[Stratosphere Temperature Watch 2015/2016]

Oh trust me, a full on wave 1 can be a harder case for the vortex to handle than a wave 2. if the vortex is strong, it can take the "energy" of the 2-wave and use it to its own advantage, remaining firm in place, unless a strong enough temp. waves are following. While in a full blown wave 1 it will get shoved out of place before it can say "is that a wave1 coming at me...". Of course with the normal disclaimers, like each scenario is different and each situation is unique.

[Stratosphere Temperature Watch 2015/2016]

I have done these posts before with images, but somehow Mr. Chionomaniac always seems to forget to add them to the first post.

 

Well, its quite simple really. Wave number 1, mean literally a 1 wave form in the stratosphere. A literal atmospheric wave, or make that like a giant rossby wave if you wish, but not that literally. 

 

An example (a quite strong one, tho, but still). Wave number 2 is the same principle, just with two waves, or two ridges instead of one. This particualr example is one of the CFSv2 ensemble perturbations for 25.12.2015. 

 

 

[Stratosphere Temperature Watch 2015/2016]

Here is some long range magic from CFSv2 daily ensembles, which I finally managed to average the right way, and they seem to be doing a good job, continuing with the promotion of the vigorous wave1.

 

  

 

Gonna add the NASA model, which I find to be very interesting, and going to 240h, it has forecasting potential together with ECMWF, FIM-9, GFS, JMA and ensemble guidance. 

 

 

Adding comparison of zonal mean zonal wind from NASA, GFS and ECMWF. 

 

  

 

And difference between NASAs GMAO and GFS, showing where GFS has weaker or stronger zonal wind mean component. Not very useful now, but will be when/if stronger dynamics get into play.

 

[Stratosphere Temperature Watch 2015/2016]

Nothing much I can add to the overall discussions, given the time I have. I admit I have put weather on side for now, so I am just occasionally peaking in here and generally in the strat development.

 

Just gonna add some ENS guidance from the new GEFS version, and also bias-corrected, which is all-in-all the best free ENS guidance for strat at the moment.

 

 

[Stratosphere Temperature Watch 2015/2016]

I was in contact with Ryan Maue from Weatherbell, discussing the ECM32, or make that ECM45. He accepted my suggestion of adding 10mb Gph+Temp fields to the ECM weeklies model page on Weatherbell, for the ens mean and control run. If it happens, it will be the best ENS support for strat yet. Of course for those subscribed to Weatherbell. It will probably be implemented sometime this month at Weatherbell, till December starts, along with the upgrade to 45 day forecast period, up from current 32 days.

[Stratosphere Temperature Watch 2015/2016]

Congrats on a new season and a great new intro post!

 

I have nothing major to share, just some forecast anomalies, showing the GEFS and GFS going for a bit less organised vortex going into November, as expected.

 

 

 

  

 

The zonal wind zonal mean forecast from GFS does show a less organised vorex, also compared to this time last year. Of course its just a forecast. 

 

 

[Stratosphere Temperature Watch 2014/2015]

 

Though I agree that a logarithmic view is indeed a better approach, with the limited capacity of the ESRL plots...

 

Oh the good old days, of being limited to the ESRL and online plotting. I know exactly where you come from, dont worry. I was just trying to add generally for all the rest, that he best way to look at the QBO is also as a zonal mean zonal wind on a latitude cross section. i was not trying to say that your plots are not right or anything. God forbid. Keep up the good work.

[Stratosphere Temperature Watch 2014/2015]

Cool to have this QBO explanation. It was explained many times, but always piece by piece, or just mentioned.

 

The best way to look at the QBO is through a latitude cross section on a zonal mean. There you can get the best feeling where its at and how strong. I am not a fan of using zonal wind anomalies for assessing the QBO, since it is already a bi-polar feature by nature. 

 

So here is an example of a zonal mean from GFS analysis and forecast, where you can see both phases of QBO. The best way to look at QBO is on a logarithmic vertical scale. 

 

  

 

And analysis focused on the QBO region, and a longitude cross section, like you did with ESRL plots.

