Stratosphere Temperature Watch 2013/2014

[Interitus]

The Aleutian high has been prominent, but what has been significant is high pressure in both North Atlantic and Pacific areas as seen in the 7-day averages, which has persisted through the depth of the atmosphere-

 

1000mb

 

 

50mb

 

 

This gives rise to the geopotential wave 2 activity which can be seen simultaneously at present forcing the polar jet near the 250mb level, and the higher levels of the strat vortex.

 

 

Observation and modelling suggests that period of a week or more is required for transient tropospheric forcing of SSW with extended periods of moderate forcing more significant than short sharp forcings which are reflected -see Sjoberg & Birner 2012 http://birner.atmos.colostate.edu/papers/sjoberg+birner12.pdf

 

The lower vortex has been squeezed and is compact, but quite strong-

 

 

And the Berlin ECM wind plots so far have shown reductions in zonal wind velocity up to 10mb as pointed out, though not <20m/s associated with a minor warming so far, while the upper vortex continues to strengthen.

It looks like it should be able to recover, though to what extent is a guess.

 

[Recretos]

GFS 12z forecast: 3D geopotential height, with colours representing temperature. Layers from top down: 10mb, 30mb, 50mb, 100mb and 300mb. Time period from 27/11 to 4/12, one image per day. The animation repeats 4 times, so try to find some propagation. The facing is the same as on the "instant weather maps", with USA in front and EU to the right. Cant wait to do these animations with some actual SSW related up/down-welling.

 

gfs.mov

 

And the CFS just doesn't want to give up on that Aleutian high.

 

 

I was reading some recent studies on the daily CFSv2 and stratosphere, and I came across this abstract from a recent study from Qin Zhang, Chul-Su Shin, Huug van den Dool and Ming Cai.

 

"...We further show that the CFSv2 has a high prediction skill in the stratosphere both in an absolute sense and in terms of gain over persistence except in the equatorial region where the skill would mainly come from persistence of the quasi-biennial oscillation signal. We present evidence showing that the CFSv2 forecasts can capture both timing and amplitude of wave activities in the extratropical stratosphere at a lead time longer than 30 days."

 

"...In particular, the CFSv2 is capable of predicting mid-winter polar stratosphere warming events in the Northern Hemisphere and the timing of the final polar stratosphere warming in spring in both hemispheres 3–4 weeks in advance."

 

It would be interesting if I could get a hold onto some daily averages from the daily CFS, except only individual run averages. As a said some time ago, I do remember some updates or fixes being applied last year, that probably affected the increase in skill scores since the bad scores in 2010 that were presented in the study by Perlwitz and Zhang. Tho the CFSv2 ensembles still have some issues with implementing the Eurasian snow extent.

 

Looks like fun times could be ahead. 

 

Best regards.

 

Edit: Thanks Lorenzo. My goal this year is to test these visualizations and products, especially if we manage to get a MMW, and analyse propagations with the help of these 3D animations.

Edited November 18, 2013 by Recretos

[lorenzo]

GFS 12z forecast: 3D geopotential height, with colours representing temperature. Layers from top down: 10mb, 30mb, 50mb, 100mb and 300mb. Time period from 27/11 to 4/12, one image per day. The animation repeats 4 times, so try to find some propagation. The facing is the same as on the "instant weather maps", with USA in front and EU to the right. Cant wait to do these animations with some actual SSW related up/down-welling.

 

gfs.mov

 

Best regards.

 

Am pretty confident that I won't see anything better than that by way of mode output from anywhere for this entire Winter.

 

This is excellent work. When we get some SSW activity these will simply be stunning. 

 

Amazing.

[chionomaniac]

I second what Tony says. Brilliant work Recretos. Here's hoping that we have a SSW to feed into the animator later this winter!

[Cloud 10]

Blimey! best buy some 3d glasses. 

[anthonym]

So, classic wave 2 response reaching top of stratosphere in about a week. Around that time, we may have some tropospheric players in-place to favor a wave 1 (N. Atlantic High / a brief return of the Aleutian Low). The uncertainty, there, will be if their transient nature reduces the amount of meridional heat flux. The deformation from the wave 2 seems to allow the Siberian-end of the dipole to wind-up the strongest. A wave 1 could help push it back towards the pole and consolidate/reconfigure the PV again? Or will the next wave 1 further break the beast down?

