Model Output Discussion 01/09/17

[Jonan92]

Very good FI charts on the 6z run, the trend shown in the models is encouraging

 

[weirpig]

The Gfs 6z  keeps with the chilly theme as we enter November   then towards the end of its run  brings heights towards Greenland  and also produces an easterly  way out in Fi   interesting output for sure

[Steady Easterly]

9 minutes ago, weirpig said:

The Gfs 6z  keeps with the chilly theme as we enter November   then towards the end of its run  brings heights towards Greenland  and also produces an easterly  way out in Fi   interesting output for sure

Certainly some interesting output as we head into late Autumn/early Winter. One eyebrow raised.

[SLEETY]

On 04/11/2017 at 22:22, LRD said:

Sorry but this sort of post is bonkers to me. There are loads of things, in the past, which have felt unsolveable but with persistence and patience those same problems have been solved by science. So EC46 might be wrong again. So what? One day in the not-too-distant EC46 will start getting it right and everyone will be lauding the overall accuracy of that and other longer range models. Waste of time and money? For trying to do the very good thing of expanding human knowledge and understanding? If our ancestors had taken that attitude we'd still be living in caves. It's funny but longer term climate models predicted the future under global warming and that's proving pretty accurate. I'll go with a continued waste of money (or, to put it another way, I'll go with a continued investment in science for the good and betterment of humankind and, maybe, even the future of the planet itself)

It up to you if you want to believe a computer program will be able to accurately predict the weather seasons in advance.

And you think soon the ec46 will achieve this

Look how much the models struggle after 5 days .Their accuaracy drops off a cliff.

How many times do you see charts in fl verify .Todays gfs06  shows  easterly winds with High pressure over Greenland .In the next run it will change again .It always does.

 

[Man With Beard]

On the EC ensembles this morning between D11 and D15, 11 out of 51 ensembles are normal fare (i.e. High pressure somewhere to the south, low pressure somewhere to the north, flow from the west), but all the others are very un-zonal, with most of them containing a high level block somewhere on our side of the hemisphere and perhaps as many as 20% are in the extreme category with outstanding blocks from Greenland through Iceland to Scandinavia.

The control run is of interest because it is quite similar to the op up until D10 - afterwards, it maintains some Atlantic influence but the massive Arctic High starts to spread towards Greenland and eventually Iceland as we move towards D15, and becomes so influential that lows start spinning back from east to west between Norway and Iceland.

To sum up, the EC ensemble trend of recent days continues, but it isn't unanimous on the high latitude blocking just yet.

To just respond to @SLEETY above, yes of course the accuracy becomes less as you go further out, and I am not going to all bet my money on the trends I currently see. But given that this is a weather enthusiasts forum, isn't it interesting to see how the story of the models unfolds??

Edited November 6, 2017 by Man With Beard

[ACCOUNT DELETED]

Wow some incredibly mild overnight lows here on the coast of South Wales tonight 12-15C (that's 10C or so warmer than last night). Safe to say: A.We'll soon be back out of nature's freezer

B. Off goes the heating as well as the thick duvet again for me tonight lol 

[MP-R]

21 minutes ago, DJ RY said:

Wow some incredibly mild overnight lows here on the coast of South Wales tonight 12-15C (that's 10C or so warmer than last night). Safe to say: A.We'll soon be back out of nature's freezer

B. Off goes the heating as well as the thick duvet again for me tonight lol 

Yes, a very brief period of mild as the active front moves through tonight, taking the mild air east and cold air following on behind. Temperatures should drop sharply behind the front leaving the west in a cold Pm airmass and afternoon temps nearer 7-9C.

[weirpig]

The Gfs  ensembles  show very much a cooling trend as we near the end of the run.  Quite a few of them go for blocking in some shape or form  and a few similiar to the OPP  bring in a easterly.     Going of this run  a very different feel from past Novembers.

[carinthian]

The fine mesh model we use for snow prediction amounts for our location was 27cm for the 12 hour period to 9am this morning. This measurement taken outside the apartment was taken at I hour ago. Just amazing prediction within a forecast confine of 2 km sq. The measurement if you cannot see it is , guess what ..27cm!!

C

[Light Without Heat]

1 hour ago, carinthian said:

The fine mesh model we use for snow prediction amounts for our location was 27cm for the 12 hour period to 9am this morning. This measurement taken outside the apartment was taken at I hour ago. Just amazing prediction within a forecast confine of 2 km sq. The measurement if you cannot see it is , guess what ..27cm!!

C

Brilliant accuracy - Rubbing it in level - High 

[Snowy Hibbo]

2 hours ago, SLEETY said:

It up to you if you want to believe a computer program will be able to accurately predict the weather seasons in advance.

And you think soon the ec46 will achieve this

Look how much the models struggle after 5 days .Their accuaracy drops off a cliff.

How many times do you see charts in fl verify .Todays gfs06  shows  easterly winds with High pressure over Greenland .In the next run it will change again .It always does.

