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

[Crocodile23]

Interesting post however the independence tests don't really work when you have selected specific data and combined them in a particular way to match the historical series which is what has been done here.

What? Why not?

 

Its not like the OPI is say a relatively simple temperature or pressure measurement at one point.

 

Well temperature is not simple either. It's a combination of billions of measurements of the kinetic energy of billions and billions molecules.

OPI is a specific value, temperature is a specific value and that's what matters when we do correlation tests. The validity of the results and the hidden variables behind is another issue.

[SomeLikeItHot]

Well temperature is not simple either. It's a combination of billions of measurements of the kinetic energy of billions and billions molecules.

OPI is a specific value, temperature is a specific value and that's what matters when we do correlation tests. The validity of the results and the hidden variables behind is another issue.

Temperature is a measure chosen for another purpose, if the october CET anomaly correlated with the winter CET anomaly then it would be interesting to measure the significance.

 

However if I took the last 24 months of CET data and combined it in such a way that it would have a 80% correlation with the next winter CET anomaly calibrated over the last 30 years. a significance test wouldn't really tell me much only a testing going forward (or using historical data its not calibrated to) would determine whether there was something there. The second situation is closer to the OPI in my opinion.

 

Which brings something to mind/ Is it possible to reconstruct OPI figures (and winter AOs) from earlier dates than the 30years the index was developed on. Isn't there reanalysis that goes back for a much longer period?

[Crocodile23]

Temperature is a measure chosen for another purpose, if the october CET anomaly correlated with the winter CET anomaly then it would be interesting to measure the significance.

 

However if I took the last 24 months of CET data and combined it in such a way that it would have a 80% correlation with the next winter CET anomaly calibrated over the last 30 years. a significance test wouldn't really tell me much only a testing going forward (or using historical data its not calibrated to) would determine whether there was something there. The second situation is closer to the OPI in my opinion.

And combined them in such a way??

I don't really understand what you are saying.

 

Here is what i've done:

The OPI team tried to create an index(OPI) that would correlate well with winter AO(another index).

 

I took these OPI values(ALL of them, i haven't selected them with any special criteria) for the Octobers and have tried to see if they correlate with the temperatures(in 850 hPa) of a specific part of the earth(your area, UK).

I have not picked special data, i have not selected anything, i have taken all years that OPI has been calculated(from 1976 till 2013) and calculated the correlations.

 

And i have found that OPI and your winter's temperatures are correlated. I have not calibrated anything, i have not picked any specific data to improve the fitting of the correlations, nothing like that at all! I have chosen ALL the OPI values.

 

So what you say is completely mistaken.

 

Which brings something to mind/ Is it possible to reconstruct OPI figures (and winter AOs) from earlier dates than the 30years the index was developed on. Isn't there reanalysis that goes back for a much longer period?

 

That would be interesting to see the correlation from 1948 to 2013 for example.

Edited October 24, 2014 by Crocodile23

[Man With Beard]

Croc23 - great post, backs up a theory that a sub -1.50 OPI leads to a much colder winter perhaps??!

However, where exactly are these figures for 2013/14 from? Surely it was warmer than that!

[Nouska]

Temperature is a measure chosen for another purpose, if the october CET anomaly correlated with the winter CET anomaly then it would be interesting to measure the significance.

 

However if I took the last 24 months of CET data and combined it in such a way that it would have a 80% correlation with the next winter CET anomaly calibrated over the last 30 years. a significance test wouldn't really tell me much only a testing going forward (or using historical data its not calibrated to) would determine whether there was something there. The second situation is closer to the OPI in my opinion.

 

Which brings something to mind/ Is it possible to reconstruct OPI figures (and winter AOs) from earlier dates than the 30years the index was developed on. Isn't there reanalysis that goes back for a much longer period?

 

Using reanalysis which goes back to 1948, I looked at the October Z500 anomaly for the three coldest winters in France, prior to the beginning of the OPI data set.

 

 

The first and last chart had a November Canadian warming which would, I assume, dramatically alter the pattern portrayed.

 

What is clear to see is that the overall pattern is very different to that which is discussed in the Taymyr theory - low geopotential heights over the peninsula on all of them.

 

May be ice loss but money would be on the recent solar changes as being the overall driver in this change of circulation pattern.

[PatrickIRE]

 

I have made a simple calculation about the correlation between October's OPI and the average temperature of following winter in the 850 hPa level for a big portion of UK. (I provide the values i used in the end of this post)

 

So i took the October's OPI value and the following winter's average temperatures in degrees Celcius in the 850 hPa level for december+january+february, for the area that is defined by 5° W to 0° E and 50° N to 55° N (that contains most of England and Wales) and have calculated Pearson's r correlation coefficient and Kendall's tau.

The values of temperatures have been acquired from NOAA and the values for OPI from a post here.

 

So the calculated values between October's OPI compared to following December+January+February average temperature in 850 hPa were:

 

Pearson's r = 0.76

 

Kendall's tau = 0.55

2-sided p-value = 0.0000014

 

Pearson's r value is relatively high and shows that something might going on with this OPI thing.

Even more, in the Kendall's calculation, we see a smaller correlation but a very interesting p-value, since the almost zero 2-sided p-value shows that assuming(the null hypothesis) October OPI and average temperature of DJF are independent, we have an extremely small chance of obtaining this 0.55 correlation if October OPI and average temperature of DJF are independent.

So we must assume they are dependent with a +0.55 (positive) correlation.

Positiv emeans that when the value of October OPI is large(get's larger) the average temperature of DJF is large(get's larger) also and when the values of October OPI decrease, the value of average temperature of DJF decrease also.

 

So all in all there is something interesting with this OPI.

And it seems to suggest that a correlation with your winter's temperature lies between them. A positive OPI brings warmer winter and a negative colder. With the aforementioned correlations of course. This is NOT a strict 1 to 1 rule of course.

 

 

The data:

(For example for year 1980:

OPI value is for October 1980

Average temperature(in 850 hPa) is for December 1980 + January 1981 + February 1981)

YEAR	OPI	AVG1976	-1,75	-3,311977	-0,95	-1,901978	-1,80	-3,081979	-0,50	-1,711980	-0,05	-2,051981	-0,40	-1,771982	-1,10	-1,841983	-0,30	-1,511984	-1,80	-2,191985	-1,90	-3,311986	-1,30	-2,381987	-0,45	-0,981988	+1,45	+0,491989	+0,25	+0,061990	+0,55	-2,611991	+1,10	+1,651992	+1,75	-0,011993	-0,15	-2,141994	+0,70	-0,841995	-0,65	-2,051996	-0,75	-1,021997	-0,75	+0,311998	+0,10	-1,061999	-0,20	-1,462000	-0,75	-1,262001	+0,45	-0,482002	-0,90	-1,232003	-0,70	-0,622004	+0,30	-1,572005	-0,70	-1,102006	+0,85	+0,472007	+0,75	+0,902008	+0,25	-1,032009	-3,15	-3,882010	-0,85	-1,902011	+0,65	-0,852012	-1,65	-1,902013	+1,60	-0,53

 

I checked the data you kindly provided and came up with the same results as yourself (along with an R² value of 0.5723704) However,  this is just taking these values on blind faith rather than substance so it leaves many questions to be answered.

Edited October 24, 2014 by PatrickIRE

[Crocodile23]

Croc23 - great post, backs up a theory that a sub -1.50 OPI leads to a much colder winter perhaps??!

However, where exactly are these figures for 2013/14 from? Surely it was warmer than that!

 

As i've said it's from NCEP's(of NOAA) reanalysis project.

Here are the complete temperature statistics for the 50° N to 55° N , 5° W to 0°E (Greenwich meridian), area**:

 

**The area:

 

Monthly average 850 hPa temperatures in degrees Celcius of the above area:

Year     Jan      Feb      Mar      Apr      May      Jun     Jul      Aug       Sep      Oct      Nov      Dec1948   -1.703   -2.286    2.857    0.303    2.981    4.046    6.110    5.970    6.456    3.316    4.646    1.2001949   -0.064   -0.148   -1.788    2.247    1.679    6.478    7.831    7.794    7.674    4.519   -0.097   -1.2871950   -0.151   -1.811    1.296   -1.948    3.037    7.128    6.450    6.303    4.570    3.111   -1.170   -5.1711951   -2.048   -3.699   -2.868   -1.976    1.050    4.783    7.396    5.386    6.802    4.106    1.427   -0.1041952   -3.832   -2.590   -0.649    0.786    4.500    4.581    6.918    6.482    1.662    1.438   -2.481   -1.9431953   -1.077   -2.124    1.382   -1.410    4.040    4.874    5.793    6.758    6.048    3.709    3.197    2.0391954   -3.292   -3.707   -0.819   -0.973    1.599    4.189    4.282    5.683    3.488    4.391    0.260    0.2311955   -1.722   -5.744   -4.092    1.593   -0.086    4.863    9.357    8.761    5.510    2.432    1.698    0.1171956   -1.807   -7.034   -0.263   -1.850    2.656    3.637    6.846    4.332    7.247    2.971    0.151   -0.1001957   -0.578   -2.006    3.277    0.188    1.304    6.432    7.089    6.826    4.528    4.702    1.377   -0.3811958   -1.882   -1.767   -3.226   -1.850    1.596    4.481    6.369    6.793    7.064    4.858    1.384   -1.2381959   -2.566    1.489    0.092    0.188    4.489    5.704    7.810    8.344    7.407    5.089    0.394   -0.1641960   -2.594   -2.537    0.506    0.027    3.971    7.022    4.797    5.371    5.164    2.522    0.189   -2.2201961   -2.328    1.381    1.767    1.292    1.774    5.136    5.380    6.222    7.433    3.313   -0.076   -0.6771962   -2.021   -3.442   -5.491   -1.038    0.010    3.734    4.646    4.927    4.538    5.257   -0.571   -1.8971963   -6.757   -6.320   -0.968   -0.441    0.421    5.036    5.048    4.418    4.612    4.927    0.691   -1.9341964    0.142   -2.376   -3.209   -1.096    3.878    4.202    6.516    6.054    6.551    2.577    1.748   -3.0511965   -3.298   -3.939   -1.608   -1.206    1.998    4.528    3.679    4.800    4.044    6.364   -2.406   -1.7731966   -2.247   -1.138   -2.280   -0.964    1.624    5.498    4.641    5.008    7.321    1.819   -2.958   -1.7491967   -1.912   -1.954   -1.652   -1.334    0.647    5.430    7.498    6.194    4.936    2.877    1.367   -1.2971968   -1.274   -4.069   -1.409   -1.061   -0.047    5.248    5.583    7.509    4.939    5.758    1.253   -1.6281969   -0.581   -6.667   -1.923   -1.567    1.938    4.906    8.756    7.413    7.106    7.591   -2.218   -3.2441970   -1.749   -5.399   -5.057   -3.316    4.520    7.346    5.639    7.110    7.471    3.652    0.883   -1.8661971   -1.070   -0.489   -3.512    0.322    2.959    3.437    7.806    6.856    7.191    5.876   -0.388    1.7571972   -3.459   -2.497   -1.021   -0.862    0.153    1.730    7.390    6.792    4.133    4.773   -0.067    1.1341973    0.403   -2.702   -0.624   -2.758    2.367    6.360    6.397    8.817    7.054    3.457    0.067   -1.4931974    0.212   -1.858   -1.099    1.026    1.392    5.101    5.473    5.920    3.690   -1.003   -0.431    0.1491975   -0.213    1.463   -3.631   -0.654    1.292    6.099    8.006    9.878    4.682    4.157   -0.161   -0.4261976   -1.031   -1.369   -1.907   -0.927    2.209    8.079    8.759    7.772    4.184    2.800   -0.287   -3.8581977   -3.707   -2.378   -0.856   -2.980    1.284    4.636    7.064    6.147    5.578    4.990   -1.724    1.0701978   -2.656   -4.100   -1.061   -2.071    3.147    3.904    5.772    6.319    6.316    5.861    1.626   -1.1931979   -5.073   -2.982   -3.756   -1.669    0.894    5.600    6.748    5.670    6.013    4.841    0.567   -1.2731980   -3.267   -0.578   -2.998   -0.514    2.384    4.187    5.213    7.374    6.719    1.536   -0.681   -1.3381981   -1.831   -2.994    0.238    0.528    2.361    4.271    6.796    8.987    6.058    0.036    1.293   -3.5901982   -0.944   -0.774   -1.899    0.059    1.853    6.118    8.487    6.547    7.313    3.412    1.053   -1.2321983   -0.099   -4.202   -0.081   -2.039    0.567    6.027   11.769    9.552    6.938    3.982    2.862    1.4611984   -3.659   -2.341   -3.847    0.777    1.420    5.879    8.191    8.576    4.696    4.089    1.451    0.1341985   -5.540   -1.151   -3.010   -0.151    2.287    3.574    7.532    5.774    7.910    6.422   -3.370    0.4281986   -3.328   -7.016   -1.776   -3.334    2.114    7.490    6.613    4.213    5.167    4.366    1.233   -1.2221987   -3.388   -2.517   -2.728    2.481    1.554    3.351    7.126    7.820    5.331    2.740    0.903    1.5001988   -1.373   -3.080   -2.586    0.404    3.153    6.934    5.523    6.892    6.818    4.427    1.487    1.3611989    1.470   -1.368   -0.277   -2.667    5.129    5.209    9.891    7.632    6.549    5.448    2.592    0.5061990   -0.176   -0.143    0.953   -0.701    3.908    3.993    8.616    8.986    4.470    4.943    0.281   -1.4491991   -1.797   -4.582    1.041   -0.722    2.616    2.832    8.656    9.361    7.283    2.842    0.486    2.6121992    1.998    0.353   -0.551   -0.558    5.692    7.233    7.346    6.561    5.362    0.069    1.480    0.2941993   -0.414    0.101   -0.386    1.894    3.391    6.956    5.936    5.586    4.211    1.683    0.692   -1.9691994   -1.928   -2.517   -0.672   -1.041    1.918    6.319    9.477    6.572    4.709    4.704    4.460    0.3111995   -2.050   -0.777   -2.377    1.877    2.031    5.892    9.999   10.348    4.578    6.333    2.151   -2.3681996    0.410   -4.191   -2.208   -0.010   -0.461    6.370    7.820    6.932    5.504    4.047   -0.941   -1.4591997   -0.628   -0.977    1.903    1.084    3.880    4.622    7.317   10.556    7.918    5.392    2.571   -0.5061998   -0.350    1.794    1.041   -1.409    5.198    5.067    6.433    8.361    7.171    2.368   -0.673   -0.1171999   -1.298   -1.750   -0.559    0.997    4.086    4.136    9.122    7.277    7.710    3.764    0.421   -2.1942000   -0.850   -1.333   -0.039   -1.051    3.789    6.373    6.384    7.821    6.793    2.640   -0.346   -0.3992001   -2.043   -1.347   -2.090   -1.769    4.086    4.724    7.641    8.186    4.594    6.072    1.153   -1.2592002    0.674   -0.869    0.420    0.797    2.663    5.338    7.139    8.361    6.178    2.819    1.680   -0.0042003   -2.342   -1.347    2.000    1.710    2.212    7.530    8.431   10.322    6.638    2.171    2.173    1.0892004   -0.922   -2.026   -1.781    0.266    3.209    6.047    6.060    8.484    7.589    2.596    1.119    0.8022005   -0.996   -4.526   -0.213    0.329    1.910    7.648    8.101    7.287    7.708    6.372    0.749   -0.8822006   -0.252   -2.172   -2.449   -0.967    3.088    7.260   11.123    6.884    9.377    6.189    1.831    1.7932007   -0.473    0.096   -0.860    4.778    3.353    6.194    5.828    7.453    5.750    5.938    1.040    0.9422008    0.200    1.554   -2.881   -1.047    4.727    4.754    7.366    7.977    5.148    2.548    0.657    0.0742009   -1.142   -2.031   -0.172    1.920    3.212    5.344    6.394    8.166    7.831    5.058    1.167   -2.7122010   -4.618   -4.324   -1.389    0.957    1.177    6.758    8.019    6.103    5.978    4.298   -1.632   -3.9382011   -1.872    0.106    1.032    4.970    2.781    5.109    5.797    6.393    7.173    5.701    4.617   -1.0022012   -0.653   -0.892    3.509   -1.547    3.761    4.679    6.327    7.824    6.243    2.926    0.554   -1.2092013   -1.334   -3.146   -3.838   -1.710    1.869    5.798   11.291    7.914    7.704    4.923    0.509    0.8182014   -0.884   -1.530    1.948    2.400    3.259    5.779    8.813    5.506    7.959 850mb Pressure Level Air Temperature (C)Latitude Range used:   55.0 to  50.0Longitude Range used: 355.0 to   0.0