 

By the way, MJO-modeling right now is terrible. I would caution against its use. The wave is around phase 3 or so and will be propagating through Indonesia / W. Pac next week or so.

[Tamara]

I have been provisionally thinking that a complete SSW and disintegration of the vortex is much less likely this winter than last and split/displacement is the best chance. We saw though what happened in late Jan 2009 with a +QBO state (in early stratosphere learning days) but what test if any did we have then to know what wave responses were going on in late autumn 2008 by way of comparing this years progress so far against?

 

The ENSO state was less favourable back in winter 08/09 than it is now, and also arctic sea ice state/Eurasian albedo snow cover feedback and solar activity have become increasingly identified factors since stratosphere research began in these threads back then.

Edited November 18, 2013 by Tamara תָּמָר

[chionomaniac]

So, classic wave 2 response reaching top of stratosphere in about a week. Around that time, we may have some tropospheric players in-place to favor a wave 1 (N. Atlantic High / a brief return of the Aleutian Low). The uncertainty, there, will be if their transient nature reduces the amount of meridional heat flux. The deformation from the wave 2 seems to allow the Siberian-end of the dipole to wind-up the strongest. A wave 1 could help push it back towards the pole and consolidate/reconfigure the PV again? Or will the next wave 1 further break the beast down?

 

By the way, MJO-modeling right now is terrible. I would caution against its use. The wave is around phase 3 or so and will be propagating through Indonesia / W. Pac next week or so.

The problem is at this time of year the upper vortex is so strong and still strengthening so that any wave activity is unlikely to break it down. Mind you, not that it needs to, as previous early winter trop induced  lower strat splits have been very favourable here - but maybe not so in your neck of the woods. December SSW's are not exactly common (especially recently and why is that?), though previously there have been a number of CW's in wQBO years, notably in Dec1978, a key analog year.

 

Interestingly the latest 30 hPa anomaly for the middle third of the month is starting to take the shape of previous years analogs - but I guess any possible lower split affecting 30 hpa may now put paid to that.

 

Latest

 

 

 

Analog

 

 

If we get strong wave 1 activity  following this wave 2, would you not suspect (assuming that the siberian side of the dipole is the strongest) that the possibilty exists that the Canadian vortex weaker segment could give way allowing a CW? All conjecture of course because we are a far way away from that at the moment!

[Interitus]

 

I was reading some recent studies on the daily CFSv2 and stratosphere, and I came across this abstract from a recent study from Qin Zhang, Chul-Su Shin, Huug van den Dool and Ming Cai.

 

"...We further show that the CFSv2 has a high prediction skill in the stratosphere both in an absolute sense and in terms of gain over persistence except in the equatorial region where the skill would mainly come from persistence of the quasi-biennial oscillation signal. We present evidence showing that the CFSv2 forecasts can capture both timing and amplitude of wave activities in the extratropical stratosphere at a lead time longer than 30 days."

 

"...In particular, the CFSv2 is capable of predicting mid-winter polar stratosphere warming events in the Northern Hemisphere and the timing of the final polar stratosphere warming in spring in both hemispheres 3–4 weeks in advance."

 

It would be interesting if I could get a hold onto some daily averages from the daily CFS, except only individual run averages. As a said some time ago, I do remember some updates or fixes being applied last year, that probably affected the increase in skill scores since the bad scores in 2010 that were presented in the study by Perlwitz and Zhang. Tho the CFSv2 ensembles still have some issues with implementing the Eurasian snow extent.

 

 

Not read the paper yet, though they have a presentation which gives just enough to guess some of it. Not much to see but it's here - http://www.cpc.ncep.noaa.gov/products/outreach/workshops/CDPW38/1_Monday/2_Afternoon/Cai.pdf

 

Anyway, it's slightly odd that this is not the result that was shown at the AMS conference on the Middle Atmosphere earlier this year. The recorded presentation entitled Evaluation of the CFSv2 45 day forecasts to capture stratospheric events and usefulness for intra-seasonal forecasts of stratospheric winter hemispheric conditions by Craig Long (NOAA) can be seen here - https://ams.confex.com/ams/19Fluid17Middle/webprogram/Paper227228.html

 

In summary, the conclusion is that the CFS detects SSW within 15 days, but is no better than GFS -

 

Edited November 18, 2013 by Interitus

[chionomaniac]

I have been provisionally thinking that a complete SSW and disintegration of the vortex is much less likely this winter than last and split/displacement is the best chance. We saw though what happened in late Jan 2009 with a +QBO state (in early stratosphere learning days) but what test if any did we have then to know what wave responses were going on in late autumn 2008 by way of comparing this years progress so far against?