 

 

 

@knocker: Not sure exactly, since I hardly read any papers since 2011, but I do remember someone finding some additional features via solar cycle among other stuff, and the BDC stratosphere connection, since the QBO could connect to the troposphere also by the stratospheric and SSW dynamics. Kinda like a backdoor entrance to the troposphere. But there is more to this than just the polar vortex in this case.  

[Stratosphere Temperature Watch 2014/2015]

Speaking of model ability to forecast. The Euro/Atlantic wavebreaking, together with the wave 1 from the Pacific side, will make the first wave2 attack on the polar vortex, though with not much long term damage. but still, its not a bad attempt, and of course gives me the opportunity to quickly model it in 3D, where this is even better shown than on 2D charts.

 

  

 

 

You can see how the bottom strat and upper trop is being squeezed in at 300mb where the bottom layer is, with a lagged response to the mid strat. The top layer is at 10mb. I added the 10mb contours at the top to better show what is happening with the mid vortex as a response to lower dynamics. This kind of "trop infused" dynamics are almost standard for the early cold season. Lets not forget the brief low to mid strat split from December 2012. these trop waves are know to sometimes being precursos to more serious events, but we are too soon in the season, and on a bit different configuration than in 2012/2013. What I am trying to say, is that one must not confuse pure direct trop induced dynamics with the dynamics and mechanics of an SSW. Not every wave2 is the same, even if it might briefly look the same on a 10mb 2D chart. Besides, this is not a strong wave2 at all, due to its background, and its much more of a G-wave than a T-wave, which is almost a classic footprint of a brief direct trop induced wave2. 

 

Anyhow, let the animation roll, and I hope it gets the message across better than my ramblings. This is tho nonetheless a good opportunity to test my 3D matrix in an operational forecasting way, to help make certain dynamics more clear and to show a better picture, without having to analyse 200 2D maps. at least for me, but thats probably because I kinda got used looking at it while constructing it.

 

 

Anyhow, the vortex will recover, no sweat, and will avenge to the trop core, sending it somewhere in the Greenland/Atlantic sector, keeping the NAO at least neutral if not positive. 

kind regards.

[Stratosphere Temperature Watch 2014/2015]

Finally managed to animate my new 3D test matrix for SSW events. In this example is the 2009 SSW event, showing the 150mb geopotential height on the bottom, and the polar vortex in 3D, from 1mb to 150mb, showing the vortex split and complete disintegration of the upper, mid and bottom strat vortex.

I think it is also appropriate for beginners in this field, to help them understand how an SSW event looks in "real life", since grasping this field from 2D maps alone is not that easy. Or it is at least to show what happens to the strat vortex during the event. If this matrix proves to be stable, I will be able to run it for any SSW you may want to see. 

 

 

Also going to add the 2009 SSW sequence done in my old matrix. Some improvements are seen in the new matrix above, like a new design, better resolution, new color schemes, and overall a different 3D "engine" than in this older version, for which the matrix I have lost anyway.

 

[Cooler Atlantic]

Thank you all. Modelling the Atlantic or any other part of the world is not that easy. But its nothing compared to the atmosphere, and multiple layering and 10x more data.

https://forum.netweather.tv/topic/81567-stratosphere-temperature-watch-20142015/?p=3269615

[Cooler Atlantic]

Thanks

Its fun yes, and also very mind-heavy. Not to mention the vast amounts of patience it takes, when you come to the point when after 2 hours of work your computer freezes while rendering and loading data, and you dont know if it will un-feeze and finish the job, or you just have to kill the process and start all over again.

With my old comp the usual solution was to kill the process and start over until it freezes again and then you kill it again and start over and it freezes again and so on. And each time you change the configuration a bit, reducing the quality or sample size and subsets, hoping it will work and load, but it just doesnt cut it. You just cant drive 200mph with an old 1.2l engine, now can you.

So I decided to upgrade my 1.2 engine from 3GB RAM which I almost fried on many occasions, and a 2.4Ghz Intel proc. and a 256MB graphic card, to a 2.0 16v turbo engine.

AsRock H81 mother board

Intel i7-4770 @3.9Ghz Turbo

16GB Dual channel DDR3 RAM 1600mhz (2x8GB)

Gigabyte GeForce GTX-750ti 2GB VRAM

System disk SSD 150GB

HD SATA 2TB

Windows 10 64bit

Its not the best comp out there, but together with a full HD 22" monitor and a gaming mouse/keyboard and a full 5-year warranty, it wasnt a bad deal for 800 quid.

