Just like to make a couple of comments regarding the roller coaster ride that the last couple of months have been. I come from an IT background of over 35 years, my observations of certainly December and January is that the the computer models are struggling to manage input data, wind speed, temperature, pressure, etc. Data being collected and input into the meteorological super computers is changing in quite radical ways from day to day. This is predominantly because of the conditions in the stratosphere over the polar region, the split vortex and SSW etc.
The computer models, taking current data are having problems providing any coherent model output further than 96 hours
The meltdown on Friday was quite probably a result of disparity of data collected and processed over the previous 2 - 3 days.
I am of the opinion that under the present very mobile conditions, especially over the polar region any attempt to provide any sort of accurate model output beyond the 96 hour time frame is unproductive, and conversely be prepared for unexpected changes to a colder model at quite short time frames.