 

The ENSO state was less favourable back in winter 08/09 than it is now, and also arctic sea ice state/Eurasian albedo snow cover feedback and solar activity have become increasingly identified factors since stratosphere research began in these threads back then.

Annoyingly, I deleted most of the early winter strat charts in that thread back in those early monitoring days so I can't compare to now - Do you remember them Tamara (or should I say Zebra Danio!!! from back then)

 

I have checked and there was a large amount of autumnal wave 1 activity in 2008 but very little wave 2 activity. Still wouldn't rule out a SSW this year though.

Edited November 18, 2013 by chionomaniac

[Tamara]

Annoyingly, I deleted most of the early winter strat charts in that thread back in those early monitoring days so I can't compare to now - Do you remember them Tamara (or should I say Zebra Danio!!! from back then)

 

I have checked and there was a large amount of autumnal wave 1 activity in 2008 but very little wave 2 activity. Still wouldn't rule out a SSW this year though.

Thanks Ed

 

Yes I remember those early days and our posts... the novelty names best remain where they were though 

[Gael_Force]

Annoyingly, I deleted most of the early winter strat charts in that thread back in those early monitoring days so I can't compare to now - Do you remember them Tamara (or should I say Zebra Danio!!! from back then)

 

I have checked and there was a large amount of autumnal wave 1 activity in 2008 but very little wave 2 activity. Still wouldn't rule out a SSW this year though.

 

Don't know if you have read this paper - fairly in depth look at the event and lead up.

 

http://journals.ametsoc.org/doi/pdf/10.1175/2009JAS3320.1

Edited November 18, 2013 by Gael_Force

[Recretos]

 

Not read the paper yet, though they have a presentation which gives just enough to guess some of it. Not much to see but it's here - http://www.cpc.ncep.noaa.gov/products/outreach/workshops/CDPW38/1_Monday/2_Afternoon/Cai.pdf

 

Anyway, it's slightly odd that this is not the result that was shown at the AMS conference on the Middle Atmosphere earlier this year. The recorded presentation entitled Evaluation of the CFSv2 45 day forecasts to capture stratospheric events and usefulness for intra-seasonal forecasts of stratospheric winter hemispheric conditions by Craig Long (NOAA) can be seen here - https://ams.confex.com/ams/19Fluid17Middle/webprogram/Paper227228.html

 

In summary, the conclusion is that the CFS detects SSW within 15 days, but is no better than GFS -

 

summary.gif

 

 

 

Very interesting. I will definitely consider this one as more legitimate. But there are probably differences in the testing modes. Some used averaging of the runs, some just individual runs, etc... Too bad the full article that I mentioned is not free, so I could see what was their method and compare it to the results from Long. That pdf does give some basic info, but the methodology is not really described. 

 

 

Speaking of the 2009 event, I actually plotted some 3D visualisations (never before seen ) for it as a test. I extracted the data from the NCEP DOE Reanalysis 2 dataset, on 6-hourly intervals. 

 

The time period is from 12/01/09 to 05/02/09. The 3D surface is the geopotential height, and the colours are the temperature at that layer. Layers from top down: 10mb, 20mb, 50mb, 100mb, 250mb, 500mb and surface temperature.

 

https://dl.dropboxusercontent.com/u/16376055/09.mov

 

This is is from the pacific viewpoint:

https://dl.dropboxusercontent.com/u/16376055/09_pac.mov

 

This is just some preliminary testing. I am just trying to figure out the system, so in the future I will be able to plot basically any SSW event. Will try to figure out how to plot additional parameters too.

Best regards.

 

EDIT: As a bonus, here is another animation for the same period as above. There are only 2 layers. The layer above is the 3D 10mb geopotential height, and the second layer is the 10mb 3D temperature topography. The temperature topography is lower so both can be nicely seen, and the temp. topo kinda looks like a better indicator of the dynamics. 

 

https://dl.dropboxusercontent.com/u/16376055/09_up.mov

Edited November 19, 2013 by Recretos

[weatherjunkie]

 

 

 

Very interesting. I will definitely consider this one as more legitimate. But there are probably differences in the testing modes. Some used averaging of the runs, some just individual runs, etc... Too bad the full article that I mentioned is not free, so I could see what was their method and compare it to the results from Long. That pdf does give some basic info, but the methodology is not really described. 

