Enough agreement and consistency to forecast. **The start of March will see mid-Atlantic ridging, leading to a cold feed for the UK, with a northerly element to the airflow. Plenty of hill snow and chances of snow to lower levels.**

Rats. Time winter disappeared. Rats.

Interesting, isn't it? Must be some excitement in the coldie ranks for early spring/late winter snow!

On 21/02/2016 09:48, dawlish wrote:
Interesting, isn't it? Must be some excitement in the coldie ranks for early spring/late winter snow!

----------------------------------------------------------------------------------------------------------
Not really. It doesn't look that cold down south. What with the sea
being a degree or two above normal and Europe not spectacularly cold I
won't hold my breath!
Dave

On 21/02/2016 09:48, dawlish wrote:
Interesting, isn't it? Must be some excitement in the coldie ranks for early spring/late winter snow!

----------------------------------------------------------------------------------------------------------
That's possibly towards the end of this predicted colder spell. It
appears to start from about Tuesday. Of course it may still be around
then but currently the coldest weather looks before then.

may well be. I only forecast at 10 days, though.

On 21/02/2016 10:59, dawlish wrote:
may well be. I only forecast at 10 days, though.

-------------------------------------------------------
I know, just saying. The consistency often comes at the start of a
pattern but as this cold snap has appeared quite quickly with no
previous model consistency, this one is forecasting something that may
have already happened to an extent although I guess it helps determine
the length once it starts next week.
I think your exercise has proved the point that 10 day forecasting with
80% consistency is unlikely to be accurate more than a handful of times
a year, although most of us had perceived that.
As you have said it involves a lot of work for you so I suggest you
don't need to do it any more as the point is made and as a forecast, as
this one illustrates, it is rarely going to be that useful.
(You might like to wind it up by saying which models had to be 80%
consistent i.e GFS,JMA, ECM, UKMO etc and which parameters had to be 80%
consistent as well, i.e pressure, temp, 850 hPa, 500 hPa etc, and what
mathematical formula is used to judge 80% consistency of these
parameters in case we do want to have a stab at it ourselves at some point).
Dave

On 20/02/2016 19:28, dawlish wrote:

Enough agreement and consistency to forecast. **The start of March
will see mid-Atlantic ridging, leading to a cold feed for the UK,
with a northerly element to the airflow. Plenty of hill snow and
chances of snow to lower levels.**

Rats. Time winter disappeared. Rats.

From my pov, winter never really *appeared*!



--

Paul Hyett, Cheltenham

Well, if I'd forecast for today, this forecast would have been a good one. If I'd forecast for the 29th Feb, again, it would have been accurate enough, but that depression, which brought Tm air to much of England and Wales yesterday scuppered my forecast for St David's day.

That low wasn't there 10 days previously and it illustrates, again, the difficulties of forecasting at 10 days. One of the 20% of wrong forecasts for me. 😀 The models got the general pattern correct at the time of forecast, but not the detail for the 1st.