Model Output Discussion

[feb1991blizzard]

Temperatures really cold for Greenland. Surely thats cold for this time of year even on the ice sheet.

[Weather Boy]

Temperatures really cold for Greenland. Surely thats cold for this time of year even on the ice sheet.

Also very cool for practically all of western Europe, in particular Spain perversely cool for the time of year.

[Backtrack]

Also very cool for practically all of western Europe, in particular Spain perversely cool for the time of year.

I'm starting to wonder if there's an error with the data for Greenland. The GFS has been VERY consistent with those temperatures since the start of June.

I've had family go at this time of year, and the temperatures have been in positive figures, not to mention that the 850 temperatures are into positive figures much of the time, meaning they must be getting lots of freezing rain if that chart is correct.

Can anyone shed some light on this?

Edit: Just looked at the BBC's forecast for Greenland, and the lowest temperature on their forecast for areas around Greenland is -2C and that's at night, so there surely must be an error with that chart?

Edited July 18, 2011 by Vegito

[Gray-Wolf]

As I've already posted I sold my soul for a decent summer school hols so thurs onwards will herald the start of the Summer! (and Hammock time for me whilst the kiddies play in the fields catching Grasshoppers and such like.......)

[SP1986]

I'm starting to wonder if there's an error with the data for Greenland. The GFS has been VERY consistent with those temperatures since the start of June.

I've had family go at this time of year, and the temperatures have been in positive figures, not to mention that the 850 temperatures are into positive figures much of the time, meaning they must be getting lots of freezing rain if that chart is correct.

Can anyone shed some light on this?

Edit: Just looked at the BBC's forecast for Greenland, and the lowest temperature on their forecast for areas around Greenland is -2C and that's at night, so there surely must be an error with that chart?

Well it would be opening a can of worms again but there is a fair amount of loosely documenting evidence suggesting the GFS does not handle temperatures very well.

As for the GFS 6z (and ECM 0z) I feel they are being sucked back into incorrect evolution again, I just cannot see the broad system which is moving east all the while containing a low which is moving west to me, given how slack it is out to the east, and how high pressure is out west - well it just doesn't really make sense to me. I would expect therefore the ECM is probably a little wrong in the medium term at this point, despite it's good marks of recent. I would expect that low pressure to finish far enough east for high pressure to ridge in by the 24th. By no means heat and sunshine, but certainly can't envisage the cold northeasterly winds the GFS/ECM are trying to pull of in this particular setup in the mid term.

[shedhead]

Some very low temps being recorded today, none more so than across the SW.

13:00hrs

Liscombe - 11.0c

Cardinham - 11.7c

Plymouth - 12.0c

Here is an all too familiar looking chart for July 5th 1978, which saw the UK's coldest max for that month of 9.1c, set at Okehampton Devon.

I accept the dangers in pattern matching, but after a dull, cool and fairly wet July it did get drier, sunnier and warmer in Aug/Sept. Clutch, clutch!

As you might also expect, there is a very similar 850mb temp profile across the chart between then and now, with very cool air having

pushed right down into the NW Med and N Iberia.

Edited July 18, 2011 by shedhead

[Thundery wintry showers]

Well it would be opening a can of worms again but there is a fair amount of loosely documenting evidence suggesting the GFS does not handle temperatures very well.

As for the GFS 6z (and ECM 0z) I feel they are being sucked back into incorrect evolution again, I just cannot see the broad system which is moving east all the while containing a low which is moving west to me, given how slack it is out to the east, and how high pressure is out west - well it just doesn't really make sense to me. I would expect therefore the ECM is probably a little wrong in the medium term at this point, despite it's good marks of recent. I would expect that low pressure to finish far enough east for high pressure to ridge in by the 24th. By no means heat and sunshine, but certainly can't envisage the cold northeasterly winds the GFS/ECM are trying to pull of in this particular setup in the mid term.

