Looks like we're still on course for the lowest OPI since 1976, it will be interesting to see the final forecast sometime next month and more importantly if it's correct. I still remain sceptical due to the lack of sufficient data but what is on offer is compelling, should I be waxing my sledge runners now or should I wait a wee longer.

Looks like we're still on course for the lowest OPI since 1976, it will be interesting to see the final forecast sometime next month and more importantly if it's correct. I still remain sceptical due to the lack of sufficient data but what is on offer is compelling, should I be waxing my sledge runners now or should I wait a wee longer.

Won't be lower than 2009, which went under -3. Looks like it will be the second lowest since 1976 tho.

---

would i be correct in assuming the way the OPI is flattening out to a consistent figure of around -2.3 would be due to the ever reducing future data, ie each day means one less future days data to go through? Or does the OPI still use 15 days into the future even on the 30th October?

That would answer the flip flopping earlier in the month, and the stability we have now. Especially when you consider that sub 3 days is considered very accurate in GFS terms as opposed to 15 days in advance, which you would see up until the 15th October.

---

Does anyone know what the OPI was in the final few days of October last year? I'd also be interested to know what it was back in 2010, only a couple of months before the frigid December we had.

---

Hi David,

I did a Blog on the OPI, which contains the full list of OPI years going back to 1976 a few days ago;

http://www.gavsweathervids.com/flash.html

Go to the second Blog entry which is under the comment box.

2010 OPI actually not that low, but of course winter 10/11 ultimately turned mild

---

"Or does the OPI still use 15 days into the future even on the 30th October?"

No.
The final figure is solely the 31 days of real calculations from October's real weather.
(Incidentally, the figures published as a guideline during October use only the next 10 days of GFS output, not 15.)

---

Hi David,

I did a Blog on the OPI, which contains the full list of OPI years going back to 1976 a few days ago;

http://www.gavsweathervids.com/flash.html

Go to the second Blog entry which is under the comment box.

2010 OPI actually not that low, but of course winter 10/11 ultimately turned mild

Indeed - other than that incredible December, January & February were disappointing. With February being especially mild by day. A strange Winter when you'd normally expect Jan & Feb to be colder, the complete opposite happened.

---

Anyone know the significance of the map on the OPI web page and what it means? I have noticed that more countries are shown now than earlier in the month and the colours seems to indicate very low OPI values.

---

Hi David,

I did a Blog on the OPI, which contains the full list of OPI years going back to 1976 a few days ago;

http://www.gavsweathervids.com/flash.html

Go to the second Blog entry which is under the comment box.

2010 OPI actually not that low, but of course winter 10/11 ultimately turned mild

Thanks Gavin.

---

http://www.climatemonitor.it/?p=36712 

The OPI team have posted This link to explain another possible October index to be measured, the IZE (zonal index). The thread offers better views and opinions on this than i could attempt to present..so i won't even try!

Use Google translate to get a reasonable translation of the Italian forum post above.

As I read it, this index points to a more average AO rather than the negative AO from the OPI. However the OPI team not ready to make forecast based on these indices at this time it seems.

Joe

Hi David,

I did a Blog on the OPI, which contains the full list of OPI years going back to 1976 a few days ago;

http://www.gavsweathervids.com/flash.html

Go to the second Blog entry which is under the comment box.

2010 OPI actually not that low, but of course winter 10/11 ultimately turned mild

Indeed - other than that incredible December, January & February were disappointing. With February being especially mild by day. A strange Winter when you'd normally expect Jan & Feb to be colder, the complete opposite happened.

Indeed. January & February 2011 were IMO something of a damp squib after the memorable December of 2010. I think the best overall winter we've had recently if you like cold was 2009/10. Even after the severe spell of late Dec 2009/early Jan 2010 had passed, the weather was never really that mild at any time throughout the rest of that season and I do remember further cold and snow in places at the end of March/start of April 2010. Winter 2012/13 started off fairly mild (and wet) but seemed to get progressively colder and drier after early January, although there wasn't a great deal of snow from what I recall. March 2013 made up for that though!

---

Keep in mind though that a -ve AO is useless if the atlantic doesn't cooperate. Last year the AO was pretty negative for most of the winter but cold air on the eastern seaboard destroyed any possibility of taking advantage of it. 

---

e OPI has correlation of 0.9 with the AO -but how well does the AO relate to Winter CET?

The AO data covers the  the period from 1950 to 2014 a total of 64 winters(DJF).Of these 64 winters 24 (37%)) had a +ve  sign, 40 (63% )  having an AO - sign.The mean of winter CET's from 1950/51 to 2013/14 was 4.3C.

The correlation between the values of the AO and CET for the whole 64 winters is 0.66.For the smaller sample size of the AO + and AO - correlations are lower  figures are AO + ,0.46, AO-,0.48.

If we look at the the below mean/above mean winter  CET's and how they relate to AO- and AO + winters we find;

Of the 40 AO - winters 65% had below average CET's,35% above average CET's.

