In depth Model Discussion

[BrickFielder]

The Polar Stratospheric Vortex has become displaced from the pole and now centers over Scandinavia and Russia.

I think this will tend to lift up the tropopause and the differences in temeperature between 500hpa and 100hpa be high enough for low pressure to form. Around the edges of the vortex high pressure is likely to form in contrast.

Forecast would tend to suggest the high pressure to our west will tend to move eastwards as the stratospheric polar vortex moves towards Siberia. However things become a little tricky after that with either the still quite strong vortex migrating back to the pole (return to wet mild atlantic conditions) or a split of the vortex and early demise for it (a more blocked pattern).

Edited January 26, 2007 by BrickFielder

[BrickFielder]

Stratospheric modelling at the moment will be difficult for the models with the stratospheric vortex (This is different to the low level polar vortex) being displaced. See on this chart how there is warm air over north east Canada.

Then on the next chart the warm air has largely gone. Certainly warm air at this level cools but these charts almost suggest the vortex behaves as the polar vortex with air flows from one area to another. I don't really like this concept and tend to think of the stratospheric vortex as a column of cold air. It can bend and stretch or split in two , or develop tendrills ,but is not quite like a flow of air at a level in the troposphere. Accepting that the model is accurate means the tropospheric polar vortex developing over the north east Canada and US will be short lived and focus of the polar vortex will be solely over china and east asia. At lower levels models are not really suggesting a strong high over Siberia , which I would be expecting under the circumstances.

Ensembles show that a fair amount of uncertainty comes into the models at around the 6th of February.

This I interpret as down to how the jet stream will react to the cold air displaced south over the US. Here cold air will meet warm and we should expect the jet stream to speed up. Will it come blasting across the Atlantic giving us an almost Bartlett high situation with a north south divide in weather or will the jet tend to wrap around the low towards Canada creating a mid Atlantic ridge.

GFS ensemble panels show some of the uncertainty about how the jet will behave.

This tends to be an area where GFS can be a little weak and I would tend to go along with ECMWF view.

At the moment this looks like being a fairly short cold spell and might even miss the UK to the east, but with my unease about the stratospheric modelling proving correct things may change.

Another important factor with respect to the jetstream will be how much low pressure systems undercut the ridge to the west of the US. This seems likely to me given the strength of the Pacific jet and the polar vortex out towards china. My feeling is that this will put energy into the southern arm of the jet across the Atlantic allowing the northern arm to drift more northwards.

Not much point looking beyond 5 days on the forecasts at the moment given the uncertainty , but there are some chances of a return to cold weather.

[Osbourne One-Nil]

Thanks for this Brickfielder (and everyone who contributes). I have to say I find this thread by far the most useful one in the winter area and hope that everyone takes the time to view it, even if they don't comment on it directly. It's good to get some completely unbiased analysis.

[TEITS]

Thanks for this Brickfielder (and everyone who contributes). I have to say I find this thread by far the most useful one in the winter area and hope that everyone takes the time to view it, even if they don't comment on it directly. It's good to get some completely unbiased analysis.

I agree OON a wonderful post from BF and like you say the good thing with BF post is that there is no biased at all towards his analysis which IMO provides a much better evaluation of the situation.

I wish my biased wouldn't affect my posts but I can't help it

[Paul]

Thought I'd posts some thoughts on the differences between todays 12z and 18z among other things in here, will also try to post some stuff about the ensembles too when they come out. Would be nice to get this thread moving for those who prefer a slightly slower pace and more in depth look at things than the general model discussion!

Starting with the 850hpa temps across the country, the 18z appears to have accellerated the journey south of cold air slightly. Although the -10hpa air is a little further north than on the 12z.

12z:

18z:

By Tuesday 12z, the 18z again has the cold air further south, and has lessened the size of the slightly milder pool over the southwest. Notice how the high pressure centre is also further south on the 18z.

12z:

18z:

Fairly small differences but in what may become a marginal situation later in the week, it could possibly have a significant effect.

