of us knows how the figure is calculated. No paper has been published and the method described only vaguely. If there are no proprietary secrets or commercial interests here why not simply release the exact steps and let people verify independently. Until then, given the remarkable correlation, the suspicion is that we have 40 years of data curve-fitted and a 1 year AO forecast.

     

    I haven't changed my opinion, and the resounding silence makes me think the creators have thrown in the towel too. It's not difficult to data mine and pattern match, but there needs to be some physical basis for it. I could take the last 50 years of PDO indices and do a polynomial fit using Matlab which would then enable me to make a PDO forecast arbitrarily far into the future. It would be completely meaningless though and would fail quickly.

    Link to post
    Share on other sites
    Posted
  • Location: Cyprus and North east Swindon. 104 ASL
  • Location: Cyprus and North east Swindon. 104 ASL

    Is the OPI running again this Oct?  And if so will there be a new forum page to follow it?  Massive fail last year but it is still an indicator just like snow cover and sea temps - it would be good to give it chance again.

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: Pershore
  • Location: Pershore

    There's nothing stopping a new OPI thread being opened for this year at any time, if anyone wants to/there's any new data to discuss :)

    Link to post
    Share on other sites
    Posted
  • Location: Exile from Argyll
  • Location: Exile from Argyll

    What are you going to talk about? It is meaningless when we don't know how it is calculated. They did say they were not measuring AO index but angles of the 500mb anomalies in the AO region;

    • Like 1
    Link to post
    Share on other sites
    • 2 years later...
    Posted
  • Location: Rotherhithe, 5.8M ASL
  • Location: Rotherhithe, 5.8M ASL
    On 30/09/2015 at 23:29, joggs said:

    Massive fail last year,especially our Italian counterparts.

    Think you scared them off well done Rome wasn’t built in a day. :cold:

    Would be good to see a revival but as much chance as knocker loving cold and snow, as there was statistical significance? Shame to see work go to waste.

    • Like 2
    Link to post
    Share on other sites
    • 1 year later...
    Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.

    I am going to be calculating this based on re-analysis of previous years / months and comparing that to this Octobers charts, i will not be bothering with GFS runs predicting what might happen, that seems pointless, just analysis charts from midday and perhaps midnight, i may also do one that runs from mid October to Mid November based on my theory that climate change is causing everything to happen later and that could be what is buggering some very respectable forecasters winter forecasts in recent years.

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: The Purbeck Microclimate, Dorset.
  • Weather Preferences: Gales, T-storms, Heavy Rain, Heat, Cold - Love it all.
  • Location: The Purbeck Microclimate, Dorset.

    I cannot see how you can calculate it without the algorithmic software? Also, with quite a bit of emphasis placed on the GFS, how do you expect to even come close to what the original system calculated?

    The OPI is buried. No papers approved and now the site is dead. Speaks volumes.

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
    1 minute ago, Mapantz said:

    I cannot see how you can calculate it without the algorithmic software? Also, with quite a bit of emphasis placed on the GFS, how do you expect to even come close to what the original system calculated?

    The OPI is buried. No papers approved and now the site is dead. Speaks volumes.

    How is emphasis placed on the GFS?

    Link to post
    Share on other sites
    Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.

    This isn't dead at all, i am going to revive it - pheonix from the flames.

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: The Purbeck Microclimate, Dorset.
  • Weather Preferences: Gales, T-storms, Heavy Rain, Heat, Cold - Love it all.
  • Location: The Purbeck Microclimate, Dorset.
    Just now, feb1991blizzard said:

    This isn't dead at all, i am going to revive it - pheonix from the flames.

    Of course it is dead. I think you're just on the wind up to be honest.

    3 minutes ago, feb1991blizzard said:

    How is emphasis placed on the GFS?

    Did you even read the first post?

    • Like 1
    Link to post
    Share on other sites
    Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
    1 minute ago, Mapantz said:

     

    Did you even read the first post?

    Yes, but thats just the predictive side, one Oct 31st comes its irrelevant.

    Link to post
    Share on other sites
    Posted
  • Location: The Purbeck Microclimate, Dorset.
  • Weather Preferences: Gales, T-storms, Heavy Rain, Heat, Cold - Love it all.
  • Location: The Purbeck Microclimate, Dorset.
    9 minutes ago, feb1991blizzard said:

    Yes, but thats just the predictive side, one Oct 31st comes its irrelevant.

    So, let me get this straight...

    You're going to try to calculate the OPI again without knowing exactly what their software did? And you're not going to bother using some parts of the theory that they did provide?

    It reminds me of the infinite monkey theorem lol

    I should add that i'd applaud you if you stumbled across anything, but with those starting points, you're going to have a hard time with it.

    Edited by Mapantz
    Link to post
    Share on other sites
    Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
    1 minute ago, Mapantz said:

    So, let me get this straight...

    You're going to try to calculate the OPI again without knowing exactly what their software did? And you're not going to bother using some parts of the theory that they did provide?

    It reminds me of the infinite monkey theorem lol

    I am going to calculate it by looking at the polar profile that VERIFIED - not what was predicted.

    Link to post
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now
    ×
    ×
    • Create New...