Model output discussion - 5th December 2016 - Into Winter we go

[knocker]

19 minutes ago, Seasonality said:

No criticism of either you or knocker, but one thing this winter has shown so far is a disconnect between longer term forecasting and the resulting weather. At this range you both have a valid point and show signals both supporting and discounting such an outcome. But going back to the disconnect is it not possible we will have an unforeseen situation arise that goes against what some signals show. In a probabilistic sense we were meant to have a blocked "front loaded" (for cold) winter due to various supporting factors. The lower probability outcome has ended up prevailing due to other factors. Moving forward we could just as easily have a situation where a lower probability cold outcome could arise, potentially against some factors as Karyo has already put so well. It could be a winter of surprises in the timing of mild and cold outcomes. 

PS I do like seeing discussions like the two of you are having, improves everyone's knowledge

Actually I wasn't trying to make a point except that the 06z anomaly supports previous runs and the EPS on the possible evolution. it is not in isolation and given the circumstances it's a reasonable percentage play. Certainly better than instinct.

Edited December 20, 2016 by knocker

[carinthian]

4 hours ago, carinthian said:

No real change in the forecast chart for the big day from the UKMO. No cruel deep winter charts for most including the Eastern states of the USA , looks like a warm flow on this chart up to the Great Lakes. Hopefull;y not too far beyond the Christmas Period we will see the stubborn high pressure zone relocate to a postion further North or West so we can tap into some real winter charts.  Still early days yet and its just amazing that the promise of full throttle zonal charts just spin the lows north with no real impact against the Euro block for many folk in Euroland and Southern Britain. Just to cheer you all up , Sweden and Norway had some temps around 10c yesterday .

 C

From a purely selfish point of view this is my favourite chart of the day. Blizzards for the Alps with a frigid air mass straight from deepest Siberia to sweep the plains of Northern Europe. 1070 mb Greenland high ready to link up with UK high and to open the flood gates for super cold Arctic air to sweep through the British Isles. We can only dream, but its there ! As I said in the post above we have a blocked winter but we need to get it in the right place, plenty of time yet .

 C

[Singularity]

One thing that's struck me in the shorter term is that the progression of frontal systems has accelerated so much as to limit the amount of rain across southern parts to nothing of much note. What with the signal for a strong height rise on Boxing Day it's possible that it will end up being a fairly tepid attempt at restoring water tables toward normal levels after the prolonged largely dry spell of the past 3-4 weeks.

Longer term, recent GFS runs are reflecting the enthusiastic nature of GEFS with the MJO but only if/when the other models start to share that enthusiasm will I dare to imagine any poleward ridges capable of acting like a brick in a washing machine with respect to the vortex. As last winter showed, it has to be something that dramatic to produce more than a transient sort-of-cold spell (which failed to deliver wintry precipitation in my far-southern location, let alone lying snow).

Almost four years now since my last day of lying snow. Anyone who's been waiting less than that, what are you moaning about? 

[Methuselah]

I must admit, that my being a fan of extremes, the T+201(?) has me wondering just how warm can a January day actually be: javascript: viewimage(1);

[Interitus]

 

1 hour ago, karyo said:

The 30hpa charts show a stronger, flatter vortex centered in the Arctic.

I always follow the chart provided by  netweather : http://www.netweather.tv/index.cgi?action=nwdc;sess=

Higher altitude charts eg 30hPa compared to 100hPa, will always show a 'strong flatter' vortex because the higher number tropospheric waves i.e shorter waves, are filtered out by critical level filtering, as their phase speeds match the background wind speeds. This is why really only wave-1 and wave-2 are considered for middle and upper strat.

With regards to strat vortex strength and favourability for northern blocking, there is only a weak correlation with the upper levels of the strat and the troposphere particularly for our neck of the woods with the NAO. For example for Januaries since 1979 there are 5 strong +ve NAO daily values >2 - all had 10mb wind speeds >40m/s. There are 4 strong -ve NAO daily values <-2 - they too all had 10mb wind speeds > 40m/s.