Obviously a lot of the hype at the moment is of the possible transitional snow and cold v mild battle later in the week, but before we get to that point. It's fair to say that with high pressure close by, there isn't likely to be much in the way of shower activity away from extreme coastal extremeties.

UK Pwat (precipitable water) chart, shows just how dry the atmosphere will be:

Moving toward the end of the week though, and yes I believe there is a possible snow event there to be had, but just sticking with the precipitable water charts, you can see we're not looking at cloudbursts of snow here. Based on this run, the heaviest precip potential is in the south, which is where snow is least likely:

Just to show the level of uncertainty at this stage though, comparing the 12z and 18z for midnight Thurs, shows the difference. Snow risk charts take into account the atmospheric temps, thicknesses etc, so are a good guide without having to go through a dozen charts!

12z:

18z:

So in summary, cold but dry air on its way, so hard frosts, plenty of sunshine and occasional snow showers probably for northern Scotland and exposed coasts. Potential for some back edge snow on the 2 fronts bringing the coldest air south. Then to the 'event' way too much uncertainty at this stage - the 18z is far better than the 12z, recent years have shown though that 8 times out of 10, should there be a battle, the mild tends to win out quicker than forecast at this sort of range. Time will tell of course!!

[West is Best]

Paul, very interesting analysis there. It's extremely good to see this in depth thread up and running again. I really hope the likes of GP and SM will post here, and that we can get into in-depth discussions which even if and when we don't all agree will be interesting and conducted amicably.

I just ran through the 0z in depth and it is actually an upgrade in terms of those temps you posted: deeper uppers, further south at the key times. Here's Monday at 12:00

And Tuesday at 12:00

With regard to Thursday, I'd be interested in thoughts from NW forecast team about 'stalling' frontal systems. We have all seen stalling features where, on the leading edge, there is an easterly, bringing a really cold undercut to the advancing front. My impression at the moment with this front for Thursday is that it may be 'the wrong sort of stalling' if it's simply a case of the upper air temps lifting ahead of the ppn actually arriving on what is essentially a southerly/south-easterly draw. I remember a classic no-show event in the cold winter of 1985-6. I'm not sure of the exact date, but we had been forecast heavy snow in East Anglia and I waited, and waited, and 'nada'. The front had stalled, uppers lifted just a touch, and when it did eventually push through it fell as rain. And that was with a much colder set-up in advance. At the moment this looks to me to be marginal for low levels down south.

One other slightly strange issue for me are the projected surface temps for Wales on Thurs night after the front has passed through. I'm not clearly seeing how this:

Can yield this:

... even with clearing skies. Ah, unless it's presuming lying snow cover?

Looking further ahead, although there is a battle between the Atlantic and the putative Scandi high, there looks at the moment to me to be only one possible winner given the strength and configuration of the jet:

However, in all fairness it would be wrong not to post up this rather intriguing sight in far FI:

No real sign at the moment of the strength going into the southern arm BF do you think?

Edited February 3, 2007 by West is Best

[Stephen W]

Paul, very interesting analysis there. It's extremely good to see this in depth thread up and running again. I really hope the likes of GP and SM will post here, and that we can get into in-depth discussions which even if and when we don't all agree will be interesting and conducted amicably.

I just ran through the 0z in depth and it is actually an upgrade in terms of those temps you posted: deeper uppers, further south at the key times. Here's Monday at 12:00

Dear WIB - thanks for taking the time to post these and present your views in an unbiased way. An interesting week coming up - I prefer this model thread since it gives the opportunity to learn a lot more. As you say, the jet seems the likely winner but with cold not too far away.

Edited February 3, 2007 by CatchMyDrift
Chopped quote size to save space

[Thundery wintry showers]

So in summary, cold but dry air on its way, so hard frosts, plenty of sunshine and occasional snow showers probably for northern Scotland and exposed coasts. Potential for some back edge snow on the 2 fronts bringing the coldest air south. Then to the 'event' way too much uncertainty at this stage - the 18z is far better than the 12z, recent years have shown though that 8 times out of 10, should there be a battle, the mild tends to win out quicker than forecast at this sort of range. Time will tell of course!!