Now I am really having fun making plots, especially 3D.

full 3D animations are my next goal.

[Stratosphere Temperature Watch 2014/2015]

Well, I dont have the time for detailed explanations unfortunately, even tho I would love to write it, but instead, i decided to show you how GFS FI looks like in a more realistic way, so you can imagine it. basically where and how is the stratospheric vortex connected with the tropospheric one. In the transition zone between 150 to 300mb, the tropospheric effects start to get into play, so the game there is from both sides. but basically looking from 1mb down to 150mb (like where I made this 3D display), you get the clear idea where the main energy goes and where the two entities are connected. At least I hope that it gets the message across.

So as you can see, the main tropospheric extension of the strat vortex is actually the double feature, one over Siberia and the other in E'rn Canada. Looks like a "shoe", or where does the P.V. stand in the troposphere.

 

 

 

  

 

  

 

 

Added jet stream on these two.

 

[Stratosphere Temperature Watch 2014/2015]

QBO well positive, while negative phase lurking above. In the meantime, the polar night jet remains undisturbed mostly. 

 

 

And how do we stack up against last year?

[Cooler Atlantic]

So, have you ever wondered how exactly would N Atlantic look like if you would pump out all the water? Well, its one big mountain range (MAR - Mid Atlantic Ridge). It gives you a feeling and a different perspective about how it interact with the ocean currents and AMOC, perhaps kinda like mountains do with atmospheric flow.

 

[Cooler Atlantic]

Good work Singularity.

Tho you have to understand that it can also be reverse correlated. Constant zonal wind, or better yet constant low pressure at that area, will cause lots of surface wind stress, affecting the MLD and essentially causing mixing and upwelling and thus negativev anomalies.

Its kinda the chicken and the egg situation at first glance.

[Cooler Atlantic]

Adding more graphics of surface SST anomalies, over various latitudes, showing the progression of the cold pool this year over time. 

 

    

 

And this is a time-series, of monthly mean values form January 1985 to December 2014, so the past 30 years, at point 50N-26W. it is seen how lately the cold waters are more on a uprise, probably preceding the eventual onset of the cold AMO phase down the line somewhere.

 

[Cooler Atlantic]

Just how big of a pool this Atlantic cold pool really is? Well, it actually goes pretty deep and you could actually say that its a pool. This is a cross-section through the 50N lat line, versus the climatology and versus last year. Data is from Glosea5 analysis (FOAM).

 

 

 

 

[Winter 2015/16]

Latest POAMA model run is staying in its firm track since summer. it is also really trying to penetrate that lower pressure into Europe. 

 

[Cooler Atlantic]

Well gee, I was planning to post that one here myself, but oh well.

 

Since there was some talk about salinity, here is some recent data from my side. The most obvious feature, is of course that fresh water blob that still exists around the Beaufort Gyre, tho perhaps not at the same location as in the BFTV graphics. Could have moved in the past years. i will have to check it out with some reanalysis data. If this were the same batch of fresh water as the original "bulge", then this indicates that it has moved SW since then. Of course of note is also more fresher water in the North Atlantic.

 

  

 

And at depth, where the anomalies are less intense, but that is normal down with depth. We tho do have some lower salinity in the north cell of AMOC, which is interesting, and could perhaps indicate some mixing of the salty flow with the fresh waters from the north.. 

 

  

[Stratosphere Temperature Watch 2014/2015]

Not sure when the definition of the tropics and subtropics changed for the strat, but if there is a seperate official definition, I would love to hear it and use it.

Otherwise I will continue to use the natures own main frame definition of the main strat tropics, via the QBO variability. And by that definition, both 97 and 02 have tropics normal, to colder than normal in 02 on your charts. But certainly not warmer than normal.

[Stratosphere Temperature Watch 2014/2015]

 

 

Many researchers consider November 1968 to be a full SSW see eg. Cohen & Jones, Charlton & Polvani, Butler & Polvani etc

 

For the UK at least, the coldest weather in 1962/3 and 2009/10 was before the effects of the SSW.

 

Incidentally, what's the link for the GPH anomaly/ao charts for earlier years (1962/3, 1968/9)?