Edited December 20, 2016 by Interitus

[mountain shadow]

54 minutes ago, Seasonality said:

No criticism of either you or knocker, but one thing this winter has shown so far is a disconnect between longer term forecasting and the resulting weather. At this range you both have a valid point and show signals both supporting and discounting such an outcome. But going back to the disconnect is it not possible we will have an unforeseen situation arise that goes against what some signals show. In a probabilistic sense we were meant to have a blocked "front loaded" (for cold) winter due to various supporting factors. The lower probability outcome has ended up prevailing due to other factors. Moving forward we could just as easily have a situation where a lower probability cold outcome could arise, potentially against some factors as Karyo has already put so well. It could be a winter of surprises in the timing of mild and cold outcomes. 

PS I do like seeing discussions like the two of you are having, improves everyone's knowledge

Have to disagree a wee bit with this.

As per Steve Murr's post yesterday, with the PV cranking up the longer term forecast (4-6 weeks) becomes easy to forecast. We will get either a flat zonal jet or a mid latitude block. Obviously within this, there is the prospect of a Faux Cold High and perhaps a Northerly 12 hour toppler.

It is forecasting proper cold for these islands at that timeframe which is impossible. We had a very weak PV going into Winter hence the forecast of colder weather, unfortunately, some teleconnection or other didn't play ball.

I therefore do not see it as probable at all that a cold outcome could occur.

 

[Summer Sun]

GFS 00z para with an FI tease

Edited December 20, 2016 by Summer Sun

[Frosty.]

The outlook from the models for the new year doesn't sound bad with high pressure in charge bringing fine days and cold nights with widespread frosts and some fog..could be a lot worse (mild) and there is a carrot being dangled for high pressure to pull west into the atlantic with Northerly incursions.

Edited December 20, 2016 by Frosty.

[Trom]

3 hours ago, Nick F said:

00z EC ensemble mean and normalized standard deviation chart for day 10 suggests good confidence (greens) in a ridge building N over UK. So looks like the days of Atlantic influence are numbered, but still a long way from getting some sustained cold and wintry weather.

Hello Nick can you give me a little info on how to interpret that.  What mean is being used to calculate these measures is it the current ensemble mean or some kind of long run average?  The chart title infers the former in the first diagram. Isn't it just showing where the 500hpa heights are the most uncertain per the ensemble data and where the models show more confidence?  I only ask as the second model shows some areas +/- 25 standard deviations from the mean which would give a huge range of potential outcomes which has me questioning whether I'm interpreting it correctly. (Also as an aside is there median data available as this is less likely to be skewed by outliers).

[Iceberg]

I don't buy any of this defeatist Malarky tbh. Nothing is showing mild mush and everything is all to play for come January. 

I remember a few making comments about fantasy easterlies etc just last week. Which were never on imho. Equally talk of 4/6 weeks is barmy. 

The gfs para shows clearly what is possible. I am not talking about the fantastic uk Synoptics, but it shows considerable wave activity and warming. It shows a collapse of the pv. It's also not alone and quite a few ens from the operational supported this as we moved into Jan. 

It needs to be watched carefully. 

Edited December 20, 2016 by Iceberg

[Interitus]

2 minutes ago, Iceberg said:

It shows a collapse of the pv. It's also not alone and quite a few ens from the operational supported this as we moved into Jan. 

It needs to be watched carefully. 

Wow, sub -92°C or 181K at 10mb would be record cold, previous lowest is 182.38K.

Vortex not quite collapsing in that image.

[Seasonality]

27 minutes ago, mountain shadow said:

Have to disagree a wee bit with this.

As per Steve Murr's post yesterday, with the PV cranking up the longer term forecast (4-6 weeks) becomes easy to forecast. We will get either a flat zonal jet or a mid latitude block. Obviously within this, there is the prospect of a Faux Cold High and perhaps a Northerly 12 hour toppler.

It is forecasting proper cold for these islands at that timeframe which is impossible. We had a very weak PV going into Winter hence the forecast of colder weather, unfortunately, some teleconnection or other didn't play ball.

I therefore do not see it as probable at all that a cold outcome could occur.

 

Polar vortex forecasts viewed in isolation then yes, that is a good point. However, the polar vortex is not the sole factor driving the weather and you're working on the assumption that the forecasts will be correct, just as the initial winter forecasts must have factored in more than the vortex and also assumed those factors would be present further down the line. Furthermore the vortex remaining strong for the next 4-6 weeks is by no means a foregone conclusion. It simply isn't possible to speak of the weather with the level of certainty you just have and I'm sure you know that *winking smiley face

Could go on but I don't want to clog up the thread, I just think you've oversimplified a bit.