 

 

Speaking of the 2009 event, I actually plotted some 3D visualisations (never before seen ) for it as a test. I extracted the data from the NCEP DOE Reanalysis 2 dataset, on 6-hourly intervals. 

 

The time period is from 12/01/09 to 05/02/09. The 3D surface is the geopotential height, and the colours are the temperature at that layer. Layers from top down: 10mb, 20mb, 50mb, 100mb, 250mb, 500mb and surface temperature.

 

https://dl.dropboxusercontent.com/u/16376055/09.mov

 

This is is from the pacific viewpoint:

https://dl.dropboxusercontent.com/u/16376055/09_pac.mov

 

This is just some preliminary testing. I am just trying to figure out the system, so in the future I will be able to plot basically any SSW event. Will try to figure out how to plot additional parameters too.

Best regards.

 

EDIT: As a bonus, here is another animation for the same period as above. There are only 2 layers. The layer above is the 3D 10mb geopotential height, and the second layer is the 10mb 3D temperature topography. The temperature topography is lower so both can be nicely seen, and the temp. topo kinda looks like a better indicator of the dynamics. 

 

https://dl.dropboxusercontent.com/u/16376055/09_up.mov

 

This is some great, great stuff. Can't wait to see what comes of this. 

 

On a side note, how has your MJO reanalysis/composite project gone? That one looked quite interesting as well. 

[MattH]

Some great posts last night...

 

12Z ECM 18th continues the trend of late;

 

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwf30f192.gif

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwf30f216.gif

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwf30f240.gif

 

What also stood out was this region of warmth evident from 216hr at 10hPa originating over Asia;

 

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwf10f216.gif

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwf10f240.gif

 

Decent wave 2 activity by day 9;

 

http://users.met.fu-berlin.de/~Aktuell/strat-www/wdiag/figs/ecmwf1/ecmwfzm_ha2_f216.gif

 

As some mentioned in the posts last night though, where we go from here seems the primary question given that it is obviously the time of year when the strat is continuing to cool and strengthen, is it a case of a battle that can't really be won at the present time just from some reasonable wave 2 activity and a split...

 

Cheers, Matt.

 

NB: Latest (overnight) EC32 update continues the trend from the last update and that is to maintain high pressure near or over the UK right through to the 15th of Dec. This seems quite an extended period for a time of year when climatologically you would expect some sort of Atlantic influence. I'm not 100% sure if the EC upgrades have been incorporated into this update or not, but I would image they will have been to the update later in the week, so it'll be interesting to see if there is a sudden shift. But again its high pressure and temperatures at or around 1C or 2C below average from the EC32 well into December.

[Catacol]

Sorry for asking a dull question, but if you dont ask you dont learn - so I always tell my students.

 

What is the difference between wave 1 or wave 2 temperature plots over on the Berlin site, and wave 1 or wave 2 heat flux plots? Basically what is heat flux? I notice that the wave 2 and wave 2 flux charts go hand in hand, but I dont understand what the difference is...

 

And what about momentum flux? I'm in the dark there too....

Edited November 19, 2013 by Catacol_Highlander

[lorenzo]

Have a scan of this one CH, wave 1 generated via wave 1 pattern in NH, wave 2 due to wave 2 pattern.

 

http://www.phys.ocean.dal.ca/people/po/Whistler/waugh.pdf

 

It is also widely recognized that planetary-scale waves control almost all aspects of the extratropical stratospheric circulation during Northern Hemisphere winter and play a key role in dynamical troposphere-stratosphere coupling [Plumb, 2010].

The waves originate in the troposphere due to topography and land-sea contrasts and have vertical wavelengths that span the troposphere and

stratosphere during winter, consistent with linear theory [Charney and Drazin, 1961].

 

 

The stratospheric planetary-scale wave heat flux, which is related to the vertical group velocity of the wave,exhibits considerable variability about its time-mean value.

 

 

http://www.columbia.edu/~tas2163/Shaw_etal_2013.pdf

[chionomaniac]

Just quickly - the heat flux is the transfer of heat in the area during a given timescale - so if it is large then that is beneficial over a given area eg meridional. That would mean that the transfer of heat from south to north is great and this can have vortex disturbance properties as any warming does in the stratosphere. Poor heat flux will not disturb the vortex as much. Substitute momentum for heat and the answer is similar.

