Model output discussion - Closing in on Christmas

[sheikhy]

1 minute ago, Frosty Winter said:

You’ve only gone and jinxed it, it’ll show this now.

Oh my days please delete that chart!!im actually allergic to such filth

[Jason M]

3 minutes ago, Scott Ingham said:

I think enough of a case has been put forward to show it isnt really realistic. Its a 50/50 in terms of low level snowfall. To bluntly state we will get rain when we dont have hi res models or dew points to look at is misleading 

80% of some low level snowfall somewhere I'd have thought (at least 'falling snow). That hides regional differences though.

My thoughts

SE/EA 20%

Central & South and SW England 30%

Midlands / NW / NI maybe 50%

North/ NE England / Scottish Lowlands 80% 

High ground (i.e. Welsh Mountains / Pennines / Highland) - 99%

 

[WillinGlossop]

1 minute ago, Jason M said:

80% of some low level snowfall somewhere I'd have thought (at least 'falling snow). That hides regional differences though.

My thoughts

SE/EA 20%

Central & South and SW England 30%

Midlands / NW / NI maybe 50%

North/ NE England / Scottish Lowlands 80% 

High ground (i.e. Welsh Mountains / Pennines / Highland) - 99%

 

Buxton 60-70%  

[markw2680]

What’s actually quite amazing is that normally one model will be exactly what we are after for cold and the other will be rubbish and we end up with a halfway house which just don’t cut the mustard, in this situation we want a halfway house to hit the jackpot!!

from past experiences we would normally see a shift east of the pattern anyway so I’d expect that as it comes into the more reliable. Plenty of time yet and tbh we are not looking for a major shift east.

enjoy the ride people as these charts are rare as.

[Man With Beard]

45 minutes ago, Frosty Winter said:

ECM 12z ensemble mean has the low further east at t168 compared to the op and there’s a more direct northerly flow at t192!

Really encouraging when the D8 mean chart is chart of the day. 

[Scott Ingham]

10 minutes ago, Allseasons-si said:

OK guys, raise ya bets

Ive head of putting your head on the line..... hahahaha

[Scott Ingham]

10 minutes ago, Met4Cast said:

EC Det is in cluster 2 with 10 other members days 8,9,10. The majority cluster of 12 has a slightly stronger block up towards Greenland with a majority of the other clusters favouring an Atlantic block. 

Latest gloseas run suggests a higher probability of high pressure over the UK out until the 10th January, somewhat fitting in with the EPS signal. Greenland blocking looks transient at best on the EPS to me with a greater focus for an Atlantic ridge and low pressure to the east/over Scandinavia. 

Cold yes, chance of snow across the hills certainly, but again, not seeing anything to suggest monumental blocking besides a few det runs. Long way to go, worth taking into account the ECM over-amplification bias at times too. How many times have we been here this season already only for blocking to become massively watered down?

Realism is often mistaken for pessimism on these forums..

We will certainly see then wont we my case had been made catacol has made his youve made yours. Well see what evolves but there's overwhelming support for Greenland height rises now the ECM has been dragged kicking and screaming to the GFS alongside the UKMO

[Ben Sainsbury]

47 minutes ago, KTtom said:

With all due respect to this concept that we need to be XXX miles west at + 168 hours because it will 'always' shift east when we head towards T-0, bearing in mind these 'super' computers must cost more than ill ever earn to make and run, dont you think the software experts would factor this in it it were true? ..Just a thought..?

That's the whole point in a forecaster, to make a prediction based on model output and our own knowledge (essentially what we're doing here). The forecast models don't factor this in is because the model output is based upon approximations and equations. To factor in adjustments like 'shifting a low pressure eastwards across the UK' would have huge consequences on the weather across the globe.

[Jason M]

7 minutes ago, WillinGlossop said:

Buxton 60-70%  

High ground so pretty much nailed on I'd have thought. If it doesn't snow in Buxton it isn't going to be snowing anywhere!!

[Scott Ingham]

11 minutes ago, Jason M said:

80% of some low level snowfall somewhere I'd have thought (at least 'falling snow). That hides regional differences though.

My thoughts

SE/EA 20%

Central & South and SW England 30%

Midlands / NW / NI maybe 50%

North/ NE England / Scottish Lowlands 80% 

High ground (i.e. Welsh Mountains / Pennines / Highland) - 99%

 

Agreed. I think you have the regionals bang on mate 

[mulzy]

Ex EPS mean very good.  Significant heights to the northwest, north and way out east.  Low Euro heights (centred over eastern Spain).  A few tweaks here and there and we are looking at winter perfection.

[Scott Ingham]

1 minute ago, Ben Sainsbury said:

That's the whole point in a forecaster, to make a prediction based on model output and our own knowledge (essentially what we're doing here). The forecast models don't factor this in is because the model output is based upon approximations and equations. To factor in adjustments like 'shifting a low pressure eastwards across the UK' would have huge consequences on the weather across the globe.

Very sensible post.

If it was taken at face value there would be no need for a met office and the forecast would change every 3 hours. 

The real skill is knowing how a model behaves and factoring this in to create a forecast by T-0

Case in point already when everyone said the ECM was right because its the best model and wasnt showing amplification. When really its just not very good in this area of the globe. You need experience in bias corrections in this instance

[Scott Ingham]

1 minute ago, mulzy said:

Ex EPS mean very good.  Significant heights to the northwest, north and way out east.  Low Euro heights (centred over eastern Spain).  A few tweaks here and there and we are looking at winter perfection.

@Met4Cast they have already improved mate.....

[Mapantz]

Locking this thread now.

Please head over to the new one:

https://www.netweather.tv/forum/topic/94847-model-output-discussion-the-run-up-to-christmas

Edited December 20, 2020 by Mapantz