 

They are somewhereon the CPC site, but i have lost the link. I have saved some of the graphics tho, in case I actually lose the link, since the graphics are quite useful. I have it bookmarked somewhere on my old computer, and I will eventually recover it, or at least try to recover it.

 

 

A couple of observations w/r/t the next few weeks in the stratosphere.

 

Both GEFS and EC EPS depict a deep upper low forming over the NE Pacific around 8th October persisting for several days, followed by sequential +ve 500hPa height anomalies over the North Atlantic and then Scandinavia in the extended range.

 

gfs-ens_z500a_sd_nhem_20.jpggfs-ens_z500a_sd_nhem_33.jpggfs-ens_z500a_sd_nhem_51.jpg

 

With a 14-21 day lag applied, planetary wave activity over the Pacific low should allow for some moderate Wave 1 activity displacing the upper vortex towards Siberia and Wave 2 activity following end of the month into November. All in all, reasons to think that the stratospheric vortex will be on the weak and slightly displaced end of the spectrum come November.

 

The extended 00z EC EPS has continued to depict a warm 10 hPa layer over the tropics. This is an unusual pattern, not really seen since 2002 for the same period. 1997 was also similar in pattern for date.

 

compday_coJ4m0Om4A.jpgcompday_1ecwY0jtzS.jpg

 

Both these advertise an El Nino type pattern deeply engrained in the atmosphere. Check the correlation between 10 hPa temps and angular momentum>

 

86_152_19_204_277_5_39_21.jpg

 

All in all, a very different start to the stratospheric season beckons.

 

That is very interesting. 

Also because I just went over GEFS (new one), and it is actually colder than normal over tropics.

 

 

That includes the bias corrected version.

 

 

And the N.Hem (ignoring the south hem. since its going into summer) pattern actually kinda fits (in a very rough sense, since the its just a forecast with some errors) with the 81-14 correlation between ENSO 3.4 and 10mb Temperature. I really prefer to use this period for stratospheric correlations, because the quality of the higher atmospheric assimilated data is much better than in the 50-70 period for example which can create some noise. Besides, a 33 year period of quality data is more than enough (especially since its recent data), and better than 64 years of half-quality data. At least in my amateur opinion of course, I have to emphasise that.

 

 

And using the same correlation as above, the new period sheds a different light. and tropics wise, it looks quite similar if not the same as the ENSO 3.4 correlation, but that is normal since AAM and El Nino do go hand in hand. Also of note is that this graphic is for November, to make it the same as above from GP, and mine above is from October, since it relates to the GEFS forecast. Also of another note is the fact that the actual correlation factors are not that strong at all actually, but its not so important, since we only want to find the "signal" and not to make actual hard correlations. And the signal can be best seen in quality data IMHO.

 

 

And wondering how much bias correction there actually is? Well, not that much, generally the bias corrected version is just a bit warmer in the north hem., or lets say that the normal version is a bit too cold. Of course the bias is not completely removed in the corrected version. Some of the cold bias is probably still left. But i do wonder if the bias would really be so strong to be 6-8 degrees too cold in the tropics? That question will be answered once we get close to this period. I have saved the data files so I can re-plot and subtract data and calculate the potential error for stratospheric forecasts of new GEFS for this period.

 

If we look at the old one, its a bit more stressed out, but still keeps the similar tone. its obvious how it just cant grab onto certain processes, like its new version.

 

 

And since we are talking about 1997 and 2002, lets see how today day stacks up to the same day on those two years. Kinda like an initial conditions check.  But (there is always a but, isn't it) of course we have to understand that the south hemisphere is just coming off its winter dynamics and is in its own story, which can and does affect the overall picture. But lets just focus on N.Hem for now. Analysis data is from GFS high res analysis on 05/10 12z. Temperature data is from ECMWFs ERA Interim, and GPH data is from NCEP R2. I need temperature data in better resolution than 2.5 grid from NCEP R2, so i went for the 1 degree grid reanalysis at ECMWF. Tho for geopotential height it doesnt matter if its 1 or 2.5, since there is some work needed to be done to make the GFS GPH analsyis grid compatible with ECMWF ERA Interim grids, and i dont have the time for it at the moment. Also i will do a whole week to week comparison sometime next time, instead of only day to day, but its ok for starters for now I guess.

