Model Output Discussion 29th December - Into mid-Winter.

[nick sussex]

We're in the unusual situation here of wanting the flatter earlier solution of the ECM rather than the more amplified shortwave of the GFS. The ECM runs shortwave energy east more quickly and then amplifies the flow after this . The GFS has initial the more amplified shortwave but then becomes flatter.

I should say something re the upstream pattern whilst NCEP still haven't made a call either way New Mexico state forecast has suggested the ECM is showing better continuity than the GFS with shortwave energy in its region and was more accurate last week. Aswell as this in other areas several state forecasts mentioned that some initial outlier solutions the ECM ops may have had over the USA are being moved to by the other outputs.

The upstream trough (which the GFS operational run and GFS/EC
ensemble means from yesterday indicated would follow a kind of flat
trajectory across the Great Basin) have now swung in direction of
the EC operational run which was the outlier in developing a
stronger trough down across CA by Wednesday night/Thursday.

As you can see I'm desperately trying to fill some time until tonights outputs. We do have the SREF coming out shortly for the USA and given the timeframes that could give us a pointer to which model has the correct pattern.

[northwestsnow]

Well the meto update suggests a breakdown on sunday which leaves us in the north and west with a 3 day cold snap/spell.

Fingers crossed the snow showers push well inland.

 

[IDO]

4 minutes ago, bluearmy said:

It evolved in a similar fashion to the op. Nothing odd. Ridge a bit more suppressed south than the op but not enough to prevent a decent easterly flow to advect the low uppers west. 

Slightly different NH profiles a few days before that very low Dew point compared to current output:

Now if the ECM op was showing the latter chart 3 days before that -15c DP, I would not be as dismissive!!!

[karyo]

3 minutes ago, nick sussex said:

 

As you can see I'm desperately trying to fill some time until tonights outputs. We do have the SREF coming out shortly for the USA and given the timeframes that could give us a pointer to which model has the correct pattern.

I know the moderators don't like discussions about the Met Office update here but in this case it is very relevant given the uncertainty. They expect a northwest - southeast split for next week. The northwest unsettled the southeast dry so I think they see the ECM option as unlikely and something closer to the ensemble mean is more likely to verify.

[phil nw.]

The 00z London combined 2m temp ens. graph shows the Op runs diverge widely around the 16th- day 8.

A look at the 2 ops at that time

Another case of the different suites working out how much energy goes over the Azores ridge in deciding where the high cell will go after the northerly.

Clearly the ECM op is too amplified based on the ens graph with the GFS looking too flat in raising temperatures later on from the effects of the Atlantic.

My view would be something in-between with a high close to S.England/N.France with possibilities of some continental cooling further south.It's dicey i know looking at mean charts in this situation but the later ECM mean chart may not be far out,perhaps with the high a touch further north. 

 

[Timmytour]

10 minutes ago, Jason M said:

Hmm, last one from me on this. If events have proved the poster correct how can we question methodology? Maybe people who are saying different should question their own methodology (and that's not directed at anyone in particular). The longer term models have often signposted HLB this season and yet we have had pretty much the opposite of that. The outputs from these might be more interesting, but frankly they have been wrong. In our climate the skill is in picking out the cold spell during the winter months. Lets see if those who always say 'mild' get the outrider correct before we judge. And by that there is little skill in constantly highlighting cold set ups and claiming victory when one finally comes along.

The problem is as always that we need the cold spell before we can judge.

 

The "trick"of correctly forecasting mild is surely to point out what feature of a model indicating cold you don't expect to verify and how that will led to the mild.  A bit different to just claiming it will turn out mild  

[Nouska]

1 hour ago, IDO said:

If you believe that -15c Dew points are close to verifying, then I don't know what i can say! March 2013 was the previous time, -17c. The chart for that day very different:

I don't know how you can say that as the dewpoint was -15C not so very far from Holland, overnight Saturday. It is an extremely cold and dry airmass just to the east and should the flow change, it is certainly not beyond the realms of possibility.