The NOAA's 8-14 day outputs seem to support your assertions, so maybe they are thinking along similar lines? It's true that we did have high pressure getting far east enough to guarantee most parts dry, sunny and fairly warm weather on many of yesterday evening's runs, particularly the GFS 18Z, so I won't be surprised if the models turn out to have over-reacted to a new trend with the real reality lying somewhere in between (though the "somewhere in between" could still be quite cool & cloudy in the east).

It has been known for depressions to wind themselves up in the North Sea and retrogress westwards, but it is quite unusual in high summer.

[bluearmy]

The NOAA's 8-14 day outputs seem to support your assertions, so maybe they are thinking along similar lines? It's true that we did have high pressure getting far east enough to guarantee most parts dry, sunny and fairly warm weather on many of yesterday evening's runs, particularly the GFS 18Z, so I won't be surprised if the models turn out to have over-reacted to a new trend with the real reality lying somewhere in between (though the "somewhere in between" could still be quite cool & cloudy in the east).

It has been known for depressions to wind themselves up in the North Sea and retrogress westwards, but it is quite unusual in high summer.

ian, i assume you are looking at cpc charts from yesterday. always take care at weekends as they are raw GEFS data with no human adjustment. also, given that they are 6 day mean charts, i wonder if you would have been better looking at the 6/10 day mean output wrt this feature ??

[Thundery wintry showers]

I'm looking at the following charts based on the GFS/ECM 00Z means:

http://www.meteo.psu.edu/~gadomski/ECMWF_0z/hgtcomp.html

[Summer Sun]

I'm looking at the following charts based on the GFS/ECM 00Z means:

http://www.meteo.psu...0z/hgtcomp.html

What does that mean for us TWS more of the same, low pressure setting up home over the uk?

[bluearmy]

I'm looking at the following charts based on the GFS/ECM 00Z means:

http://www.meteo.psu.edu/~gadomski/ECMWF_0z/hgtcomp.html

as far as i am aware, they're not NOAA output Ian. infact, this mornings output is a perfect example of why i'm a tad cautious about these charts. if you check the ecmwf website (i posted the link in a response to mushy yesterday), you'll see the ecm mean 500mb charts for the period T168/T240 are not really reflected by the 'meteo.psu' charts as the 00z ecm op was far too aggressive in dropping in the trough day 10. almost an outlier in 500mb parlance.

[Evening thunder]

I'm starting to wonder if there's an error with the data for Greenland. The GFS has been VERY consistent with those temperatures since the start of June.

I've had family go at this time of year, and the temperatures have been in positive figures, not to mention that the 850 temperatures are into positive figures much of the time, meaning they must be getting lots of freezing rain if that chart is correct.

Can anyone shed some light on this?

Edit: Just looked at the BBC's forecast for Greenland, and the lowest temperature on their forecast for areas around Greenland is -2C and that's at night, so there surely must be an error with that chart?

That would probably be partly because I would have expected your family to visit areas/any towns near the coast, which are actually shown in positive figures on that map, and by the same nature the BBC forecasts are for towns/cities and all of these will be located along the coast not in the middle of a huge few thousand metre high ice sheet.

Also the reason it's shown so cold there is probably because it has a huge ice sheet around it and most importantly is a few thousand metres high in places. This leads on to a question I have about 850hpa temps.

I have asked this before, but no one answered, perhaps because no one really knew, but I always thought the 850hpa temp was the temperature in the atmosphere where the mean air pressure was 850hpa, and I thought this was about 1500m above sea level. If this is correct how can the GFS and other models show the 850hpa temp over Greenland when for a reasonable area of the country this elevation would be buried in the ice sheet? Or does it show about 1500m above the ice sheet instead, in which case I really doubt the temperature at 4500m elevation above Greenland would be above freezing? especially when the 0C isotherm height charts show that only happening at latitudes near Africa. Or if it defaults to surface temps why would the 2m temps be so much colder?

To me I can't see it as anything else than the charts sort of making this up to be 'what it would be if the Greenland ice sheet didn't exist' Surely when the actual model it's self is run it doesn't do this or it would be a huge inaccuracy? Can anyone shed any light on this because I've wondered this for ages?