 

I think he is talking about the future of the EC46. Like 5, 10, 20, 30 years from now. Of course the EC46 will improve.

Back in the 60s and 70s, they could only imagine how far we have come with weather prediction. They probably couldn't predict much, beyond what the sky, clouds and wind told them. But now we can accurately forecast out to 7 days. I think it's pretty incredible. Humanity will obviously work out the puzzle of sub-seasonal forecasting. There's already work to improve it. EC46 is getting improved all the time, and NOAA and it's sister agencies just launched SubX, a sub seasonal forecasting system, that is assessing the work of 6 different models. 

 

[Nick F]

Certainly continues to be encouraging signals from the models for colder blocking patterns to develop mid-late month, with no real signs of the Polar Vortex organising itself yet, a sprawling Siberian High and +ve heights over the Pole too. Early days, but the AO and NAO forecasts look similar from 00z/06z GEFS and 00z EPS with a downward / increasingly negative trend for both indices.

 

 

[johnholmes]

Back in the 60s and 70s, they could only imagine how far we have come with weather prediction. They probably couldn't predict much, beyond what the sky, clouds and wind told them. But now we can accurately forecast out to 7 days. I think it's pretty incredible. Humanity will obviously work out the puzzle of sub-seasonal forecasting. There's already work to improve it. EC46 is getting improved all the time, and NOAA and it's sister agencies just launched SubX, a sub seasonal forecasting system, that is assessing the work of 6 different models. 

In the 60's there were no operational weather computers. By the 70's these were beginning to be used by oursleves, the USA and a few other countries. The degree of complexity was huge to even get the original 3 level Met Office model. It is, even to one involved with this change, to read now of the enormous number of levels, data input etc that is now available. Of course the increase in accuracy is slowing down but there is likely to be a very very slow improvement in the 'operational' type, the 1-10 day forecast period I mean. Possibly a faster increase in the 1 month upwards over the next 10 years. Just how far our atmosphere can be solved is giving a lot of debate in meteorological circles. As good a place to read about this is within the met office web site. Not sure just where you will find it.

[phil nw.]

It looks like next weekend will be a re-run of the one just gone with another burst of polar air from the north west.They all look similar- this is the latest run-GFS06z

Keeping things quite seasonal with more frosts and some snow over the higher levels.

[DR(S)NO]

2 hours ago, carinthian said:

The fine mesh model we use for snow prediction amounts for our location was 27cm for the 12 hour period to 9am this morning. This measurement taken outside the apartment was taken at I hour ago. Just amazing prediction within a forecast confine of 2 km sq. The measurement if you cannot see it is , guess what ..27cm!!

C

Jealous much Carinthian

https://goo.gl/images/AH1Prq

[knocker]

26 minutes ago, johnholmes said:

Back in the 60s and 70s, they could only imagine how far we have come with weather prediction. They probably couldn't predict much, beyond what the sky, clouds and wind told them. But now we can accurately forecast out to 7 days. I think it's pretty incredible. Humanity will obviously work out the puzzle of sub-seasonal forecasting. There's already work to improve it. EC46 is getting improved all the time, and NOAA and it's sister agencies just launched SubX, a sub seasonal forecasting system, that is assessing the work of 6 different models. 

In the 60's there were no operational weather computers. By the 70's these were beginning to be used by oursleves, the USA and a few other countries. The degree of complexity was huge to even get the original 3 level Met Office model. It is, even to one involved with this change, to read now of the enormous number of levels, data input etc that is now available. Of course the increase in accuracy is slowing down but there is likely to be a very very slow improvement in the 'operational' type, the 1-10 day forecast period I mean. Possibly a faster increase in the 1 month upwards over the next 10 years. Just how far our atmosphere can be solved is giving a lot of debate in meteorological circles. As good a place to read about this is within the met office web site. Not sure just where you will find it.

John when I first joined the office I worked as gopher at Dunstable where they had just installed the new state of the art Ferranti Mercury computer. It was 1965 when the office went operational with their next generation of computers.

Caption: Met Office Mercury computer equipment, in the Central Forecasting Office. The Met Office, the UK's national weather monitoring and forecasting service, began to use computers in the late 1950s and early 1960s. This Ferranti Mercury was the first advanced computer owned by the Met Office, who purchased it in 1959, shortly before the move in 1961 from Dunstable to new headquarters in Bracknell. It used paper tape and valve electronics, and was capable of many thousands of calculations a second.

http://www.sciencephoto.com/media/435870/view

A link to the obituary of Fed Bushby which is very much pertinent to the evolution of numerical forecasting in the METO

http://www.chilton-computing.org.uk/acl/associates/users/bushby.htm

 

 

Edited November 6, 2017 by knocker

[DR(S)NO]

Carinthian, cannot actually tell you how jealous i am of your picture!

if I could get that just once this Winter... I'd be a happy man

[johnholmes]

3 minutes ago, knocker said:

John when I first joined the office I worked as gopher at Dunstable where they had just installed the new state of the art Ferranti Mercury computer. It was 1965 when the office went operational with their next generation of computers.

