Tuesday, October 7, 2014

Historically-Accurate Pattern Predictor Suggests Cold Winter Ahead

A winter predictor that has had incredible success in past years, commonly referred to as the October Pattern Index (OPI), is supporting another cold winter ahead.

The above image shows anomalies of the October Pattern Index over the last several days, since the start of October. The OPI, the concept of which was brought about by a group of Italian scientists, says that monitoring of the atmosphere during the month of October can yield great hints at what the coming winter will bring. October is a month well-known for big winter-predictors showing their cards for the coming cold season (i.e. the LRC, and Judah Cohen's Snow Advance Index (SAI)), but it may interest many to know what the OPI may be one the best, if not the best predictor of the upcoming winter season out of the three mentioned above.

The explanation page of the OPI tells of the index's incredible accuracy, around 90%, of being able to predict the December-January-February Arctic Oscillation. In the winter, a negative phase of the Arctic Oscillation (AO) means the polar vortex is weak and is more prone to sending cold outbreaks to the mid-latitudes, while a positive AO indicates a strong polar vortex, hence a warmer winter increases in probability.

Though we aren't to the middle of October yet, early indications are that the OPI is sustained in a negative phase, today plummeting below -3, off the chart as shown above. If the OPI stays in that negative territory for the rest of October, chances are the winter may have a negative Arctic Oscillation. The AO doesn't define the winter in terms of cold overall, but it can give a good hint at the temperature pattern for the coming winter.



Anonymous said...

It just keeps getting better & better...Thank you so much Andrew!

bree (:

P.S.They say if you hate something (winter)with all your being it just makes that something draw to you even more so....sooo
I'm changing my emotion...I love winter, cold, & Oh yes,snow... tons of snow! Bring it!!

Anonymous said...


Thanks again for the insightful data. The upper midwest temps have been below normal for 5 weeks with a few exceptions which fit your forecasts to a tee. I saw a post awhile ago from someone about the old black/red banded caterpillar farmer forecast. I saw one about 10 day's ago and it was totally black, no red at all. Also the snow geese migrated way early this year. Looks like you and nature are on the same page.

Anonymous said...

the latest GFS 18z 10/7/2014 run supports this. freezing temps all the way to northern Georgia. I know it's just a model and relatively far out at that, but looks promising.

DH said...

What's up with the PDO? That was something you mentioned earlier in the year and seemed to play a large role in last years weather.


Shawn said...

That's awesome news Andrew! Thanks! :)

Anonymous said...

I applaud Brie. I live in Mpls and love non winter weather except very hot & humid weather we sometimes get here.

Since 2008 I've changed how I handle winter weather by making really great soup or stew and then shoveling or going for a walk. Started reading again for enjoyment too. It's made a difference for me. BTY old relatives miss the old style winters and said they keep you healthy.

Also what was the gentleman saying about GFS 18z could you explain.

Thanks Mark in Mpls

Anonymous said...

Incredible success? It has been used as a predictor for just one year and it turned out wrong(+1.64 vs +0.18 observed AO), correlating the past doesn't imply good future predictive skills.

BearCub said...

The index today is almost negative 4. Andrew, correct me if I am wrong, but the more negative the number the further West the cold will setup for the Winter? I thought last year's final reading was +1.X.

Anonymous said...

NOAA announced that it might not be a cold winter after all. Based on what you have suggested do you concur.

Andrew said...

DH: The PDO remains positive, with above-normal sea surface temperatures in the northeast Pacific.

Anonymous at 9:54: The prediction was for a positive AO last winter, and it verified. Getting into exact numbers is digging a little too deep into the correlation. As an example, there's nearly a 0.500 correlation between the summer PDO and the following winter PNA, but exact values from the PDO and PNA aren't compared using a ~0.500 correlation- that's silly.

BearCub: This index is still in its 'initial' development phases, so I'm just using it now to predict a positive or negative AO.

Anonymous at 9:35: My Official 2014-2015 Winter Forecast will be out this Saturday at 12:00 PM Central Time with my temperature outlook.

Anonymous said...

Last year prediction was for a +1.64 +- 0.5 AO, 0.18 is outside the uncertain range provided by the authors.