Edit: also isn't the reason why it looks so cool for the time of year for W. Europe and Spain because it's a chart for 6am?

Edited July 18, 2011 by Stormmad26

[johnholmes]

replying to post 292 from ba

yes I would agree ba, all the available data has to be looked at before making a decision. I rarely do anything until I have seen outputs from all centres in the time scale I'm interested in, so those Ian links to, along with NOAA, ECMWF whatever is available, then spend time seeing how they compare with their last few outputs, not just on one day, although I was perhaps a touch guilty of that last evening. Usually I look at a minimum of 3 days from the outputs mentioned before I then take a look at other outputs in the synoptic scale-do they appear consistent? Then usually a look at further out teleconnections to see what probable effect they may be having.

Its a longish process but it does seem to give a much more stable feel to events in the relatively possible synoptic scale. This I judge to be 15-20 days perhaps 25 in some situations. Beyond that then its into even further out teleconnections and I have never pretended I am any expert at that. An enthusiastic and interested amateur at that range is me!

I have asked this before, but no one answered, perhaps because no one really knew, but I always thought the 850hpa temp was the temperature in the atmosphere where the mean air pressure was 850hpa, and I thought this was about 1500m above sea level. If this is correct how can the GFS and other models show the 850hpa temp over Greenland when for a reasonable area of the country this elevation would be buried in the ice sheet? Or does it show about 1500m above the ice sheet instead, in which case I really doubt the temperature at 4500m elevation above Greenland would be above freezing? especially when the 0C isotherm height charts show that only happening at latitudes near Africa. Or if it defaults to surface temps why would the 2m temps be so much colder?

I would send an e mail to NOAA they are very good at replying usually. I should know the answer but for the life of me at the moment I cannot quote what the height actually is-sorry

Edited July 18, 2011 by johnholmes

[bluearmy]

replying to post 292 from ba

yes I would agree ba, all the available data has to be looked at before making a decision. I rarely do anything until I have seen outputs from all centres in the time scale I'm interested in, so those Ian links to, along with NOAA, ECMWF whatever is available, then spend time seeing how they compare with their last few outputs, not just on one day, although I was perhaps a touch guilty of that last evening. Usually I look at a minimum of 3 days from the outputs mentioned before I then take a look at other outputs in the synoptic scale-do they appear consistent? Then usually a look at further out teleconnections to see what probable effect they may be having.

Its a longish process but it does seem to give a much more stable feel to events in the relatively possible synoptic scale. This I judge to be 15-20 days perhaps 25 in some situations. Beyond that then its into even further out teleconnections and I have never pretended I am any expert at that. An enthusiastic and interested amateur at that range is me!

understood John which is why you tend to post on here rather less often than some. i just wanted to highlight the point as i fear some less expereinced (not you Ian!) may place too much trust in those meteo charts which can sometimes be quite misleading if the op run is an outlier of sorts.

[johnholmes]

understood John which is why you tend to post on here rather less often than some. i just wanted to highlight the point as i fear some less expereinced (not you Ian!) may place too much trust in those meteo charts which can sometimes be quite misleading if the op run is an outlier of sorts.

hi ba

thanks for your reply-a question?

Do you have a full list of links for the ECMWF outputs, I never seem able to get what I want for the 500mb data?

thanks in advance

John

ps

and to highlight your comment about the Op run affecting things. The ECMWF 500mb anomaly output this morning is unusually, for it, fairly different to its last couple of runs I saw. Hence why I never forecast on just one day of output, its the pattern over several days that folk need to watch. If its consistent, and with all 3(ECMWF, GFS, NOAA) then the upper pattern is rarely much different from their progs. If they vary then forget them until they settle back down again?

Edited July 18, 2011 by johnholmes

[feb1991blizzard]

That would probably be partly because I would have expected your family to visit areas/any towns near the coast, which are actually shown in positive figures on that map, and by the same nature the BBC forecasts are for towns/cities and all of these will be located along the coast not in the middle of a huge few thousand metre high ice sheet.