 

[Interitus]

26 minutes ago, Trom said:

Hello Nick can you give me a little info on how to interpret that.  What mean is being used to calculate these measures is it the current ensemble mean or some kind of long run average?  The chart title infers the former in the first diagram. Isn't it just showing where the 500hpa heights are the most uncertain per the ensemble data and where the models show more confidence?  I only ask as the second model shows some areas +/- 25 standard deviations from the mean which would give a huge range of potential outcomes which has me questioning whether I'm interpreting it correctly. (Also as an aside is there median data available as this is less likely to be skewed by outliers).

The large deviations in the second image are of the operational from the ensemble mean, so for example look at large deviation related to the heights over Finland - details like that are washed out in the ensemble mean.

edit: also should add the lower deviations on the left are normalised, quickest here to cut & paste from ECMWF site

The left panel, meanwhile, shows the 'normalised standard deviation'; to compute this we take the standard deviation from the right panel and divide it by a 'mean standard deviation'. This 'mean standard deviation' is a pre-computed field, that is a function of lead time and of geographical location. If the initiation time of the selected forecast is 00 UTC it represents the mean of the standard deviations of the 30 most recent 00 UTC ECMWF ensemble forecasts. Similarly for a selected 12UTC forecast the 30 most recent 12 UTC forecasts are used.

So, as an example, if the spread at day 5 in a particular ensemble forecast set (right panel) seems to be large, but has of late also tended to be equally large at day 5 in the same area, then the left panel shading will denote a value that is close to 1 (i.e. no colouring). Conversely if the spread in a particular area at day 5, in one ensemble, exceeds the spread that had recently been seen there at day 5, then the left panel shading will indicate a value rather greater than 1 (purple shading).

More details and interpretation - http://www.ecmwf.int/en/forecasts/charts/medium/ensemble-mean-and-spread-four-standard-parameters?time=2016122000,0,2016122000&parameter=MSLP&area=Europe

Edited December 20, 2016 by Interitus

[karyo]

1 hour ago, Interitus said:

 

Higher altitude charts eg 30hPa compared to 100hPa, will always show a 'strong flatter' vortex because the higher number tropospheric waves i.e shorter waves, are filtered out by critical level filtering, as their phase speeds match the background wind speeds. This is why really only wave-1 and wave-2 are considered for middle and upper strat.

With regards to strat vortex strength and favourability for northern blocking, there is only a weak correlation with the upper levels of the strat and the troposphere particularly for our neck of the woods with the NAO. For example for Januaries since 1979 there are 5 strong +ve NAO daily values >2 - all had 10mb wind speeds >40m/s. There are 4 strong -ve NAO daily values <-2 - they too all had 10mb wind speeds > 40m/s.

Thank you for the response. I don't disagree with what you are saying but I am wondering why net weather only has the 30hpa charts. It makes it easy to make the assumption that this is the most relevant chart while as you explain it is not.

[Banbury]

2 hours ago, Interitus said:

They really aren't cold temperatures for midday on 26th December.

Chilly not cold I said but too be fair below the average for most of the UK

[Dennis]

57 minutes ago, Summer Sun said:

GFS 06 para with a FI tease

 

[Interitus]

Just now, karyo said:

Thank you for the response. I don't disagree with what you are saying but I am wondering why net weather only has the 30hpa charts. It makes it easy to make the assumption that this is the most relevant chart while as you explain it is not.

Well charts such as 30hPa are good for seeing the strength of the strat and warmings etc, it's just that they don't always predict the troposphere very well, depending on coupling.

[Trom]

31 minutes ago, Interitus said:

The large deviations in the second image are of the operational from the ensemble mean, so for example look at large deviation related to the heights over Finland - details like that are washed out in the ensemble mean.

edit: also should add the lower deviations on the left are normalised, quickest here to cut & paste from ECMWF site

 

The left panel, meanwhile, shows the 'normalised standard deviation'; to compute this we take the standard deviation from the right panel and divide it by a 'mean standard deviation'. This 'mean standard deviation' is a pre-computed field, that is a function of lead time and of geographical location. If the initiation time of the selected forecast is 00 UTC it represents the mean of the standard deviations of the 30 most recent 00 UTC ECMWF ensemble forecasts. Similarly for a selected 12UTC forecast the 30 most recent 12 UTC forecasts are used.