Of the 24 AO + winters 83% had above average CET's, 17% below average CETs.

Finally a brief look at the stronger AO signal winters >1 and <-1 we find the following years.

AO + winters >1  72/73 CET 4.9,88/89 CET 6.5,89/90 CET 6.2,91/92 CET 4.6,92/93 CET 4.7,99/00 CET 5.4,06/07 6.4. The seven years all have above average CET.

AO - winter <-1 52/53 CET 3.5,55/56 CET 2.9,59/60 CET 4.6,62/63 CET -0.3,64/65 CET 3.3,65/66 CET 4.4,68/69 CET3.2,69/70 CET 3.3,76/77 CET 3.3, 77/78 CET 4.1,78/79 CET 1.6,84/85 CET 2.7,85/86 CET 2.9,95/96 CET 3,00/01 CET 4.5,09/10 CET 2.4,12/13 3.8. The CET in these 17  AO years was below average on 88% of occasions.

Maybe the OPI method will prove a better predicter of the AO than predictions of the NAO.It would be interesting if METO were to release figures showing  how successful their forecasts of winter NAO have been. Be it NAO or AO forecasts  the link with CET is still not very precise with the majority of years having but a weak link. 

 

Keep in mind though that a -ve AO is useless if the atlantic doesn't cooperate. Last year the AO was pretty negative for most of the winter but cold air on the eastern seaboard destroyed any possibility of taking advantage of it. 

 

Not according to this is wasn't:

http://www1.ncdc.noaa.gov/pub/data/cmb/teleconnections/ao-5-pg.gif

 

09/10, 10/11 and 12/13, were mostly negative however (and all cold winters, excluding early 2011, where the AO was correspondingly very positive).

Keep in mind though that a -ve AO is useless if the atlantic doesn't cooperate. Last year the AO was pretty negative for most of the winter but cold air on the eastern seaboard destroyed any possibility of taking advantage of it. 

 

Not according to this is wasn't:

http://www1.ncdc.noaa.gov/pub/data/cmb/teleconnections/ao-5-pg.gif

 

09/10, 10/11 and 12/13, were mostly negative however (and all cold winters, excluding early 2011, where the AO was correspondingly very positive).

This is why I use a tabular Format http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/monthly.ao.index.b50.current.ascii.table

January was significantly negative (-0.97/17%)

February was neutral (0.04/52%)

Admittedly December was strongly positive (1.48/93%)

And ironically despite having the most positive AO, December actually had the lowest anomaly of the three months (February had the highest). 

This is why we should take this OPI with a pinch of salt, even if it is true that this has a 0.9 Correlation with AO, then that does not mean it also has a 0.9 correlation with CET. NAO imo is more important than AO; a negative NAO usually comes with a negative AO anyway. 

 

To this end if we look at the NAO we have:

December was significantly positive (0.95/83%)

January was slightly positive (0.29/61%)

February was strongly positive (1.34/91%)

I've also converted the indicies to gaussian percentiles which might mean more than just the number. 

 

---

While I'm not disputing that -AO does not necessarily equal -NAO, or that the NAO is more important for UK weather, you can clearly see from that table that the coldest months of the last 5 years coincide with highly negative AO values.

The correlation between OPI and CET will certainly not be anwhere near the 0.9 correlation with AO, it seems likely there is probably a strongish correlation. That doesn't mean we *will* have a cold winter of course even with the fairly negative OPI.

While I'm not disputing that -AO does not necessarily equal -NAO, or that the NAO is more important for UK weather, you can clearly see from that table that the coldest months of the last 5 years coincide with highly negative AO values.

The correlation between OPI and CET will certainly not be anwhere near the 0.9 correlation with AO, it seems likely there is probably a strongish correlation. That doesn't mean we *will* have a cold winter of course even with the fairly negative OPI.

It seems to me that a good way of looking at it is you will probably get a cold winter if your NAO is significantly negative, how cold the winter will be then is more about the AO. However a negative AO, positive NAO situ is not that great for cold winter lovers. I think an NAO usually implies a negative AO but I'm not sure the reverse is true.

---

OPI -2.23 with one day to go;

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

---

imo, a poistive nao will be over ridden by a strongly negative AO re nw europe.

Here's how I've come to see the AO/CET correlation after reading the above and considering what I've seen over the years:

 

If you've got a negative AO at the low end of the range, say up to around -1.5, then associated high latitude blocking needs to be positioned in just the right place to deliver cold conditions to the UK. As blocking with this level of negative AO tends to cover less than a third of the hemisphere (though that's just a rough observation based on experience), the blocking will more often than not situate too far west or east to bring much of a drop in temperatures, if any.

If you've got one in the high end of the range, the blocking is more extensive and has a much higher chance of affecting the UK, perhaps very high indeed when you're nearing the magnitude of 2009's minimum values.

 

What do you reckon?

---