Also the reason it's shown so cold there is probably because it has a huge ice sheet around it and most importantly is a few thousand metres high in places. This leads on to a question I have about 850hpa temps.

I have asked this before, but no one answered, perhaps because no one really knew, but I always thought the 850hpa temp was the temperature in the atmosphere where the mean air pressure was 850hpa, and I thought this was about 1500m above sea level. If this is correct how can the GFS and other models show the 850hpa temp over Greenland when for a reasonable area of the country this elevation would be buried in the ice sheet? Or does it show about 1500m above the ice sheet instead, in which case I really doubt the temperature at 4500m elevation above Greenland would be above freezing? especially when the 0C isotherm height charts show that only happening at latitudes near Africa. Or if it defaults to surface temps why would the 2m temps be so much colder?

To me I can't see it as anything else than the charts sort of making this up to be 'what it would be if the Greenland ice sheet didn't exist' Surely when the actual model it's self is run it doesn't do this or it would be a huge inaccuracy? Can anyone shed any light on this because I've wondered this for ages?

Edit: also isn't the reason why it looks so cool for the time of year for W. Europe and Spain because it's a chart for 6am?

I would hazard a guess although i stand corrected that its the same logic applied to surface pressure charts, i am 600ft above sea level so the pressure level would be lower where i live than suggested on the charts because they are not really surface pressure charts they are sea level pressure charts and they show what the pressure would be if there was no mountain range and we were on a plain. I would guess the same logic applies to the upper 850 temp charts, it is what the temp would be at the 850hpa level if there was no mountain or ice block.

[bluearmy]

hi ba

thanks for your reply-a question?

Do you have a full list of links for the ECMWF outputs, I never seem able to get what I want for the 500mb data?

thanks in advance

John

ps

and to highlight your comment about the Op run affecting things. The ECMWF 500mb anomaly output this morning is unusually, for it, fairly different to its last couple of runs I saw. Hence why I never forecast on just one day of output, its the pattern over several days that folk need to watch. If its consistent, and with all 3(ECMWF, GFS, NOAA) then the upper pattern is rarely much different from their progs. If they vary then forget them until they settle back down again?

see my post 871 from yesterday morning (my copy/paste not working)

Edited July 18, 2011 by bluearmy

[johnholmes]

see my post 871 from yesterday morning (my copy/paste not working)

ok ta I'll go take a look

[knocker]

I would hazard a guess although i stand corrected that its the same logic applied to surface pressure charts, i am 600ft above sea level so the pressure level would be lower where i live than suggested on the charts because they are not really surface pressure charts they are sea level pressure charts and they show what the pressure would be if there was no mountain range and we were on a plain. I would guess the same logic applies to the upper 850 temp charts, it is what the temp would be at the 850hpa level if there was no mountain or ice block.

I would suspect that is probably correct. Interesting looking at the Bloemfontein midday ascent which is nearly above the 850mb level. Bloemfontein is at 1358m. I got confused recently reading a paper when some of the diagrams had surface pressures of around 830mb until the penny dropped that the station level pressures at Jan Smuts were being used.

[Summer Sun]

GFS 12Z is now coming out, looking cool at present for Saturday in the North anyway,

http://www.wetterzen...s/Rtavn1321.png - http://www.wetterzentrale.de/pics/Rtavn12617.png A ground frost maybe on Saturday night / Sunday Early hours - http://www.wetterzen.../Rtavn13217.png - http://www.wetterzen.../Rtavn13817.png

Temperatures look like recovering on Sunday

http://www.wetterzentrale.de/pics/Rtavn1501.png - http://www.wetterzentrale.de/pics/Rtavn15017.png

So a cool and wet day to come for those in the east on Saturday, never mind summer it will be more like mid October.By Monday thing's do start to settle down for a few day's.

Edited July 18, 2011 by Gavin D

[knocker]

I'm starting to wonder if there's an error with the data for Greenland. The GFS has been VERY consistent with those temperatures since the start of June.

I've had family go at this time of year, and the temperatures have been in positive figures, not to mention that the 850 temperatures are into positive figures much of the time, meaning they must be getting lots of freezing rain if that chart is correct.

Can anyone shed some light on this?

Edit: Just looked at the BBC's forecast for Greenland, and the lowest temperature on their forecast for areas around Greenland is -2C and that's at night, so there surely must be an error with that chart?

Probably quite irrelevant but the 850mb temps for midday.

[Backtrack]

That would probably be partly because I would have expected your family to visit areas/any towns near the coast, which are actually shown in positive figures on that map, and by the same nature the BBC forecasts are for towns/cities and all of these will be located along the coast not in the middle of a huge few thousand metre high ice sheet.

Also the reason it's shown so cold there is probably because it has a huge ice sheet around it and most importantly is a few thousand metres high in places. This leads on to a question I have about 850hpa temps.

I have asked this before, but no one answered, perhaps because no one really knew, but I always thought the 850hpa temp was the temperature in the atmosphere where the mean air pressure was 850hpa, and I thought this was about 1500m above sea level. If this is correct how can the GFS and other models show the 850hpa temp over Greenland when for a reasonable area of the country this elevation would be buried in the ice sheet? Or does it show about 1500m above the ice sheet instead, in which case I really doubt the temperature at 4500m elevation above Greenland would be above freezing? especially when the 0C isotherm height charts show that only happening at latitudes near Africa. Or if it defaults to surface temps why would the 2m temps be so much colder?

To me I can't see it as anything else than the charts sort of making this up to be 'what it would be if the Greenland ice sheet didn't exist' Surely when the actual model it's self is run it doesn't do this or it would be a huge inaccuracy? Can anyone shed any light on this because I've wondered this for ages?

Edit: also isn't the reason why it looks so cool for the time of year for W. Europe and Spain because it's a chart for 6am?

Hey Stormmad, great post.

I don't think they visited the coast, they went to Lapland, unsure how close this is to the coast though. I can't remember if it was Lapland either, but it's where santa is.

Thing is, those charts are showing the air temperature, and I can't see the air temperature being -25C in the height of summer, and where as slightly further North, sea ice is melting. Surely if it can reach -25C in Greenland in the height of summer, then it can only be even colder further North?

Regarding the 850's I thought that they were taken at a set height above sea level, meaning that in places the 850's would be as cold as the air temperature meaning that the 850's would be as low as -25C?

I honestly don't know if there is an error with the chart, or I am just interpreting the information very wrongly. Would really like John Holmes to shed some light on this, as I believe he is the best guy on the forums to deal with this matter.

[Summer Sun]

GFS or Giving False Summer, is at it again in FI,

http://www.wetterzen...s/Rtavn2161.png

http://www.wetterzen...s/Rtavn2401.png

Before pressure slowly fall as July comes to a close

http://www.wetterzen...s/Rtavn2881.png

Edited July 18, 2011 by Gavin D

[William of Walworth]

This coming weekend especially : 12z models seem to me to suggest something of an W / E divide, or more to the point a SW / NE divide, largely. For Saturday in particular.

I know temperatures (at this stage) don't look very likely to be warm for most places til after the coming weekend, but in the W / SW at least, there seem to be a much higher chance of drier conditions this w/e as the HP just starts to influence those regions by Sat and Sun.

Also, in the coming days' putput and still talking about the forthcoming w/e, we stil have time for small yet possibly? significant tweaks in synoptic positioning -- changes to the exact timing of when the Low starts to move away significantly, could alter conditions on the ground a fair bit, in some places.

GFS or Giving False Summer, is at it again in FI,

http://www.wetterzen...s/Rtavn2161.png

http://www.wetterzen...s/Rtavn2401.png

Before pressure slowly fall as August comes to a close

http://www.wetterzen...s/Rtavn2881.png

Just seen this one, surely you mean as JULY comes to an end

Don't give more ammunition to the August/summer writers-offers than they'll find for themselves

[The Northern Ramper]

Looks very consistant the gfs, it has counted down a fairly settled looking weekend from fi into the near reliable.