A very good post overall, but I'm going to query the last bit of the post. I agree that when we're talking a reliance on high pressure to the NW, the models tend to overdo it and bring milder air in quicker. However, as was well illustrated in the Februarys of both 2005 and 2006 (and also March 2006), it often seems that when Atlantic systems come up against a block to the East, they quite often overestimate the speed at which mild air moves in. (I get the impression that this was far more true of the 1980s than of now, but I've still detected that tendency).

I still think that the block doesn't stand much of a chance against the jet, but there's an outside bet for the jet to undercut the high, and the high to fill the gap to our N and NE.

The projected 2m temps, I think, are assuming extensive cloud cover and possible low environmental lapse rates- it is often the case in a battleground situation that snow occurs from 850hPa temperatures above -5C (indeed this even happened in February 1996).

[johnholmes]

Yes, good to have constructive in depth discussions. Partly the reason I now have a weather blog. At least I can post my views, get hold of them quickly, and not have them cluttered up with what I call ‘ya boo’ posts.

Good post there Paul, we’ll make a forecaster of you yet

I largely agree with what you have said and shown. Certainly cold but as to ppn once the genuine cold air above the surface has arrived, then a large question mark.

I think the only place likely to see snow before shall we say the expected event is the N Isles and parts of n/ne Scotland. Here 3-5cm could fall as the trough/low, whatever it ends up being swings south and then into the N Sea.

Going on to the comments from WIB and how/why the temps are shown so low.

Its something I tried to get across on my last blog(3). Once the first Occlusion has cleared, possibly, although by no means certain, this far out, with sleet/rain at low level and perhaps snow for a time above about 300ft, skies will clear, and effectively, as the front is an Occlusion, the lower layers will be very similar, albeit modified milder to an extent, to what they were before the front. Thus, I think the models efforts to show 2 m temps s quite realistic.

As to what ‘might’ happen in terms of snow amounts on the first or second occlusion, again like on my 2 blogs, we have to wait until no more than 48 hours before the expected event to be really able to give any idea of where and amounts. My own gut feeling is that the Pennines northwards for all high ground, especially western and central areas are likely to see some snow, at least for a time.

Regards

John

[J10]

What has also happened in recent winters, is that the block has won out in the short term, to avert the battle ground scenario, and mild air slowly seeps in usually from the NW, with no snow.

I seem to recall in either early 02 or 03, Rob McElwee predicted some heavy snow from a battle ground event for the following Tuesday, but the front never made it, and the mild air did indeed creep in from the NW.

This may indeed by a symptom of the polar jet being further north than in the past.

Regarding the scenario for next week, there are still major changes from run to run.

Tuesday 12Hz Situation

http://www.wetterzentrale.de/pics/brack2a.gif

http://www.wetterzentrale.de/pics/brack3.gif

Wednesday 12Hz Situation

http://www.wetterzentrale.de/pics/brack3a.gif

http://www.wetterzentrale.de/pics/brack4.gif

Looking at these a slight improvement in the scenario for Tuesday, but if anything the angle of approach for Wednesday's charts is not as good as previously.

[johnholmes]

hi

excuse these being a triple post.

The first part of the cold spell seems fairly certain with the models being in quite close agreement.

Little snow seems in prospect for the main bulk of the UK. The exception to this, as I see it, mid day Saturday, is for the northern half of Scotland, and especially the N Isles. Here, as the trough/low swings south then into the N Sea during Monday into Tuesday then a spell of snow is likely in the far ne. A scattering of sleet, soft hail and snow showers will occur in much the same areas before this.

Frost for almost all from now on each overnight period, and some quite low values seem probable as time goes on.

As to Wednesday onwards then the models are not in major agreement here on either timing or the situation into Saturday. As time goes on the cold spell appears to be lengthening not shortening. Just how much snow will come out of the attempts by the Atlantic to push in remains to be seen? As I keep stressing its pointless expecting to know much more than 48 hours ahead of the expected event as to what will really happen.

