Model Output Discussion

[bluearmy]

There is a good reason for that bluearmy, its because these things need to undergo extensive testing and if they find anomolous results they have to go through more testing/development. Whats the point in naming a date when they cant guarantee it will be ready, if you do that then you end up having people complain that its either not working or not on time. We all know models are complex things and i'd have thought developing them is a very complex task.

dont disagree - ecm and gfs have both made upgrades in the past two months and their websites kept us fully informed as to the planned live dates and whether there were any issues with these dates. maybe uk met have done this aswell - i doubt it though i stand to be corrected if i'm wrong.

[johnholmes]

Met Office

Provisional Model Upgrade

Timetable 2010

• January 2010

• Global model to 25km
• Upgrade Global ensemble to N216 (~60km) / L70
• Upgrade regional ensemble to 18km/L70
• Replace the stochastic kinetic energy backscatter (SKEB1) and
  stochastic convective vorticity (SCV) schemes with the SKEB2
  scheme (MOGREPS-G).
• Revised screen T diagnostic
• New soil ancillaries and van Genuchten soil hydraulics

• Spring 2010

• Physics upgrade including new PC2 cloud scheme

May I ask where you got that from and is it for general realease?

many thanks

[cooling climate]

Met Office update as expected after discussing this with my contact yesterday. We seem to have missed the opportunity to get hold of some cold uppers this weekend. The bulk of the cold is slipping into France as illustrated in the latest model runs. It looks likely that there will be milder air mixed up with next weeks Northerly - snow only for higher elevations. We are definately going to have a cold spell - however it is going to be a fairly mundane affair for most with a wintry mix of precipitation the best we can hope for together with some frost. By the way it's a lovely day out here today clear blue skys and an increasingley warmer sun

I think you would be very wrong to take this as gosbel. Watch the models(MJO, stratosphere etc etc)

they will give us and the Met the answers.

The Met are only responding to what they see at the present time and imo they are wrong.

The GFS 12z showing some activity for east Anglia and the southeast tonight and tomorrow

I see.

[Cecil Hogmanay Melchett]

I think you would be very wrong to take this as gosbel. Watch the models(MJO, stratosphere etc etc)

they will give us and the Met the answers.

The Met are only responding to what they see at the present time and imo they are wrong.

The GFS 12z showing some activity for east Anglia and the southeast tonight and tomorrow

I see.

yes but check out the temps,id say rain/sleety mix going off GFS?

[full_frontal_occlusion]

There is a good reason for that bluearmy, its because these things need to undergo extensive testing and if they find anomolous results they have to go through more testing/development. Whats the point in naming a date when they cant guarantee it will be ready, if you do that then you end up having people complain that its either not working or not on time. We all know models are complex things and i'd have thought developing them is a very complex task.

Not quite right Pinball. The models are modular with very specific parameters passed between the physics packages (ie. the mathematical descriptors for the gas laws, etc), the main partial differential equation solutions (conservation of energy, momentum etc), the main sequencer (resolution, stochastic iteration etc) and then finally the GUI (Graphical User Interface, houskeeping, diagnostics etc).

Point is, all of these items can be developed in parallel and integrated into the final model as and when they reach a level of test compliance independent of other modules.

Software development goes through very definitive stages from Analysis, Specification and Definition through to Coding, Testing, Integration and Qualificaton.

Each of these stages can be estimated with increaing accuracy as each stage of development is completed. The final delivery date can be quite accurately predicted once the definition phase is completed.

If the development lifecycle goes through successive iterations of testing and modification as you suggest, then this is a sign that not enough risk analysis and/or risk reduction was included in the early phases.

Software development at thus scale is incredibly expensive and mistakes made at the start just get compounded eponentially if they are not found early on.

Giving 'provisional' dates for release is a way of managing expectations and does not reflect the fact that the Met does not know when - any other way and serious questions would be asked about how much the whole thing was going to cost with quite a few heads rolling as a result!

ffO.

[mountain shadow]

Not very different to the 6z up to T+120, angle of attack maybe a touch less favourable.

http://charts.netweather.tv/gfsimages/gfs.20100209/12/120/h500slp.png

[kold weather]

The 12z shows the utter western limit of the upper high over Greenland if we are going to get any decent northerly airflow, shows further west/weaker then the results would probably be similar to what the ECM shows...

[mountain shadow]

12z looking very good at T+144, a little further west but not to much.

[karyo]

12z looking very good at T+144, a little further west but not to much.

I'd say it is a significant shift to the west and not a good trend! We need the ECM to start going with the gfs idea not the other way around!

Karyo

[mountain shadow]

I'd say it is a significant shift to the west and not a good trend! We need the ECM to start going with the gfs idea not the other way around!

Karyo

I make it about 150 miles further North West at T+156, not exactly significant, at least not for here.

[kold weather]

Thats a big difference actually MS, esp for the south. As I said the 12z is the utter west limit we can have for most of the country bar maybe Scotland as I think the 12z from 144-180hrs shows perfectly, we are right in the upper low, shift it another 100 miles west and we go into a slowly warming flow IMO...

