Severe Cold Will Start Late January, Intensify February

I expect the worst of this winter's cold to start up in Late January as more moderate cold, and really get going into February.

I want to first analyze the observed snow cover anomalies over the North Hemisphere in the month of October 2012. There has been work done which suggests a negative correlation (Factor A goes up, Factor B goes down) between October snow cover anomalies and the Arctic Oscillation (AO) in the following winter. So, in times of below normal October snow cover, the corresponding part of winter would then be expected to have warmer temperatures across the US. Likewise, above normal snow cover would hint at a negative AO and thus cooler temperatures.

If we take a look back at observed temperatures for the month of December 2012, we find 'Above Normal' to 'Much Above Normal' temperature anomalies in much of the central and east US. This corresponds very well with below normal snow anomalies in the first half of October. This proves that this winter appears to be lead by this Siberian connection, meaning it is a valid hypothesis to say you could use this correlation to forecast the rest of winter. If we jump ahead to today, we find that the first part of January has been fairly warm, once again matching up with Siberia's below normal snowfall near the halfway mark of October.

However, once the cooldown at the end of January hits, we have then reached the equivalent of just after the halfway point of October, where snow cover was slightly above normal, and thus cooler temperatures would be predicted for the US. Moving on towards the end of October, Siberia encountered anomalously high snow cover, meaning that February could very well lead to well below normal temperatures not experienced in years. This happens to coincide with the potential time period that the polar vortex may come crashing down, something we will discuss later on.

Another piece to the puzzle involves something called the Madden-Julian Oscillation, or MJO. The MJO involves the certain location of anomalous precipitation across the Pacific Ocean. The above forecast graph is cut into eight sections, and each section defines a different placement of this precipitation. Phase 1 is the furthest west, with the highest precip centered in the Indian Ocean, while Phases 7-8 are the furthest west, putting the precip in the far east Pacific or even into the Caribbean. Right now, we are in Phase 6, which is known as one of the phases that causes warmth to overrun the nation during the winter months. In order to get back to a pro-cold solution, we want the MJO to cycle around to Phases 7-2. While that does look to happen, the strength of the MJO appears to be dissipating with time, shown as the forecast models' (colorful lines) tendency to go towards the middle circle, where the MJO is too weak to have any profound affect on the weather. If we saw the MJO sustain its strength and head towards Phases 8-1, we would be in a very good place as far as cold weather potential goes. If we did happen to reach Phases 8-1, it would likely be during the first week or two of February, fitting in nicely with the aforementioned Siberia connection.

Another big thing to watch for temperatures is the sun. Shown above is a graph of observed sunspot numbers in the 30-day sunspot cycle. During the 30 day sunspot cycle, the number of sunspots starts out low, spikes mid-cycle, and drops back down for the end. This process takes about 30 days, hence the 30 day sunspot cycle title. During the height of the 30 day sunspot cycle, temperatures tend to warm as well, due to the activeness of the sun is then directing more heat and energy out to space, including Earth. During the minimum of the 30 day sunspot cycle, temperatures tend to be on the cool side. If the current cycle started around December 24th, the minimum should be coming around January 24th or within a few days of that date, which happens to coincide with the potential collapse of the polar vortex into the US, which I will dive into more below.

0z GFS Hour 252 500mb height forecast

0z ECMWF Hour 216 500mb height anomaly forecast

Both the GFS and ECMWF models - two very respected and accurate forecast models - are showing the polar vortex collapsing into the United States. For those unfamiliar with the term polar vortex, it refers to the low pressure that is stationed over the Arctic, and its varying strength drives the amount of cold air let down to lower latitudes. When it shifts south, frigid air flows with it, and if it comes into the US, you can expect brutally cold air. This collapse is expected to happen around the 22nd - 25th of January. If you look closely at the two models and what timeframe they're forecasted for, you will see significant differences in timing exist between the two models. As small as a 32 hour difference between the two models may be, it is more than enough time for something to happen to possibly prevent this whole thing from occurring. That's why we need the models to match up in order to even see the potential of this happening.

Andrew