Model madness continues but there's an underlying increase in support for very cold air to make its way in from the northeast within the 3rd week of the month, likely as part of a cyclonic setup (i.e. unstable i.e. high snow potential).
The GFS 12z was one of the colder outcomes, giving me a CET estimate to 26th in the mid-1s °C.
Yet there have been a fair few much milder outcomes appearing too, which while a dwindling cluster in the ensembles, can't be ignored. Those lift my CET estimate right up into the 3s or even 4s °C!
Funny really - coldest opening 10 days in the CET series in 128 years, while down here it's been nothing of much note (overall) and about half the days reached close to 5°C which is what I'd call 'very chilly' rather than cold.
No snow here either - but it did deliver a rare subzero maximum and one of the only ones I can recall with no snow cover. So it's had its moments!