[lorenzo]

The Shaw paper above is a good read, whilst the main part of the paper goes about examining the difference between low and high top models in order to recognise the new methodology proposed i.e quantifying impacts based on heat flux between 60-90 degrees North.  HLSHF - High Latitude stratospheric Heat Flux.

 

The key points I guess relate to the relationship between Heat Flux anomalies and the positioning of the Jet.

 

+ve Heat Flux anomalies leading to an equatorward shift in the Jet, with -ve Heat Flux leading to a Poleward shift.

 

Heat Flux plots from Berlin Site - Wave 1 / Wave 2 / Overall

 

 

 

Re the Wave 1 and Wave 2 queries this excellent explanation by Forecaster may help paint a picture, also this loop is currently showing the Wave 2 500mb anomaly nicely.

 

Shaw writes that the -ve heat flux events are likened to the NAO and the downward wave coupling is similar to the positive phase of the NAO.

 

Edited November 19, 2013 by lorenzo

[Interitus]

Have a scan of this one CH, wave 1 generated via wave 1 pattern in NH, wave 2 due to wave 2 pattern.

 

http://www.phys.ocean.dal.ca/people/po/Whistler/waugh.pdf

 

.

 

Although they quote that "All stratospheric events are preceded by bursts of EP flux" it is important to note that not all bursts of EP flux lead to stratospheric events.

It may be absorbed or reflected - see Harnik 2009 or Sjoberg & Birner 2012.

This presentation by Colucci considers polar cap potential vorticity tendency to be possibly the determining feature as always required -

https://ams.confex.com/ams/19Fluid17Middle/webprogram/Paper227033.html

 

However, as we only get to see most indices like this after they've already occurred, we have to guess what is going to happen!

Edited November 19, 2013 by Interitus

[Mucka]

 

 

 

Very interesting. I will definitely consider this one as more legitimate. But there are probably differences in the testing modes. Some used averaging of the runs, some just individual runs, etc... Too bad the full article that I mentioned is not free, so I could see what was their method and compare it to the results from Long. That pdf does give some basic info, but the methodology is not really described. 

 

 

Speaking of the 2009 event, I actually plotted some 3D visualisations (never before seen ) for it as a test. I extracted the data from the NCEP DOE Reanalysis 2 dataset, on 6-hourly intervals. 

 

The time period is from 12/01/09 to 05/02/09. The 3D surface is the geopotential height, and the colours are the temperature at that layer. Layers from top down: 10mb, 20mb, 50mb, 100mb, 250mb, 500mb and surface temperature.

 

https://dl.dropboxusercontent.com/u/16376055/09.mov

 

This is is from the pacific viewpoint:

https://dl.dropboxusercontent.com/u/16376055/09_pac.mov

 

This is just some preliminary testing. I am just trying to figure out the system, so in the future I will be able to plot basically any SSW event. Will try to figure out how to plot additional parameters too.

Best regards.

 

EDIT: As a bonus, here is another animation for the same period as above. There are only 2 layers. The layer above is the 3D 10mb geopotential height, and the second layer is the 10mb 3D temperature topography. The temperature topography is lower so both can be nicely seen, and the temp. topo kinda looks like a better indicator of the dynamics. 

 

https://dl.dropboxusercontent.com/u/16376055/09_up.mov

 

 

 

This is brilliant work.

I know you gave the time line but is there anyway you could add that to the animations?

Not a criticism at all, one again great work and thank you.

[Catacol]

A big thank you to those that posted an answer to my flux question. Plenty there for me to read and get my head around. There are times when I wish I was more scientist and less historian... though those moments tend not to last very long!! :-)

[johnholmes]

You lot are making my head hurt. I am getting too old so have thrown in the towel on this area of meteorology I am afraid. But there is certainly a huge amount of work being done by some of you-keep going, it may well bring some further light on this fascinating topic. I'll stick to my very amateurish look at the 30mb temperature profile.

[Catacol]

No Berlin update today? Is this because the new ECM is being set up? 

[bluearmy]

No Berlin update today? Is this because the new ECM is being set up?

Berlin updates at 6 am from the previous days 12z op run. I am very interested to see what the strat split looks like tomorrow after a much flatter op run this side of the NH.