 

The temperature pattern is clear, that this year, or today at least, we have a warmer pole the Greenland area and especially Asia/Siberia. Week to week mean comparisons should give a better picture.

 

 

GPH wise, the differences are actually interesting, with the common thing being more cyclonic energy over Greenland area, which is where the vortex actually is today. One more reason why we need at least a 7-day mean to make comparisons, and that should look quite tasty, since I am trying a new "look" of my graphics. 

 

 

 

Till next time. 

 

Regards

 

p.s. to Glacier Point: It is interesting just how different data we got. We are using different datasets in essence, and I do prefer the EPS over GEFS, even the new one, but I really love this, since we get to make a real-time test of the new GEFS.

 

P.s.2:

 

Forgot to add above, GEFS does have the wave2-ish pattern, coming mainly from the Scandi-Atlantic duplex ridge attack and of course the Aleutian low effects that GP has mentioned.

 

    

[Stratosphere Temperature Watch 2014/2015]

I was plotting some monthly data, and I came up with the plots of ECM32, from the end of August (Run 20 days ago). 

 

The purpose was to test how well it verifies with current obs.

The ECM32 Control run was actually very decent in its forecast. It forecasted a wave-2ish look, with a high over Asia and North America, and it got surprisingly close to the vortex minimal geopotential height and also the maximum of 314 over North America.

 

 

 

Temperature wise, it was also extremely good for this forecast period, with getting the max/min temp very close if not the same, and generally saw the warmer area of Pacific and Asia.

 

 

 

Its forecast further on is for a not so defined vortex. 

 

 

 

Similar as was the ECM32 ENS mean. it could be just some model bias, or the natural tendency toward climatology.

 

 

 

Another forecasting system I am testing, is the CMA monthly forecast, but it has a bias of developing a too strong vortex on heights and not cold enough.. 

 

 

 

There are also other monthly forecasts I will try to plot, but the data usually gets out with a 10-20 day delay. That is reasonable since we are talking about data which is usually not for free, being available for free under the S2S project. But since some forecasts got to 30 or 60 days, I will be able to get some forecasting power out of it.

 

The fact of the matter is, that if we compare the current model situation with the one about 2 or 3 years ago, its simply phenomenal how much more modelling is now available for the stratosphere, since we only had GFS then and ECM to 240h from berlin. Or actually, the data was there all along, we (I) just wasn't able to plot it yet. But times have changed and my quest to improve stratospheric modelling or at least promote it, is in full swing. I was (and still very much am) really trying to pioneer the modelling and visualisation of the stratosphere, since I do believe that it is being very neglected in this aspect, when compared to its actual potential/importance. 

[Stratosphere Temperature Watch 2014/2015]

True. Considering it is esentially just 21 calculations of the new GFS on a bit lower resolution, it shouldnt do too bad at all, since the new GFS proved to be an upgrade (strat-wise) over its previous version.

We are still waiting for the vertical resolution increase in the GFS tho...

[Stratosphere Temperature Watch 2014/2015]

A new post, but a different topic. 

 

A general comparison between the old and new GEFS. I intentionally made the color scale range of only 15K, to better show dynamics. 

 

 

 

And the actual grid, which is in this case the same, at 1°. but it is really obvious from the very start, just how much more dynamics the new GEFS has. Looks quite superb stuff.

 

 

 

And the vortex itself.

 

 

And this is the GFS grid, which is so fine, that its hard to tell which one is interpolated and which one shows the actual grid points. The one without contours (second) shows the actual grid points. Compared with the interpolated low resolution grid from InstantWeatherMaps.

 

 

 

 

 

 

 

As far as the stratosphere goes, we have some decent modelling power this year to work with. Especially cool is the GEFS upgrade, with the new GEFS now having the model top at 0.2mb, up from 2mb. its new configuration makes it the best free ensemble forecasting system for strat. One whole logarithmic scale model top increase, makes a big difference, together with improved resolution and up-to-date physics.

 

Personally, this season I will/can plot strat data from GFS, GEFS, JMA GCM, FIM9, NAVGEM, NAEFS, some monthly ens, and occasionally perhaps some from ECM ens. And some CFS or something like that.

 

Cheers.