12z gives a London snow-row total of 104; I think it topped 120 once this season- but, boy, what a contrast this year, forecast-wise even, compared to the previous 4. The better winter continues.

I've certainly seen more snow falling this winter than in the last two or three, but it hasn't resulted in anything more than a dusting at home, and a temporary couple of cm on the top of the Surrey Hills earlier in the month.

It also seems that the variability both between models and between runs has been at an all-time high, or maybe I'm just looking at more different models these days. 

I've added in 5 and 10 day temp anomaly charts. They compare the GFS op to the 1981 - 2010 reanalysis data from NCEP. They are logically correct but I have some reservations. Hopefully they'll be of some use, the direct links are: 

5 day temp anomaly 1981 - 2010

10 day temp anomaly 1981 - 2010

Maybe, but I've felt that too- although a member for <3 years, a watcher for longer. I think maybe the models deal with less frequent weather patterns less well than the frequent. Stands to reason- scientifically, greater sample size correlates with greater confidence levels. The odd & often conflicting signals this season have meant that even the most knowledgeable on TWO seem perplexed at times.

As I suspected, a  large variation in the 2m temperatures on the SW Holland ensembles from day 5, which then widens to make pretty much anything possible from day 8: 

https://www.weerplaza.nl/15daagseverwachting/?type=eps_pluim&element=&r=zuidwest

Cooling then warming trend discernible though.  But a different picture will be painted again in 12 hrs time.

Mean.

Thanks Brian. That is a very useful addition to your charts.

I have just calculated the latest anomalies for the CET thread and I came up with an anomaly of -0.3C for the next 10 days. That is using data from the Meteogroup app which I understand is derived from the ECM model.

I tried using the site below to estimate the anomaly using GFS data - using Luton AP as a proxy for Rothamsted, Leeds/Bradford as the best available proxy for Stonyhurst and Brize Norton as a proxy for Pershore.

http://wxweb.meteostar.com/sample/sample_C.shtml?text=EGVN/

A bit of a rough and ready calculation but this gave me an anomaly of -0.8C. So a bit higher than ECM but still somewhat less than the -2C anomaly shown on your 10 day chart for central England.

This chart shows the anomaly for the next 8 days as in the 0C to -2C range for central England. 

http://wxmaps.org/pix/temp4.html

My calculations have the anomaly for days 1-8 as -1.2C (as opposed to the -0.8C for days 1-10). 

So your anomaly figures seem about right - possibly a little too much on the negative side. But certainly no more than a 1C variance which is easily accounted for by the rather rough and ready nature of my quick calculations.

Cant agree with that at all.  Model fatigue has set in after months of cold chasing for many punters, sadly just as the charts start to get really interesting.

Take the ICON 18z offering - for t120 - fascinating and an upgrade:

http://www.meteociel.fr/modeles/icone_cartes.php?&ech=120&mode=0&map=2&archive=0