Mid day yesterday after some of the cold air had been pushed back east.

No great leap of faith to envisage the mass moving west or south west with an easterly drift.

Leuven (central Belgium) ensemble from ECM - that's cold for there!

I'm taking note of the forecast for down here anyway - just on the edge of the -12C isotherm.

Just to add the day nine Mean SLP chart from ECM.

Perhaps more likelihood of the cold travelling south west rather than directly west but nothing there to say the det run is an impossibility.

Edited January 8, 2017 by Nouska
Editing on the hoof!

[Frosty.]

With sunshine and snow showers expected for Fri / Sat I would say the Ecm 00z is favourite, for saturday at least compared to the very progressive 6z which cuts the Northerly off before next weekend, anyway, sunshine and snow showers with frosts and icy patches will beat the hell out of what we have seen so far this winter! 

Edited January 8, 2017 by Frosty.

[nick sussex]

3 minutes ago, karyo said:

I know the moderators don't like discussions about the Met Office update here but in this case it is very relevant given the uncertainty. They expect a northwest - southeast split for next week. The northwest unsettled the southeast dry so I think they see the ECM option as unlikely and something closer to the ensemble mean is more likely to verify.

Regardless of the UKMO I still think the easterly has a chance. That nw/se split could be in evidence with a continental surface flow.The GFS can only verify if the ECM is wrong between T96hrs and T120hrs over the USA, and that UKMO outlook could implode if the ECM is proven to be correct between those timeframes, indeed if the GEFS have the wrong modelling of shortwave energy over the USA at that time then they can go in the bin. Of course the same can be said of the ECM ensembles. Its for this reason that I'd be wary of all the ensembles until we get agreement upstream. The ECM has the cleanest Atlantic with only a weakish shortwave, the GFS is much more messier and the UKMO is a halfway house.

[bluearmy]

12 minutes ago, IDO said:

Slightly different NH profiles a few days before that very low Dew point compared to current output:

Now if the ECM op was showing the latter chart 3 days before that -15c DP, I would not be as dismissive!!!

I rather suspect we wouldn't be chatting about dew points in holland if that were the case!

couple of points

a) you are comparing dp's in Amsterdam which is coastal so only offshore flows are likely to deliver low dp's. not ridiculously relevant as we are discussing easterlies but it needs to be stated when making the judgment

b) those are the numbers that the control delivers and I agree they are skewed too low because of the latitude the ridge gains.

I am always intrigued when a ridge throws itself so far without any apparent strong forcing to do so. One would summise its to do with the ECM high resolution feeding back a more erroneous solution with each calculation but then why would the control also go more than half way to the same place. surely we won't get four out of five ec ops doing the same thing at the back end? 

[beng]

Something that caught my eye somewhat is that this morning's ECM is the first morning ECM run to bring the Easterly on the Op; the previous times were both on the 12z Op run. Whether we can deduce much from that though, I'm not sure without knowing the reason that the 12z runs appear to have been more amplified up until now.

Edited January 8, 2017 by beng

[Seasonality]

 

26 minutes ago, Timmytour said:

The "trick"of correctly forecasting mild is surely to point out what feature of a model indicating cold you don't expect to verify and how that will led to the mild.  A bit different to just claiming it will turn out mild  

Exactly! You beat me to it Detailed analysis rather than the same old tiresome FI mean or ensemble trick which anyone can do. The real skill lies in the analysis.

Edited January 8, 2017 by Seasonality

[IDO]

6 minutes ago, bluearmy said:

I rather suspect we wouldn't be chatting about dew points in holland if that were the case!

couple of points

a) you are comparing dp's in Amsterdam which is coastal so only offshore flows are likely to deliver low dp's. not ridiculously relevant as we are discussing easterlies but it needs to be stated when making the judgment

b) those are the numbers that the control delivers and I agree they are skewed too low because of the latitude the ridge gains.