At the moment there is the POSSIBILITY for considerable snowfall from about the Peak District northwards for all high ground, say about 1,000ft in the south of this area, and about 3-500ft over the north of Scotland, as this event develops.

The models are generally suggesting that an air frost is possible even into the Midlands on Friday morning and on Saturday morning for parts of the border counties and further north along with parts of N Ireland.

So lots to watch for over the next few days. Certainly a colder spell than most have seen all this winter in prospect.

Posting this in this thread, the techie one and into blog 4.

I will do another blog sometime this afternoon with charts from the 12z run if I have time.

John

[Snowyowl9]

http://www.wetterzentrale.de/archive/ra/19...00119860105.gif

http://www.wetterzentrale.de/archive/ra/19...00119860108.gif

I just ran through the 0z in depth and it is actually an upgrade in terms of those temps you posted

One other slightly strange issue for me are the projected surface temps for Wales on Thurs night after the front has passed through. I'm not clearly seeing how this:

Can yield this:

... even with clearing skies. Ah, unless it's presuming lying snow cover?

http://www2.wetter3.de/Animation_00_UTC/132_30.gif

http://www2.wetter3.de/Animation_06_UTC/126_30.gif

It shows that the 6z run only has a very brief spell of snow compared with the 00z run with much more but it turns to rain in the end.

The 2 links at the top are from the 1986 battleground snow not as I remember that.

The mild air won eventually in the end,but the 0z run was interesting.

[West is Best]

I think we have to be realistic about next week. Even if this were coming out of a really cold spell the set-up would be extremely marginal to my mind. Had I have looked at this even 20 years ago my inclination would have been 'leading edge snow only for a few hours from Northamptonshire northwards'. Coming as it does with so little cold air intrinsically in the mix (and certainly little to draw on from the east on the front edge) it looks a poor prospect to me. The problem with the slightly slower arrival of the front is that the temps are going to lift ahead of it. I am not, and have never been, a staller fan until or unless the front really hits into the cold air over the UK. Then if it stalls, it's great. But if it's stalling to the west of Britain we can forget it. By the time the ppn is shown to arrive on Thursday, the temps just look too marginal to me, notwithstanding evaporative potential:

I suppose a little front edge snow might fall, but I'm really not convinced, at least until the front hits the Pennines: then it might be different.

Looking ahead, the models seem to my casual eye at this stage to be in reasonable consensus, even if the ensembles are not fully (when are they?). The Atlantic jet pattern has moved a little further north recently. Even at T168 it's not exactly a blocked looking scenario:

[nick sussex]

I'm getting worried I agree with WIB again.

As I've said a few time already the favoured spots are Scotland and then NE although I'm beginning to think NE will be Marginal as well.

Yes that could be a problem with eastern coastal regions in particular with the surface flow ahead of the front being a likely east to se bringing those dew points up. Going back to the possibilities of a staller, its nigh on impossible to forecast this until almost the day, although the ukmo will normally drop the odd hint if they think theres a chance. Theres also the possibility that the front doesnt get any further than the west and weakens as pressure builds from the east. I certainly wouldnt like to be in the ukmo this week, what a difficult forecast to make.

Going back to the longer term an onslaught of the atlantic looks likely with a deep trough running across the atlantic but exactly where the lows track is in a little more doubt because of pressure building to the north again esp over scandi.It will be interesting to see if these positive anomalies continue to grow as this could take the track of the lows further south.

Heres the mean height comparisons.

http://www.meteo.psu.edu/~gadomski/ECMWF_12z/hgtcomp.html

[J10]

Please restrict this thread to detailed comments about the models, other more general comments about the models can be posted in the original thread. Any such general comments will be moved back into the "original" model discussion thread.

Many Thanks

[PersianPaladin]

I at first though that it would be a marginal event as well however having had a look a the data, i have come to the conclusion that with the precipitation forecast to be fairly heavy, evaporative cooling should be expected and east of the Pennines, snow will fall.

Well, it is progged quite heavy over the northern pennines; but I fear it may stall there and only light precipitation getting through eastwards into North Yorkshire, Tyne&Wear, Durham and Aberdeenshire (perhaps further south if things change).