So, as an example, if the spread at day 5 in a particular ensemble forecast set (right panel) seems to be large, but has of late also tended to be equally large at day 5 in the same area, then the left panel shading will denote a value that is close to 1 (i.e. no colouring). Conversely if the spread in a particular area at day 5, in one ensemble, exceeds the spread that had recently been seen there at day 5, then the left panel shading will indicate a value rather greater than 1 (purple shading).

More details and interpretation - http://www.ecmwf.int/en/forecasts/charts/medium/ensemble-mean-and-spread-four-standard-parameters?time=2016122000,0,2016122000&parameter=MSLP&area=Europe

Ok I've got that now.  So right shows actual standard deviation of operating run from ensemble mean.  The left shows the standard deviation divided by long run mean standard deviation so we can get an idea of whether the right hand diagram is particularly unusual. or close to the normal spread.  Thanks for your time.

[Iceberg]

33 minutes ago, Interitus said:

Wow, sub -92°C or 181K at 10mb would be record cold, previous lowest is 182.38K.

Vortex not quite collapsing in that image.

the tropo response was a big fat high pressure, not a pv endured low pressure. The tropo vortex is pretty dead. 

[Paul]

A couple of pre-Xmas upgrades to share with you. 

Firstly the NetWx models have had a big update in the last couple of weeks, increasing the vertical levels by almost double and adding a new convective scheme, both of which have shown to improve various facets of the modelling. 

http://www.netweather.tv/index.cgi?action=netwx-sr;sess= 
http://www.netweather.tv/index.cgi?action=netwx-mr;sess=

Secondly, we've just put a parallel viewer online for the GFS 2017 upgrade (due May). There's a standard and an Extra one, the link below will take you to the extra one if you're a full subscriber and logged in, otherwise it'll take you to the standard version. 

http://www.netweather.tv/index.cgi?action=gfspara;sess=

Hope they're useful  

[Iceberg]

Re polar vortex. 

This is the easiest piece I can find that is from a reputable site. 

http://scijinks.jpl.nasa.gov/polar-vortex/

thanks to NASA. 

Now look at the last operational and para from gfs as we enter January. 

I think it's clear that the pv is broken in the Atlantic as we enter January. 

We've seen the strato struggling so far this autumn winter with a disconnect. 

[Scandinavian High.]

This is what I like to see maybe a change to a colder outlook is lurking.

Edited December 20, 2016 by abbie123

[Paul]

Another quick one, and this is really just a community tool rather than something we'll widely be showing on the site. But if you want a quick comparison between the GFS and GFS para you can here (00 and 06 runs only)

http://www.netweather.tv/index.cgi?action=gfscomp;sess=

[gottolovethisweather]

4 hours ago, Banbury said:

Chilly not cold I said but too be fair below the average for most of the UK

I personally don't understand all the debate over whether something is cold or cool or balmy as to me it's all barmy and doesn't do much to assist confused or those less-experienced in model reading. But today was cold in Newbury, Berkshire today with a 5.5c Maximum at midnight and a daytime high of 4.4c; that's cold enough for wintriness given the right parameters. As for the upcoming days, it is suggested the big day itself will be mild, but even that could change right now, aside from the more cool rPM air than anything else with the briefest warm sectors in between. Come, Boxing Day into the turn of the year, it may well become more settled and more appropriately less STORMY, not that the South should see much in the way of rain or gales, bar tomorrow if current expectations go to plan. 

 

Edited December 20, 2016 by gottolovethisweather

[jvenge]

1 hour ago, Paul said:

Another quick one, and this is really just a community tool rather than something we'll widely be showing on the site. But if you want a quick comparison between the GFS and GFS para you can here (00 and 06 runs only)

http://www.netweather.tv/index.cgi?action=gfscomp;sess=

What are your thoughts on its usefulness? With a para run and considering how many GEFS usually diverge after day 5 to 6, how do you weigh it against say potential improvement vs just variability between different ensembles?