I am always intrigued when a ridge throws itself so far without any apparent strong forcing to do so. One would summise its to do with the ECM high resolution feeding back a more erroneous solution with each calculation but then why would the control also go more than half way to the same place. surely we won't get four out of five ec ops doing the same thing at the back end? 

If the 12z doesn't backtrack then we will have to re-evaluate as it is rare when ECM is wrong in these setups they run more than 3 "wrong" runs in a row!

The ECM 0z would be a game changer for mid Jan onwards if it verified and as Nick says it is early in the run so it has to be right or wrong in the next couple of runs.

Edited January 8, 2017 by IDO

[ArHu3]

13 minutes ago, bluearmy said:

I rather suspect we wouldn't be chatting about dew points in holland if that were the case!

 

I just showed the plumes for De Bilt because they show the most info on EC ENS without having to pay for it. You can also compare the plumes for the different regions and get an idea if there is a front somewhere across the Netherlands and you know that if the plume for western netherlands sucks and looks good for the east your will surely suck even more

Edited January 8, 2017 by ArHu3

[Tim Bland]

Personally I'm not buying the last 2  operational ECM runs. We know the Metoffice don't support it either according to their long range update in last hour. Ian F on twitter also states that only 10% of ECM ensembles support the operational. I urge coldies to expect the easterly to be dropped on the 12z ECM, that way you won't be disappointed! 

[Frosty.]

9 minutes ago, Tim Bland said:

I urge coldies to expect the easterly to be dropped on the 12z ECM, that way you won't be disappointed! 

At the moment I'm more interested in seeing the Arctic Northerly last throughout next weekend as per the Ecm 00z compared to the Gfs 6z which quickly shunts the cold away eastwards, hopefully the slower ecm solution will win out!

Edited January 8, 2017 by Frosty.

[snowbob]

Well I urge coldies to wait and see what the 12z brings

big set of runs coming up.

cant see the ecm and ukmo being that wrong against the gfs 

In recent years isn't it the gfs that drops the cold runs to go out on its own against ecm and ukmo.

only to pick up on them again a few runs later and fall in line,(sometimes lead the way )

not saying that will happen just remember the gfs doing this in cold spells previously.

[PersianPaladin]

22 minutes ago, Tim Bland said:

Personally I'm not buying the last 2  operational ECM runs. We know the Metoffice don't support it either according to their long range update in last hour. Ian F on twitter also states that only 10% of ECM ensembles support the operational. I urge coldies to expect the easterly to be dropped on the 12z ECM, that way you won't be disappointed! 

I'm sure you know already that the Operational Runs are at a higher resolution than the Ensembles. If there's a subtle pattern detection, then the higher-resolution the better as small differences can make a larger outcome for cold as far as this part of the world is concerned. Still, I share the view that the ECM will stop showing that easterly in the next run or so.

Edited January 8, 2017 by PersianPaladin

[nick sussex]

27 minutes ago, Tim Bland said:

Personally I'm not buying the last 2  operational ECM runs. We know the Metoffice don't support it either according to their long range update in last hour. Ian F on twitter also states that only 10% of ECM ensembles support the operational. I urge coldies to expect the easterly to be dropped on the 12z ECM, that way you won't be disappointed!

Lol! I think many of us are now immune to the easterly traumas. We've built up a lot of resistance over the years.

Does this give you confidence in the GFS solution?

The main conundrum now is both the 00Z GFS and ECMWF are towards the
deeper end of the 07/12Z NAEFS ensemble spectrum with respect to
lower heights aloft crossing the forecast area Friday. However, an
early look at 08/00Z NAEFS members (and specifically the GEFS
members) shows a decided shift towards this stronger solution likely
indicating something in initial data assimilation was finally picked
up in the GFS that was previous seen in the ECMWF initializations
for the past several runs.