Caption: Met Office Mercury computer equipment, in the Central Forecasting Office. The Met Office, the UK's national weather monitoring and forecasting service, began to use computers in the late 1950s and early 1960s. This Ferranti Mercury was the first advanced computer owned by the Met Office, who purchased it in 1959, shortly before the move in 1961 from Dunstable to new headquarters in Bracknell. It used paper tape and valve electronics, and was capable of many thousands of calculations a second.

http://www.sciencephoto.com/media/435870/view

A link to the obituary of Fed Bushby which is very much pertinent to the evolution of numerical forecasting in the METO

http://www.chilton-computing.org.uk/acl/associates/users/bushby.htm

Yes I knew this but it was not used in operational forecasting, sometimes early 70's I 'think' the first operational runs were made. Fred, yep got to know him quite weel, bark worse than bite comes to mind.

sorry folks bit of chit chat from 2 old met men!

delete if you wish team.

Must go and read your links k

 

 

 

[Daniel*]

Newly updated AO forecast has it plummeting like a stone overwhelming signal. 

[mountain shadow]

1 minute ago, Daniel* said:

Newly updated AO forecast has it plummeting like a stone overwhelming signal. 

Yowser. Haven't seen a signal like that in November since 2010!

[phil nw.]

Our Greenland trough could be cut off in week 2 as some ens members indicate a build of pressure towards NW Europe from the the Azores high.Coupled with those +ve hts to the ne then a squeeze could be on.

I think these currently are the 2 options.

We could see a High close by with dry conditions,light winds,chilly with night frosts or the Greenland trough remains intact as it extends se into Europe with further feeds of polar air from the nw.as we are seeing now.

A look at the gefs from the 06z day 10-see the options here. 

Whichever way it goes things continue to look more seasonal with the PV under developed at this stage and a consequential weaker Atlantic jet.

 

 

 

 

Edited November 6, 2017 by phil nw.

[johnholmes]

26 minutes ago, mountain shadow said:

Yowser. Haven't seen a signal like that in November since 2010!

Just remember though that AO and NAO forecasts are just that, they are very prone to being off the mark beyond about 7 days in my experience

[Allseasons-Si]

The gfs at 192hrs  is like water off a ducks back

seriously though,the 12z is more amplified than it's counterpart 06z with heights going up towards Svalbald,will that trough to the NW in future runs disrupt against this block and head SE.

it's thinking about it

and i can see another holy grail of a run in FL

 

Edited November 6, 2017 by Allseasons-si

[mpkio2]

1 hour ago, phil nw. said:

It looks like next weekend will be a re-run of the one just gone with another burst of polar air from the north west.They all look similar- this is the latest run-GFS06z

Keeping things quite seasonal with more frosts and some snow over the higher levels.

Yup.

This actually ties in perfectly with where all the models are now in agreement with each other.

A few days ago, I mentioned that this point on 02/11 that all models (up to t120) were in cross model agreement up to 07/11 into the runs.

This is what they forecast:

12z GFS 02/11 t120:

And this is what tomorrow will look like:

12z GFS 06/11 t24:

Very good verification from all models there!

And this is where all models are in cross model agreement at the moment:

12z GFS 06/11 t120:

12z UKMO 06/11 t120:

0z ECM 06/11 t120:

All show the same theme with strong heights in the Atlantic ridging up to Greenland and trying to join to the small high in the Arctic, Low over Scandi bringing down cool N/NW winds down to the UK, with a Low off Newfoundland/NE America. I do believe this is where we are heading, which also ties into the METO updates of cool/frosty mornings and cool showery weather with snow on high ground.

Where we go from here is all theoretical (IMO) seeing how the models can just about predict/forecast whats gonna happen in a few days, let alone two weeks in advance. It does seem, however, many models are leaning towards a huge Siberian High forming, which could help us to lock into some cold weather (Which some members in some suites (As @phil nw.  points out with the GFS 06z pet members - some going for Scandi high .)

This coupled with the forecast AO plummeting like a stone (As @Daniel* points out with his post) and NAO also forecast (At the moment) going negative too:

Not a bad place to be in for the beginning of November.

However!!!!

Big caution still must be in place, as we've all been here in the past few Novembers that have passed where everything looked good for cold for the UK, only for the models to scrap the idea and go full on zonal, with the Atlantic train bawling through. As some have pointed out, these are only, for now, FORECASTS, and a computer model can easily be wrong, especially with something as complex as our weather (Remembers the fiasco of THAT ECM run from DECEMBER 2012 ).

So, best thing to do is to keep the positivity going, bust stay very cautious and to keep expectations way in check.

As we all know, nothing is written in stone and anything can happen...

~mpkio2~

[Mucka]

(Still in cheeky ramp mode)

 

Winter is coming.