850's east of the pennines look good for the event tho, and dewpoints stay at -1C or below even for the far east.

850's: -

--------

viewimage.pl.png

viewimage2.pl.png

viewimage3.pl.png

viewimage4.pl.png

Dewpoints: -

------------

dewpoint.pl.png

dewpoint2.pl.png

dewpoint3.pl.png

Precip water charts show a typical amount of 12mm of water per hr in the atmosphere within a possible 6hr period of snowfall for the region stated. Looks like a light affair for those away from pennine slopes; although some places on eastern or western lees could see a couple of inches - albeit highly localised.

All this may change of course.

dewpoint4.pl.png

Edited February 3, 2007 by PersianPaladin

[nick sussex]

I think sometimes too much is made of the 850s when in reality its the culmination of this with many other factors like thickness levels and dew points etc that determine if its going to be snow or rain. For example snow can fall with 850s of just below freezing if other factors are positive, we've seen many times snow events with a surface flow ahead of the front from the east or se because as a rule this will produce better other factor conditions.

The one thing I would say is that its crucial for the proposed breakdown next week that the surface flow ahead of the front is no worse than se or more preferably easterly, the latter would more than likely produce the best chances for more areas.

[The PIT]

I'm just thinking perhaps someone should do a check on these prec charts like TWS does on the 12oz Models. I don't have much faith in them so it would be interesting to see if my mistrust is misplaced. It would also be interesting to see how much they modified closer to time.

[Snowyowl9]

For example snow can fall with 850s of just below freezing if other factors are positive.

The one thing I would say is that its crucial for the proposed breakdown next week that the surface flow ahead of the front is no worse than se or more preferably easterly, the latter would more than likely produce the best chances for more areas.

Yes there`s going to be a few factors to take here as the upper air is between -2 and -5c.

The more SE to east wind the better, what happened last mid-March the wind was S/SE and it surprised me how long the cold air lasted last time must of been 1996 but the wind turned more to the east,must of been some very low dew points then the milder air come at night as the wind turned to the S/SW I knew it was over then,but this looks different,a NW is coming in.

http://85.214.49.20/wz/pics/Rmgfs1202.gif

http://85.214.49.20/wz/pics/Rmgfs1232.gif

http://85.214.49.20/wz/pics/Rmgfs1262.gif

The surface temps also at freezing or below so it could even be freezing rain too, but of course it can still change for better or worse.

http://85.214.49.20/wz/pics/Rmgfs1238.gif

It could easily be a bigger snow event than some think.

Edited February 3, 2007 by Snowyowl9

[J10]

I'm just thinking perhaps someone should do a check on these prec charts like TWS does on the 12oz Models. I don't have much faith in them so it would be interesting to see if my mistrust is misplaced. It would also be interesting to see how much they modified closer to time.

As a general rule of thumb, the GFS precipitation forecasts tend to overplay the rainfall amounts, that's why you shouldn't put too much credence in them and as you move into the more realistic timespan, then can then be checked by TV forecasts as an example.

[Paul]

The thing about the precipitation forecasts is that they're much more sensitive to the resolution of the model than any of the other parameters. There only needs to be a splash of rain forecast for one part of a grid square for the whole grid square to show that amount. On the half degree GFS data (upto 180 hours), this is reasonable as the grids are relatively small, as shown by the map below:

Beyond 180 hours the resolution increases to 2.5 degrees and at that point the precip forecasts becoming increasinly inaccurate due to this. The grids are huge...

[The PIT]

Okay since this in depth.

High level blocking over Green land is very important. I presume the rule also applies over Europe and if doesn't why not?

[Evo]

Should we use 850hpa temperatures to determine surface temperatures?

I've noticed several people over the last couple of days refer to the 850hpa temperature and use this as the basis of surface temperature and precipitation type forecasts.

850hpa temperatures are indeed useful to estimate front locations and therefore air masses.