The above from the Arizona state forecast doesn't fill me with confidence that the GFS is handling the upstream pattern well. If it took several runs to see something that the ECM saw several runs earlier. Of course the ECM maybe wrong we don't know yet but in the last week the GFS has been the last one at the party upstream.

Edited January 8, 2017 by nick sussex

[Eskimo]

Anyone have the ECM precip charts?

[Mark Neal.]

6 minutes ago, PerfectStorm said:

Anyone have the ECM precip charts?

https://www.wunderground.com/wundermap - Got to model data on the right tab. 

http://en.vedur.is/weather/forecasts/atlantic/#type=prec

https://www.yr.no/place/United_Kingdom/England/Manchester/long.html - This site uses ECM data to derive its forecasts.. 

Edited January 8, 2017 by Mark N

[shotski]

14 minutes ago, snowbob said:

Well I urge coldies to wait and see what the 12z brings

big set of runs coming up.

 

Totally agree, I think the next couple of days model watching will define the rest of winter. Looking forward to the 12z ECM. 

[Nick F]

24 minutes ago, bluearmy said:

I am always intrigued when a ridge throws itself so far without any apparent strong forcing to do so. One would summise its to do with the ECM high resolution feeding back a more erroneous solution with each calculation but then why would the control also go more than half way to the same place. surely we won't get four out of five ec ops doing the same thing at the back end? 

I tend to look at the formation of a Scandi ridge in two layers, a cold surface high formation once the upper trough over Scandi edges away E/SE while aloft a 'warm' ridge builds through strong WAA northeastward due to the strong height gradient / wind on the forward side of the Greenland polar vortex. The combination of these two layers allows the ridge to be thrown quite far northeast.  I imagine in my head that there may also be an empirical rule programmed into the EC computer model that will churn out a Scandi high if the evolution beforehand signals possibility of a ridge NE, 

The shape, intensity and position of the lobes Canadian / Greenland trop vortex IMO seems to be crucial in determining whether we get enough amplification in the flow to get a ridge to build NE towards Scandi and the Scandi trough next week to relax to allow a cold surface high to form to our NE and WAA spread aloft from the SW. 

Given the stubborness of the GFS ops to continually keep the upper flow flat when the EC high res wants to amplify - with not a great deal of ens support for the EC high res easterly, you would have to go for the idea that 12z EC  will model the ridge to still build but further south without the easterly, as per EPS mean.

But, the wildcard for me is: are the models handling the shape of the Canadian/Greenland PV correctly? As this has ramifications for the upper flow over the N Atlantic and potential for amplifications.

[bluearmy]

2 minutes ago, shotski said:

Totally agree, I think the next couple of days model watching will define the rest of winter. Looking forward to the 12z ECM. 

I checked my diary and it's the 8th Jan 

[snowice]

1 hour ago, nick sussex said:

We're in the unusual situation here of wanting the flatter earlier solution of the ECM rather than the more amplified shortwave of the GFS. The ECM runs shortwave energy east more quickly and then amplifies the flow after this . The GFS has initial the more amplified shortwave but then becomes flatter.

I should say something re the upstream pattern whilst NCEP still haven't made a call either way New Mexico state forecast has suggested the ECM is showing better continuity than the GFS with shortwave energy in its region and was more accurate last week. Aswell as this in other areas several state forecasts mentioned that some initial outlier solutions the ECM ops may have had over the USA are being moved to by the other outputs.

The upstream trough (which the GFS operational run and GFS/EC
ensemble means from yesterday indicated would follow a kind of flat
trajectory across the Great Basin) have now swung in direction of
the EC operational run which was the outlier in developing a
stronger trough down across CA by Wednesday night/Thursday.

As you can see I'm desperately trying to fill some time until tonights outputs. We do have the SREF coming out shortly for the USA and given the timeframes that could give us a pointer to which model has the correct pattern.

Hi Nick by 96hrs 12z gfs we should know once and for all?