Model Output Discussion

[Eugene]

Runs are very interesting tonight, the second GFS FI with extensive northerly blocking i see.

[nick sussex]

I have been critical of the UKMO especially with their obvious obsession with GW and poor LRF but I think we need to separate out the longer range from their shorter term forecats. In terms of their advisory they were quite within their rights to put this out because regardless of what the gfs says theres still a chance. The gfs in terms of precip forecasts and 850's isn't the most reliable and i don't know why it's always the world according to the gfs in here!

Take a look at the new 96hrs fax chart to see why the UKMO put out the warning.

http://www.meteociel.fr/modeles/ukmo2.php?ech=96&carte=2000

[Nick L]

I have been critical of the UKMO especially with their obvious obsession with GW and poor LRF but I think we need to separate out the longer range from their shorter term forecats. In terms of their advisory they were quite within their rights to put this out because regardless of what the gfs says theres still a chance. The gfs in terms of precip forecasts and 850's isn't the most reliable and i don't know why it's always the world according to the gfs in here!

Take a look at the new 96hrs fax chart to see why the UKMO put out the warning.

http://www.meteociel...h=96&carte=2000

I think the main problem is because we have access to masses of GFS data, but nowhere near as much UKMO/ECM data - the GFS is the main source of data that we can quote for short term events.

[john mac]

Well looking at theGFS 18z some decent cold in FI iy shows the atlantic being held at bay by the block, the lows dont even get past iceland until deep FI. But looking at esembles F1 starts at t72, all toplay for IMO

[bluearmy]

the GFS ens certainly catching onto the pattern in Fi now

[Spurry]

I think the main problem is because we have access to masses of GFS data, but nowhere near as much UKMO/ECM data - the GFS is the main source of data that we can quote for short term events.

You share my frustrations ::o

Edited January 16, 2010 by Spurry

[kev238]

Indeed- there is so much data available from GFS that it is easy to take it as gospel. To be honest, I expect somewhere in the warning advisory will get some snow but it will be a narrow band and obviously altitude will be a good thing to have on your side!

[kumquat]

Sorry to intervene - but is the Met Office Fax chart "literally" hand - drawn and if so, are there any videos of the process around?

[Skyhi]

We must back the GFS at this range; if we look at the event between Xmas and New Year; the fax charts would have us believe that the PPN would make it no further North than the M4/South Midlands but the GFS correctly called it making into Northern England, it also called the temp rise ahead of the PPN correctly. Of course it doesn't always get it right, but it's a global model based in America up against the local based METO output who, supposedly, have access to a super computer specifically designed for the UK.

Which is better pound for pound ? Go figure as they say !

Maybe, just maybe, this 'super-computer' has some gremlins floating around it's circuit boards causing all these charts to be all over the place!!!

[J10]

We must back the GFS at this range; if we look at the event between Xmas and New Year; the fax charts would have us believe that the PPN would make it no further North than the M4/South Midlands but the GFS correctly called it making into Northern England, it also called the temp rise ahead of the PPN correctly. Of course it doesn't always get it right, but it's a global model based in America up against the local based METO output who, supposedly, have access to a super computer specifically designed for the UK.

Which is better pound for pound ? Go figure as they say !

But conversely the UKMO had last week (Tuesday) right, with GFS even up to 48hrs before wildly pumping plus zero uppers well over Wales, as it turned out this was nowhere near the case and GFS and NAE got this one right. Of course it doesn't look cold enough for snow here next week on any model.

[Thundery wintry showers]

Personally I would generally trust the UKMO and NAE over the GFS because latest model accuracy stats from NOAA suggest that the UKMO has caught up with the ECMWF over the last 2-3 months and pulled away from GFS. GFS has had one-off days when it's verified better than the other models, but the emphasis is definitely on "one-off".

However I think the Met Office are just warning the public of the possibility of a snow event on Wednesday- because it certainly exists. Personally, though, I think when we get to the time, we'll see something colder than what GFS shows but only enough for snow on high ground and maybe sleety stuff for a time in some lowland areas.

[TEITS]

Very surprised nobody has mentioned the change in the model output this evening wrt to ECM & GFS beyond +168.

As I said this morning to GP I disagree with the ideas of the trough moving into Europe with the PV dropping into Scandi. The models have now backed away from this with the block to our NE remaining in situ. If anything we see the trough back W along with the PV.

Beyond this and it becomes fasinating because we see a weakening of the PV along with some extensive blocking coming out of the Arctic. All of this ties in nicely with the drop in the AO plus the tumbling ensembles.

