A new system of analog guidance I composed in recent days, showing success in the limited number of trial runs I have conducted, is forecasting a predominantly warm month of December, followed by a cold January.
|
ESRL |
The image above shows analog guidance's temperature forecast for the month of December. As the image shows, millions across the country would average out over a degree (Celsius) above normal, primarily in the Plains and Midwest. Only the Southwest would see a colder than normal start to winter.
Now, how accurate is this analog guidance? I just got this system up on its feet last night, and these are the first results of the forecast. Some development is still needed, but that will come with time. I've noticed the analogs seem to understand the general pattern (i.e. negative Arctic Oscillation, positive PNA, etc), but still do struggle with less-synoptic features. For now, I'm presenting the data here, allowing it to verify and examine how it does later in January.
|
ESRL |
In January, my analogs see the mass of cold air being stored in Canada on the last image bulldozing southward, bringing below-normal temperature anomalies to much of the nation, especially the northern Plains. The South then looks to see slightly above normal temperatures. Oddly enough, this pattern is surprisingly La Nina-like, even though I made sure to factor in El Nino analogs.
|
ESRL |
The 500mb height forecast anomalies for January show sustained ridging over the Arctic,
forcing what appears to be the brunt of the Polar Vortex south into western North America. This then bleeds east, allowing for those chilly temperature anomalies in the Central US.
I'll end this post with a look at the analog forecast for the stratospheric portion of the polar vortex, since this image only shows the tropospheric version.
|
ESRL |
If you look closely, you'll see the polar vortex has been nearly obliterated...
To summarize:
- A new set of successful analog years is predicting a warm month in December, followed by a cold month in January.
Andrew