The link below claims that 850 temperatures were historically used by forecasters to estimate the surface temperatures because of the poor surface modeling at the time. 10 to 15 degrees would typically be added to obtain a mean surface temperature, then it would be adjusted for the particular location by using a fudge factor based on experience and the topography of the region. If the surface layer is saturated a value of 7 to 10 degrees is used instead. However, it is implied that this is generally no longer the case.

Models such as the GFS do produce surface temperature predictions, which most seem to treat with scorn. I'm not sure if this is based on experience or suspicion. I'd like to hear Paul's input on this - do you factor any adjustments for your automated site forecasts?

I remember Nick F mentioning that using the hydrostatic equation, it is possible to forecast the mean temperature for a layer of the atmosphere. Perhaps it would make more sense when looking at snow potential to look at the mean temperature of the lower 1500m or so of the atmosphere by using the 1000 and 850 heights?

It's possible to use the hydrostatic equation:

Z2 - Z1 = (R * T * ln(p1/p2)) / g

Where:

p1 and p2 are the lower and upper levels respectively

Z1 and Z2 are the heights of the isobaric surfaces p1,p2 (in metres)

R is the universal gas constant for dry air

g is the gravity constant

Which can be arrange as:

T = ((Z2 - Z1) * g) / (R * ln(p1/p2))

This works fine for 500-1000 thicknesses, for example

p1 = 1000hPa

p2 = 500Pa

Z1 - Z2 = 5280 (528 decametres or 528dam)

R = 287

g = 9.81

T = (5280 * 9.81) / (287 * ln(1000/500)

T = 51796.8 / (198.9332)

T = 260K or -13c

So the mean temperature of the layer between 1000hPa and 500hPa is -13c

However, when using this on the 1000hPa to 850hPa layer, we need to know the 1000hpa to 850hpa thickness, which I can't find published on any normal chart. We are given the 850 geopotential height, though, so we could find the thickness if we had the forecast mslp and the forecast surface temperature:

Thickness = Z1 - (pmsl - 1000) / (g × 1000) * Tsfc

Z1 = 850hPa height

pmsl = Sea level pressure

g = 9.81

Tsfc = Surface temperature

But surely this defeats the object (needing to know the predicted surface temperature)? Can anyone add anything to this?

Ideally, we need to know the 1000hPa height and then this would allow us to calculate the thicknesses.

Anyway, moving on, the reason for wanting to use this method is the research outlined by Martin Rowley in the link below.

He states that statistical analysis has shown the following probabilities for snow to fall, given the 1000-850 thicknesses:

Probability:.............................90%......70%.....50%....30%.....10%

850-1000 hPa(gpm)..............1279.....1287....1293....1297....1302 (unadjusted)

850-1000 hPa(gpm)..............1281.....1290....1293....1298....1303 (adjusted - see below)

Boyden's work in the 60s determined that the thickness values should be adjusted to take into account the height above sea level of a particular area when using the thickness values to forecast snow chances (amongst other things).

(Z-h)/30

Z = 1000hPa geopotential height

h = height of ground above sea level

This factor is applied to the thickness value and then the corrected thickness is used in the table above.

Additionally, Rowley outlines research by Callen and Prescott to determine surface maxima from 1000-850 thickness values as follows:

T = -192.65 + (0.156 * Thickness)

Note the thickness value above is the 1000-850 thickness.

This value is then adjusted using the following table, given the expected cloud cover:

Class 0: Low and medium cloud generally less than half cover. High cloud not overcast. Fog only around dawn, if at all.

Class 1: Roughly 50% cloudiness. If fog occurs, it clears slowly during the morning.

Class 2: Mainly cloudy. If fog occurs, it clears by midday, but slowly.

Class 3: Overcast with rain/snow etc. Persistent Fog.

JAN

Class 0: -4

Class 1: -4

Class 2: -5

Class 3: -5

FEB

Class 0: -3

Class 1: -3

Class 2: -4

Class 3: -5

MAR

Class 0: -1

Class 1: -2

Class 2: -3

Class 3: -4

APR

Class 0: +1

Class 1: 0

Class 2: -1

Class 3: -2

MAY

Class 0: +2

Class 1: +1

Class 2: 0

Class 3: -1

JUN

Class 0: +4

Class 1: +3

Class 2: +1

Class 3: 0

JUL

Class 0: +4

Class 1: +3

Class 2: +1

Class 3: 0

AUG

Class 0: +3

Class 1: +2

Class 2: +1

Class 3: 0

SEP

Class 0: +1

Class 1: 0

Class 2: -1

Class 3: -1

OCT

Class 0: -1

Class 1: -1

Class 2: -2

Class 3: -3

NOV

Class 0: -2

Class 1: -3

Class 2: -4

Class 3: -4

DEC

Class 0: -4

Class 1: -4

Class 2: -5

Class 3: -5

Obviously these figures were derived empirically, rather than theoretically so are not perfect.

So, although I've not really shown the best way to predict surface maxima, I hope I have shown that is far more complicated than just looking at an 850hPa temperature and adding 10c onto it.

Sources:

http://homepage.ntlworld.com/booty.weather...topics.htm#850T

http://homepage.ntlworld.com/booty.weather/tthkfaq.htm

http://www.metoffice.gov.uk/research/nwp/p...381/FRTR381.pdf (Met Office NWP Forecast Techniques Review Board)

[johnholmes]

yes it is highly complex and even professional meteorologists struggle with the maths and sometimes the actual physics behind it.

To try and help most people on here the 'will it snow' idea on the guide I did should help in most situations, as long as you do a correction for your height above sea level.

Ill dig the link out

John

here is that link, hope it helps

http://www.netweather.tv/forum/index.php?showtopic=27989

Edited February 4, 2007 by johnholmes

[West is Best]

Evo that really is a terrific post. It demonstrates how extremely complex is the answer to the 'will it snow in ...' .

I must admit that as an amateur with about 30 years of studying the weather I've tended to combine the upper air temps + 10C and then applied a fudge factor based on what seems to have happened on previous occasions with certain set-ups. Hardly scientific I'm sure you will all agree, although surprisingly not always perhaps that inaccurate. It is great to see proper science being brought into the equation. Can you apply it to the up-coming week I wonder to see how it shapes up?

We are certainly in for an interesting week. Much deeper cold is now projected by the GFS just to the east of the UK:

At the moment this never quite makes it, with the milder Atlantic-driven air pushing across eventually:

This looks a more genuine mild vs cold battle to me, although not really of the ilk of winters past; and I still think there can be only one winner with a jet pattern like this:

It's interesting to see the ECM being so much more progressive at the moment than the GFS, with the fronts straddled right across the UK on Thursday:

This is in contrast to the 0z GFS which has the front parked over the west all day Thursday:

Indeed, the GFS then shows the front breaking up and petering out entirely, before a much milder second wave strikes on Saturday:

By this stage the 850hPa's currently suggest all talk of snow will be banished:

For cold lovers this may be a frustrating time, with doubtless talk if this GFS output were right about 'so nearly' and 'if only'. The 'problem' to my mind remains the jet pattern, and an absence therein of true upstream blocking to allow any Scandi high scenario to build and hold. This looks for all the world to me like a phantom therefore (at this stage). There's certainly little upstream comfort from the ECM output over North America:

For those looking at teleconnection blocking forecasts, the AO seems to be showing mixed signals to me, from neutral to strongly -ve only on the 14 day forecast:

Although the ECM does have a southern movement in the jet (hinted at still occasionally by the GFS) this would still be a westerly and north-westerly dominated regime, with fairly considerable precipitation potential. Much depends on where the putative Scandi high ends up (assuming the answer to that isn't Scandinavia!). The best scenario for cold is probably the least blocked one at this stage because the alternative, hinted at yesterday on the GFS outputs, is for it to sink slowly but inexorably towards Poland, leaving us with the kind of synoptic pattern that so dominated last autumn: a blocked German/Polish high with a strong jet slamming into Europe and leaving the UK in a warm southerly feed. That's only one possible scenario amongst many it seems to me from here ...

Edited February 4, 2007 by West is Best