[mountain shadow]

I'm assuming we want the scandi trough to be deeper to allow it to kick south east? If we get flabby lows then pattern moves North West. General pattern remains good though.

Certainly is either going to very wet or very white somewhere. Still the chance of some significant snowfall especially with a little height.

Edited February 9, 2010 by mountain shadow

[sandstorm2]

Is there not an awful lot of cold air stagnating in the low settled over the UK, if there is there will be a lot of snow.

SS2

[mountain shadow]

Is there not an awful lot of cold air stagnating in the low settled over the UK, if there is there will be a lot of snow.

SS2

The problem will be slighly less cold in the mix moderating temps in the boundry layer. Still looking good for Scotland and Irelans and many parts of England and Wales in my view, apart from Ibrox where it will rain.

[Bobby]

GFS looking very snowy, I think a potential problem with this Northerly could be the lack of real cold about. Most of it is right up at the pole and we don't seem to tap into it. So probably be marginal but as long as keep the uppers below -5 it should be OK, especially if it stays a bit further East. Could be very widespread heavy snowfall though.

[Thunder Snow]

If we go by the 12z run, there could be alot of snow around next Monday sitting across the same areas for a long period.

And some widespread snow by the middle part of the week.

[sandstorm2]

The problem will be slighly less cold in the mix moderating temps in the boundry layer. Still looking good for Scotland and Irelans and many parts of England and Wales in my view, apart from Ibrox where it will rain.

Now now steady on there lol, yes you are probably right about the mixing, lets get optimistic.

SS2

[karyo]

Thats a big difference actually MS, esp for the south. As I said the 12z is the utter west limit we can have for most of the country bar maybe Scotland as I think the 12z from 144-180hrs shows perfectly, we are right in the upper low, shift it another 100 miles west and we go into a slowly warming flow IMO...

Indeed! This run shows the coldest air flooding south over the Atlantic rather than over us! The only saving grace is the low on the east coast of Canada which is angled in a way that stops the high pressure block moving further west.

The UKMO is slow in updating today.

Karyo

[Weathizard]

If we go by the 12z run, there could be alot of snow around next Monday sitting across the same areas for a long period.

And some widespread snow by the middle part of the wee.

I'm presuming this snow won't reach the SE?

The models look a lot better for your part of the world but not so sure about other regions..

[Iceberg]

Overall a very good GFS run, it keeps the snow and -5 850's across the country until the 23/24th of Feb and in some places in the east keeps the snow going until T384. No need to mention it's a snowfest again.

Meto follows it reasonably well too.

[cooling climate]

Overall a very good GFS run, it keeps the snow and -5 850's across the country until the 23/24th of Feb and in some places in the east keeps the snow going until T384. No need to mention it's a snowfest again.

Meto follows it reasonably well too.

Meto is not out on meteociel yet, can I ask where you managed to view it.

[Thunder Snow]

Northern and Western areas will fair better for a change lol, if there is any snow it will be lighter and patchy. However there could be some more widespread snow towards midweek.

However this is only one run.;

[Michael Fishermans Friend]

Can't see snow for southern parts of the UK from that, lower air will be modified to much by the time it reaches these parts even though the uppers are reasonable. For Scotland, N Ireland and Northern England could have a good spell of wintry conditions, but feeling a little short changed at the moment over the supposed Wintry North Easterly which is now getting more and more 'watered down', so much that nobody is even mentioning it now.

[PB22]

Can't see snow for southern parts of the UK from that, lower air will be modified to much by the time it reaches these parts even though the uppers are reasonable. For Scotland, N Ireland and Northern England could have a good spell of wintry conditions, but feeling a little short changed at the moment over the supposed Wintry North Easterly which is now getting more and more 'watered down', so much that nobody is even mentioning it now.

I'd agree Timmy. The GFS outlook ties in very well with today's Met Office 6 - 15 day outlook. A very good outlook for northern hills in particular in terms of snow potential. Lower down and towards the south, I wouldn't expect much other than cold rain if the 12z run came off.

What north-easterly?!

[Thundery wintry showers]

Not entirely sure about that, judging by the GFS outlook it seems that in the northerly towards FI, overnight minima are mostly below zero everywhere so the outlook would mostly have sunshine and snow showers but with a thaw in the daytime sun across southern England. Maybe any frontal events from shortwaves would have rain in the south and snow in the north, but we know how the GFS overdoes shortwaves beyond about 48 hours out.

The upcoming NE'ly is indeed watered down, the issue is the proximity of high pressure rather than the lack of cold air, and while 850hPa temps are favourable, thicknesses are less so. The GFS has been proved right in frequently positioning the HP further south and the UKMO/ECM have had to play catch-up, in stark contrast to the 6th-10th January when it was the opposite way around.

Nonetheless there should be some snow showers about in East Anglia and the SE over the next three days, though probably with only limited accumulations and so always vulnerable to the daytime sun.