Recretos

Since I am not at home these days, I cant provide any updates. But I will jump home in the morning after work (night shift) and plot the 00z GFS run to see where we stand. It takes around half an hour to render all the layers and all times once I load the datafiles, and it turns my RAM into liquid. 3GB is the very minimum. It took me around two hours the other day to make this layout so I can now just change the input files with a new run and the rest is automatic. I couldnt use the layout I use for reanalysis data, since ECM ERA data has parameters coded in a different way than GFS. I could only use the same outer grid. If I wont be too sleepy I will give an interpretation in the morning, hoping the 00z wont be too much of an outlier. EDIT 07:10am: My RAM wants to give out, since I am getting a memory error message. Too tired at the moment to try and resolve the issue.

No I havent applied any lag, because there is not enough dates. But if I apply a "virtual" lag, its not much of an improvement.

The past days which were under MJO 7, showed a bit poor relation to the phase 7 composite for January El Nino years. At some places even a negative correlation applies. The only bright points of correlation seem to be the Euro negative anomaly, which I am a bit sceptical to relate only to MJO, but I do give it some credit for its possible influence. And the next correlation point is of course the N Pac. low, but that is a clear ENSO signal, not really MJO on its own. Generally, I am personally not impressed with the MJO at the moment, and given how the whole trop-strat dynamics are soon going to get a re-configuration, I would not bet too much on it. But thats just me.

The MJO and ENSO will be temporarily put on hold, or to silence, from days 5-12, while the strat does its dance.

You cant have a canadian warming without warming, now can you? Its more of a matter of heights in this case, which makes it more personal between the trop-strat, than when temperatures are involved. Appart from the the upper strat, this "mole event" is height driven almost entirely. Thanks, you need to learn how to walk, before you can run.

That is a good idea. Tho I think 1 month is quite enough, since each month has its own climatological features, and you will have less work to do. If its not too much of a problem, you could take 5 or 10 random points, and just make the average result of all the locations. Would certainly give more merit to the end result. And again, great work.

This is kinda the thing that interests me the most in this case. I would say that given the overall stratospheric "driver" and support, this situation should remain fairly stable in the forecasts, but of course that is as far as the longwave goes. As for the details and the specific cyclonic areas, that is to be determined over time. I do expect just the general patter to remain more or less fairly stable, despite being in the operational runs at T200+. The situation is just simply different. This is not some tropospheric extension of a stable P.V. core, when model are trying to forecast which way the small amount of residual energy is going to go. These processes that are to happen, are massive energy movements across a very deep layer of the bottom atmosphere, and when the strat is in charge, the verification scores tend to increase. As always, time will tell.

First I have to say good work to Weather Master. That was a very fun post. But it is true that model verifications are made on a much larger scale, and not point wise, because that makes no real sense whatsoever. But I am sure you know that.

Well, I dont have any analogues, but I have something else. The full 16-day GFS animation. The usual rules apply: Top layer 10mb, mid layer 30mb, and bottom is 150mb. Color and 3D surface is Geopotential height. The animation is in full temporal resolution. The data files are quite huge so it took me a while. I will do a short interpretation after the video. So just of fun note is this: Watch at around 26.1 on, how the heights in the ridge wave at 10mb and 30mb start to increase and embed around the vortex, like a mole going underground, bulging the heights up. Looks really cool, and especially from the Pacific perspective in the second part of the video, you can really see it there. As for the brief interpretation: At the beginning you can see the shared elongation of the vortex via wave2 on all layers. And you can also see the connection of the wave 1 strat ridge with the tropospheric pacific ridge over N Pac. and W USA. That entire ridge starts shifting eastward and in the process it dispositions the strat vortex, with the troposphere core following soon after. Also in the process via wave response starts emerging an Atlantic ridge around 25.1, which seems to start quickly rising slightly on all layers, appearing as small bulging and violet coloring, creating a triangular patter shape. After that tho, the above mentioned bulging of the heights starts (the mole analogy), first at 10mb and 30mb, and with some small delay later around 29.1 it appears on the 150mb, as pressure from an Atlantic ridge, really squeezing and sending a deep longwave trough into Europe. Bottom line: As these video always seem to emphasise, the relationship between the strat and the trop is never ever linear. And in this case the strat does seem to have the upper hand, and the first end net result of this dynamics is the deep trough over Europe, evident on both GFS and ECM. And this situation is not some normal model calculation that lingers for a run or two before disappearing. This is largely stratosphere driven in the process, giving more security and weight to the situation with this trough, tho of course there will still be changes, especially how and when exactly this "mole" ridging drives the trough. On a side note: I have never seen such type of height rises, a.k.a, "The mole rise" as I will call it. Lets hope this "mole rise" will gives something interesting down below. And the tropospheric 500mb charts (Under the ECM Licence, WMO-essential charts can be shared publicly): The arrow is the Pac. ridge connecting up into the strat, and the M is our emerging Mole ridge. What to say. My days of watching only instantweathermaps individual charts are almost over. The stratosphere-troposphere dynamics are so complicated that its time to bring the strat analysis to a new level, which I am trying to pioneer for a while now. And this situation is a good example why I am trying to emphasise that geopotential height are also very important, not only the temperature. