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

[SMU]

From the very first post of this thread:

 

"The software analyzes the previous consolidated daily charts plus 10 forecast charts as given by the GFS model. For example on the 15th of October, the software calculates the OPI index based on the 15 consolidated daily charts (1-15 October) adding 10 forecast charts as given by the GFS model (16-25 October)."

 

It's always seemed pretty clear to me. It's just the nature of the calculation itself that is unknown (yet).

Hi yarmy you miss the point - we knew the number on the top right of the link was the historic data + 10 days forecast

Whats certainly been a liitle unclear is the graph at the bottom

It should have been the actuals from the end of each dau to be able to calculate the running mean, not as its transpired the forecast values just transferred down...,

S

[Steve C]

make of this as you will

 

 

Not a lot tbh. What are your thoughts then?

[Mucka]

It fluctuates because it's still using 9 days of GFS output in addition to the actual data from the previous days this month. In the coming days it will use less and less GFS output so there will be less deviations until eventually by the end of the month we get the final figure. The daily figures are only really useful for giving an idea of what the final number will be as you get closer to the end of the month. The final number is the only important one.

 

I think even more pertinent tot he fluctuation is the fact that the actual data posted on each day also contains 10 days of forecast data. Therefore it makes the daily values even more open to big swings when that forecast data changes from one day to the next.

 

Personally i am going to sit back and wait for the final figure and stop trying to figure it out. 

Edited October 22, 2014 by Mucka

[Mucka]

Hi yarmy you miss the point - we knew the number on the top right of the link was the historic data + 10 days forecast

Whats certainly been a liitle unclear is the graph at the bottom

It should have been the actuals from the end of each dau to be able to calculate the running mean, not as its transpired the forecast values just transferred down...,

S

 

Indeed and we do not know what the daily total is. Is it a mean of the 00z 06z and 12z so that the 00z is on its own, then the 06z value is mean of 00z/06z, and 12z mean of 00z/06z and 12z?

 

The only thing we can really evaluate is whether the forecast data is more positive or negative than in the last OPI given.

Edited October 22, 2014 by Mucka

[Yarmy]

Hi yarmy you miss the point - we knew the number on the top right of the link was the historic data + 10 days forecast

Whats certainly been a liitle unclear is the graph at the bottom

It should have been the actuals from the end of each dau to be able to calculate the running mean, not as its transpired the forecast values just transferred down...,

S

 

The bar chart or the map of Europe with the geopotential anomalies?

 

If the bar chart, my understanding was that the bar chart always shows the latest overall OPI for the month. So for each day it updates after the 00z, 06z, and 12z. The other days are the values post 12z up to that point in time.

 

The geopotential anomalies are a different matter: are they predicted average anomalies for DJF? If so, the massive negative anomaly for Spain would be good news regardless of the OPI's magnitude.

Edited October 22, 2014 by Yarmy

[SMU]

No Yarmy the bar chart of the daily OPI Index above the map of europe!

 

thanks

S

[shaky]

Ecm has stronger blocking than both the gfs and ukmo lol!!!even manages a minor ridge in the atlantic!!reckon the opi on the Ecm run would be well into the sub 2 category!!

[terrier]

Surely our winter in the uk has much more to do with other factors than the opi. My niggling thoughts are why are the met going for above average winter again. Surely they must take into account the opi index El nino qbo etc. just think this opi is been blown out of all proportion. But time will tell I guess.

[Crocodile23]

The OPI figure now (-1.89) cannot be the running mean total because it fluctuates so wildly, even around the -1 to -3 regions.  Surely some forecast data must have caused it to jump? Weird though, as you say the 12z was a pretty good run!

 

A huge leap from -2.75 to -1.89!!

That is completely dependent, and the reason this has happened, on the GFS 12z versus 06z forecasts of the next 9 days.

So simple!

 

 

 

 

[bluearmy]

Not as I understand it. The daily value is the mean of the consolidated data up to that date + up to 10 forecast days.

Maybe I wasn't clear mucka. From the beginning, my take was that the OPI generated on the link was the up to date OPI for the month which included actual and forecast data. no need to keep a track on what each day showed as the number on the link is the most accurate current best forecast. And that's the thing - until the end of the month it remains a forecast but as the final third of the month ticks by, it should slowly trend to a consistent number.

last year under a strong and organised vortex, the polar field and areas adjoining it which go into the equations were, as is normal under such conditions, well modelled. The OPI varied little post the 20th. this year is different and it may well be that until the 25th/26th we could yet see some variations towards -1 and -3, depending on the nuance of the run.

[karyo]

Surely our winter in the uk has much more to do with other factors than the opi. My niggling thoughts are why are the met going for above average winter again. Surely they must take into account the opi index El nino qbo etc. just think this opi is been blown out of all proportion. But time will tell I guess.

When did they issue their forecast? If it was weeks ago then they wouldn't know whether the OPI would be negative.

[BLAST FROM THE PAST]

make of this as you will

 

 

Beautiful, more of it please....PV split FATBT

BFTP

Edited October 22, 2014 by BLAST FROM THE PAST

[neige57]

When did they issue their forecast? If it was weeks ago then they wouldn't know whether the OPI would be negative.

 

Their forecast comes out in early November

[bobbydog]

I think steve has put into fairly simple terms how the OPI graph works.

The final value can only be calculated accurately from the ACTUAL daily data. As Guido explained, the graph shows the figure derived from both actual AND forecast data from 2 runs of the GFS per day. The OPI value on the graph for any given day is really only the result from the 12z run + actual to that date. As we know, model runs can fluctuate wildly the longer out they go, which would explain the larger differences in the daily graph values. therefore as we near the end of the month, the less influence the forecast has on the final figure and by the 31st, the OPI is calculated by actual data alone.

what we dont see on the graph (or anywhere else for that matter) is the actual data for each day. What we do see is a rough estimate which only becomes truly accurate on the 31st.

in conclusion, the final figure is not a running mean of the graph, it is a running mean of actual data, (which again, we dont see) only available on the 31st of october.

Its not rocket science.

its meteorology. Which is harder...

[karyo]

Their forecast comes out in early November

Thanks, then they may well consider the OPI in their forecast.

[Radiating Dendrite]

I doubt they consider the OPI as it's not peer reviewed etc. Going off their past forecasts they tend to lean heavily on the seasonal models.

[Timmytour]

It's interesting in terms of research but IMO far too many people on this site are getting too hung up on all this, reacting with every nuance. I post on TheWeatherOutlook forum and Ukweatherworld as well as here and it hasn't drawn anywhere near the interest it has on this site. 40 pages already this thread?!

God knows what the mood will be if next year it returns a large +ve value, judging by some of the posts reacting to values that are not as negative etc. Winter over posts galore and it will be at the start of November!

But now we know, when it sits at +2.75 on the afternoon of the 29th October, that the upcoming 0z can change everything, and we can live in hope :-)

 

[SMU]

Right, my last post in this thread for a while as I will now be busy looking into the Winter preliminary forecast - to release on Sunday then the full forecast on 20 Nov.

I will try & stick in 1 more update around 25th/26th

 

My tact has changed regarding the graph on the OPI link.

 

Todays 'score' is '21 actuals & the forecast out to the last day of the 31st' '

 We know the GFS model has been to progressive in removing the blocking over the pole & indeed around 60N as well, so with that in mind you could say the GFS has a positive bias.

 

Thinking statistically to come to a figure of -1.89 (call it 1.9) I would estimate the following.

 

Current rolling Actuals to day 21         ~ -2.2

Projected 10 Day Mean last 10 Days ~  -1.3

This then Nets out @ -1.9

 

With a correction for the GFS 's inability to see blocking I would predict the actuals will be lower at between -1.5 & 1.75 - So for arguments the prediction is last 10 days. -1.6

 

This makes MY OPI forecast for Oct.

 

2.2 X 21 + 1.6 X 10 / 31 = 

 

-2.00              ON THE NOSE.............

 

 

I doubt they consider the OPI as it's not peer reviewed etc. Going off their past forecasts they tend to lean heavily on the seasonal models.

 

Which, the moment they land in full final development are probably 5 years out of date........

 

S

Edited October 22, 2014 by Steve Murr

[feb1991blizzard]

Shall we have a competition, a bit like the CET, lets have a predict the OPI final landing figure competition, entries by 12 o clock tomorrow night, I will start the thread tomorrow morning, if the moderating team don't object please.

[Man With Beard]

Well that's an enormous difference to last night - which had a fairly tight PV - its clear that the forecast models are, in FI, overestimating the PV on a regular basis ATM ... appreciated it isn't the only factor for the OPI but it must play a decent part, so maybe the OPI could yet dip again?

Edited October 22, 2014 by rjbw

[SMU]

So just to add to the post...

 

Here are the correlation results for OPI to AO averaged Winters of Negative.

 

 

Sample size 23 Negative OPIs ( ALL NEG ONLY) landed 19 Negative overall Winters

 

Correlation r = 0.80 rounded up.

 

Result Details & Calculation

X Values
∑ = -24
Mean = -1.043
∑(X - M_x)² = SS_x = 12.657

Y Values
∑ = -18.73
Mean = -0.814
∑(Y - M_y)² = SS_y = 20.893

X and Y Combined
N = 23
∑(X - M_x)(Y - M_y) = 12.919

R Calculation
r = ∑((X - M_y)(Y - M_x)) / √((SS_x)(SS_y))

r = 12.919 / √((12.657)(20.893)) = 0.7944

Meta Numerics (cross-check)
r = 0.7944

 

 

OPI index of months below -1 Sample size 8

 

r = 0.86

 

Result Details & Calculation

X Values
∑ = -15.1
Mean = -1.888
∑(X - M_x)² = SS_x = 3.029

Y Values
∑ = -12.16
Mean = -1.52
∑(Y - M_y)² = SS_y = 8.39

X and Y Combined
N = 8
∑(X - M_x)(Y - M_y) = 4.318

R Calculation
r = ∑((X - M_y)(Y - M_x)) / √((SS_x)(SS_y))

r = 4.318 / √((3.029)(8.39)) = 0.8566

Meta Numerics (cross-check)
r = 0.8566

 

 

OPI Index at -1.5 or below. Sample size 5

 

r = 0.8

 

Result Details & Calculation

X Values
∑ = -12.5
Mean = -2.083
∑(X - M_x)² = SS_x = 2.088

Y Values
∑ = -11.48
Mean = -1.913
∑(Y - M_y)² = SS_y = 4.157

X and Y Combined
N = 6
∑(X - M_x)(Y - M_y) = 2.367

R Calculation
r = ∑((X - M_y)(Y - M_x)) / √((SS_x)(SS_y))

r = 2.367 / √((2.088)(4.157)) = 0.8035

Meta Numerics (cross-check)
r = 0.8035

 

That is some very strong results ( however with the caveat that its a hindcast)

 

In Summary the correlation that the OP used as 0.9 wasn't far off ( although rounded ) using the pearson correlation calculator for negatives.

 

I haven't got time to punt through the positive months.

 

Also It would be good to check the neg months with easterly QBO's!!

S

Edited October 22, 2014 by Steve Murr

[Isolated Frost]

Curiously most of you will be waking up to a -2.99 OPI, based on the 0z. With the missing GFS data issues and what not, I'd say the spread is possibly larger than usual.

 

But then again, I don't have a clue at all, nice to see it dip so negative though.

[bluearmy]

Curiously most of you will be waking up to a -2.99 OPI, based on the 0z. With the missing GFS data issues and what not, I'd say the spread is possibly larger than usual.

 

But then again, I don't have a clue at all, nice to see it dip so negative though.

The vortex angle of 7% is lower than yesterday's angle. We are learning much about how viewing the OPI on a daily basis. Clearly, following a negative figure is quite different to following a positive one. If nothing else, following the index this year has served to show that modelling a hemisphere that has a strong organised vortex is easier than one which is split and struggling. But we all know that anyway!

[Penrith Snow]

Looking at Steve's analysis I would say the OPI is a very good forecasting tool when the value is above 1 or below -1 then it seems almost 100% correct.

Between 1 and -1 I suspect other factors can overwhelm the OPI signal.

If I am correct then last winters mild washout was inevitable as is the coming winter 2014/15 freeze!

We will see.

Andy

[Gael_Force]

This index and its calculations fascinate me with such huge swings at this stage of the month, Time for another side by side comparison to have a look at what is different.

 

Edited October 23, 2014 by Gael_Force