Cold Snap, T0-180Hours Only

darkman2 wrote: »

My understanding is that the starting point changes - 00hrs the models start with the data collected from all over the world (bouys, met stations, aircraft measurements etc) at the time the run started (in this case 12pm) and the data is always different to varying degrees to that originally forecast so the outcome changes. The changes get more and more wide as the run progresses and the variables get bigger and bigger. Essentially they are operating off a sort of chaos theory which science can never overcome. The data is processed through supercomputers and complex algorithms to produce a forecast. The forecast is always different to the last one.

Talking as a professional modeller we have the following points.
Firstly most models use historic data as a point of reference. The issue here may be that this type of setup is historically rare & or the time when the setups happened the level of detail taken was poor.
Secondly the most complex models I work with have perhaps 10 'Gross Influence' parameters, those things that alter models beyond others. I guess in weather forecasting there are more than 10.
Thirdly the starting position moves.
Fourthly the 'butterfly effect' (read up on that one)
Fifthly the sheer complexity of the models, which are orders more complex than anything I have created (probably more complex than all the models I have ever created put together).

Even with my own simple models in comparison, forecasting to 80% accuracy in the short term (3 days for weather) is classed as a successful model. A change in maybe 2 of the 10 gross influence variables radically changes the output.

So always take a pinch of salt with the forecasts and take 2 pinches with relatively rare setups like this.

Trotter wrote: »

Thanks for the reply! What element of the high pressure would stop the precip in this case? Is it that the cold air / high pressure travelling west towards us would be high enough to hold any precip away from it?

If pressure is too high then convection could not occur over the Irish sea (convection would produce snow showers in such cold uppers). However, if there is a sufficient temperature gradient (ie difference between the cold air aloft and the warm waters of the sea) then convection becomes much more likely even under relatively high pressure. I remember before the cold spell last November/Dec 2010 people were saying pressure was too high for snow showers but, as it turned out, the upper temperatures were cold enough when passing over the warm Irish sea to produce those great snowfalls

Btw I am no expert, just going on what I have learnt over casual following of the weather over the last few years.

ch750536 wrote: »

Talking as a professional modeller we have the following points.
Firstly most models use historic data as a point of reference. The issue here may be that this type of setup is historically rare & or the time when the setups happened the level of detail taken was poor.
Secondly the most complex models I work with have perhaps 10 'Gross Influence' parameters, those things that alter models beyond others. I guess in weather forecasting there are more than 10.
Thirdly the starting position moves.
Fourthly the 'butterfly effect' (read up on that one)
Fifthly the sheer complexity of the models, which are orders more complex than anything I have created (probably more complex than all the models I have ever created put together).

Even with my own simple models in comparison, forecasting to 80% accuracy in the short term (3 days for weather) is classed as a successful model. A change in maybe 2 of the 10 gross influence variables radically changes the output.

So always take a pinch of salt with the forecasts and take 2 pinches with relatively rare setups like this.

I agree with your entire post except for one point, I asked over on netweather a few years ago about the data input into the models and apparently they dont have any historical data fed into them at all.


Models looking excellent this evening bar the UKMO- You NEED that to come on board before been anywhere near certain of severe cold.

Hopefully the GFS 18Z is good- even if it isnt dont worry, its not very reliable.




Dan

BEASTERLY wrote: »

Its at times like this you'd miss the analysis of Maq, thetonynator or Deep Easterly

no disrespect to those guys, but darkman, su campu, weathercheck, harps, enderman, black briar and not forgetting Dan(Blizzard7) provide just as good analysis as anyone.

nacho libre wrote: »

no disrespect to those guys, but darkman, su campu, weathercheck, harps, enderman, black briar, provide just as good analysis as anyone.

They sure do, but more the merrier and all that!

pistolpetes11 wrote: »

A little video of how the GFS 6z and 12z evolved

https://www.youtube.com/watch?v=te8o8Sx4dGA

Am really impressed with your vid skills. Well done.

but will it be there tommorrow because of the rise in temp and rain predicted will it rain or snow at that height?
thanks

Who?

thanks a lot:D