So if you take:

December 2013 = 0.818

January 2014 = -0.884

February = -1.530

 

.....and do (0.818-0.884-1.53)/3 you get -0.532 ~= -0.53 as i've used in my original post.

 

And anyway why do you think it was warmer? Warmer from what?

As you can see from the below it(the aforementioned area) was a warmer winter than your climatological average(1980-2010) but overall was below 0 °C.

 

Edited October 24, 2014 by Crocodile23

[karyo]

No update from the 12z sadly!

[Summer Sun]

No update from the 12z sadly!

 

Yep looks like they are ignoring the 12z

[karyo]

Yep looks like they are ignoring the 12z

Coincidentally, they also ignored last Friday's 12z and Sat's 0z.

[Ninman]

Coincidentally, they also ignored last Friday's 12z and Sat's 0z.

 

Well it is Friday night....probably sinking a few glasses of prosecco

[DAVID SNOW]

Coincidentally, they also ignored last Friday's 12z and Sat's 0z.

 

Its the weekend and its Vino time.......again, let the guys have a break  

Well it is Friday night....probably sinking a few glasses of prosecco

You win; you called SNAP! first

[karyo]

 

 

You might be right guys but I need my OPI fix! lol

[bobbydog]

Yep looks like they are ignoring the 12z

Which begs the question- why?

Being 'selective' with the input data wouldnt be very scientific!

Just had a quick flick through the ensembles and the op run seems to be at the bottom end of the scale for air pressure in the areas used for the OPI. Maybe outlier runs are omitted to avoid inconsistent results??

[Crocodile23]

Which begs the question- why?

Being 'selective' with the input data wouldnt be very scientific!

 

Of course, but this is just for the useless in-between results we get every day(which are just for the fun of it).

The actual final OPI value in November 1st, will be based on input of real values of the atmosphere for the 31 days of October, so nothing unscientific is going on at this aspect.

[bobbydog]

That is true but with only a few days left, they may be ignoring any runs they feel are outliers to try to get a better picture of the final value. Just trying to answer my own question really lol.

Anyway, 00z is -2.57

Looking good

[Yarmy]

That is true but with only a few days left, they may be ignoring any runs they feel are outliers to try to get a better picture of the final value. Just trying to answer my own question really lol.

Anyway, 00z is -2.57

Looking good

 

They mentioned in one of the email exchanges that runs with too much variance are discarded. However, it isn't clear what the variance is measured against: the ENS, other models, previous runs?

Edited October 25, 2014 by Yarmy

[bluearmy]

There really should only be 2, perhaps 3 days worth of forecasted runs which can offer any meaningful variance in the end figure. It would be a big surprise if the number was not sub -2 and at this stage trending towards -2.5 is looking favourite.

This piece of the winter jigsaw will be in place for coldies - we now have to wait for riccardo's team to give their prediction based on the data etc.

and then we wait for mother nature to see if she has read the script!

[Penrith Snow]

-2.5 sounds good, so off to B&Q today to buy a large bag of rock salt a new snow shovel a sledge for the grand children and snow chains for my Fiesta.

That confident? Absolutely!

Andy

[radiohead]

It's almost certain now that we'll end up with a final OPI between -2 and -2.5, probably closer to -2.5. In other words, the second most negative OPI year since 1976.

 

[Penrith Snow]

Looking at the historical OPI data every October since 1976 with an OPI below -1.5 has been followed by a cold winter in the UK without exception, not always very cold but below average never the less.

Ok it's only 38 years of data but the evidence is compelling.

It's a shame Ricardo's team haven't gone back another 40 years which would be even better.

Andy

[Cloud 10]

Assuming the final OPI figure is -2.5 or below then the winter of 2009/10 is still the nearest analogue we have for comparison,so i would expect something along the lines of the below charts for their forecast when released.

 

 

 

There is just a chance it isn't that simple however.

[Duane S.]

Well, -2.57 makes a > -0.2 end OPI figure unlikely now I would have thought. This is great news that on the one hand it would appear to show that winter 14/15 is likely to be a good one and the other is it is going to be a great test for the theory. If we get another winter like the last one it's probably going to bust it.

[NorthernRab]

Has the area of analysis been extended to include Poland and the Czech Republic recently? Will this be extend further as we approach the month's end?

[MR EXTREMES]

There really should only be 2, perhaps 3 days worth of forecasted runs which can offer any meaningful variance in the end figure. It would be a big surprise if the number was not sub -2 and at this stage trending towards -2.5 is looking favourite.

This piece of the winter jigsaw will be in place for coldies - we now have to wait for riccardo's team to give their prediction based on the data etc.

and then we wait for mother nature to see if she has read the script!

judging by the model runs today and lastnight we could well end up with a nice OPI index figure at the end of october.

then were have to wait and see what the effects will be interesting stuff very scientific.