Cold Spell Chart, Forecast Analysis & Warnings- 29/11/2010 onwards

Anyone seen the 18Z run yet

I hate the Pale. East wind and the snow only goes a bit inland. NW wind and it'll rain here but snow heavily in the midlands and east. UGH!

Frezing rain wll be the big issue over the next 24 hours. so don't drive unless it's an emergency

Showers will make their way across the country from east to west in the eary hours. The showers will be of rain and sleet in most places, except for the midlands, parts of the north, and the northeast

thetonynator wrote: »

More cold uppers on the way for tuesday . . .

It doesn't make too much difference for that chart but just in case you didn't know, thats actually the GFS control run. It's the same as the operational run but it's at a lower resolution so its not usually as accurate in the medium term because it can miss things that the operational run doesn't.

From my research, the most likely error in model output past day five in the current situation would be under-predicted retrogression. In other words, large-scale features may actually move west faster than depicted (if depicted at all). And looking at the output we have today, that would mean a greater tendency to cold northerly flow patterns re-emerging as they did before the recent very cold week.

I would look for several shots of arctic reinforcement later this month, and perhaps at least one as impressive as this past one was -- timing for these colder shots would be something like 6-7, 12-13, 19-21 and 28-29 Dec. Around those times you could expect waves to be dropping south or southeast and after those dates, colder weather than before those dates. The energy peak on 19-21 could be a very significant winter storm event for Ireland and the U.K. Watch for this to emerge on reliable time frames of models as we get into mid-month. Various sketchy solutions will come and go before then, the models tend to be less accurate than my research indices at day 7-8 on average, and I blend the consensus of reliable models with the research output for my own 1-2 week outlooks.

I've seen some cases where the GFS is totally wrong on large-scale patterns past day 7 and sometimes by day 16 (the end of its run) what is depicted is not even close to a verifying pattern even if you shift it around considerably. My explanation for that is fairly complicated, but the short version is that energy cycles in the atmosphere have typical life spans of 7-10 days, while large-scale features can also change considerably in that time. Therefore trying to blend short-range computer forecasting methods, no matter how fine the grid for initial time may be, is a hopeless way to construct a 10-20 day model (let alone anything much further out).

Let's just put it this way -- one of the more significant weather events of the winter is likely to take place around 20-21 Dec.

Mayo Exile wrote: »

There was an interview with John Hammond (one the weather forecasters on the BBC) on the BBC News channel. He said there was no indication whatsoever of a return to an Atlantic westerly regime and even hinted at current conditions (blocking anti-cyclone setup) lasting into the new year!

Here is the link for that http://www.bbc.co.uk/news/uk-11913530

oooops sorry just posted it now