Those who were convinced of a spell of mild SW,lys may have to think again.

http://www.wetterzentrale.de/pics/Rtavn1681.png

So in summary a possible snow event next Wed as clearly shown on the +96 fax chart. Beyond and a spell of average temps (not mild). However beyond +168 and the outlook is far from certain. Probably best just to wait and see before coming to any conclusions.

I will add though that I strongly disagree with the idea of a mild Feb. The month of Feb has always been the month that blocking is more likely. The combination of this with the extensive snow cover on the continent will mean blocking will be even more dominant than we have already seen.

Edited January 17, 2010 by THE EYE IN THE SKY

[casparjack]

I have been critical of the UKMO especially with their obvious obsession with GW and poor LRF but I think we need to separate out the longer range from their shorter term forecats. In terms of their advisory they were quite within their rights to put this out because regardless of what the gfs says theres still a chance. The gfs in terms of precip forecasts and 850's isn't the most reliable and i don't know why it's always the world according to the gfs in here!

Take a look at the new 96hrs fax chart to see why the UKMO put out the warning.

http://www.meteociel...h=96&carte=2000

I agree, they should stick to the 120 hr time frame and leave the 6-15 days to us and the year 2100 to the politicians!!

[Cheese Rice]

We must back the GFS at this range; if we look at the event between Xmas and New Year; the fax charts would have us believe that the PPN would make it no further North than the M4/South Midlands but the GFS correctly called it making into Northern England, it also called the temp rise ahead of the PPN correctly. Of course it doesn't always get it right, but it's a global model based in America up against the local based METO output who, supposedly, have access to a super computer specifically designed for the UK.

Which is better pound for pound ? Go figure as they say !

The GFS didn't predict snow today (Saturday), but it happened in many areas, the Meto however predicted this when most of us were baffled as to why the Meto would predict snow. So because of this I back the Meto.

[Thundery wintry showers]

Well, I'm genuinely interested now to see how the future progression of the weather goes, because if I'm reading the teleconnections experts' posts right, those teleconnections are suggesting the trough heading to Scandinavia with some HP possibly ridging across from Siberia to Greenland, implying a relatively chilly start to February before it perhaps turns mild with a Euro ridge. But the models aren't showing much sign of bringing the trough towards Scandinavia except way out in FI, and there is no conclusive Atlantic breakthrough modelled until at least the back end of next week.

Although it has turned milder from the west, early next week temperatures still look like being a little below the January average for most parts of the country with a ridge of high pressure moving across from the continent. Yes, the Atlantic looks stronger than the Siberian block at this stage, but nothing like as convincing as the teleconnections suggest.

As I hinted earlier, while the teleconnections have proved pretty reliable so far this season, it only takes one or two variables to override them to cause the forecast to veer wide of the mark- take late last February for example. So while the upcoming spell of weather probably won't have too much of interest, synoptically speaking this is an interesting test for the traditional medium-range forecast models vs the teleconnections.

Edited January 17, 2010 by Thundery wintry showers

[TEITS]

A few more interesting ensembles for you.

Cambs SLP mean between 1015-1020mb.

http://charts.netweather.tv/ensimages/ens.20100116/18/prmslCambridgeshire.png

Oslo SLP mean doesn't drop any lower than 1010mb.

http://charts.netweather.tv/ensimages/ens.20100116/18/prmslOslo.png

Iceland mean shows a sharp rise to 1020mb.

http://charts.netweather.tv/ensimages/ens.20100116/18/prmslReyjavic.png

So what does this all mean.

Well after the possible snow event on Wed a short period of rather benign weather. However as you can see from the ensembles it looks as though blocking will be a feature rather than endless LP systems bringing mild weather. The uncertainity is with regards to the extent and positioning of the blocking.

[Thundery wintry showers]

I've checked out the specifics of the GEFS ensemble charts, and it seems that towards T+240 many of them have the PV splitting, one "split" heading towards Scandinavia and another retreating west of Greenland, allowing elongated northern blocking to develop between Siberia and Greenland. However although the PV heads to Scandinavia not many of the ensembles show particularly low pressure there which is probably why Oslo's mean doesn't drop below 1010mb.

I'm pretty sure that there were some hints at this pattern from the teleconnections. No obvious signs of a Euro High although this was projected to happen more likely after the first week of February.

[grab my graupel]

Please find a new Model thread by following the link below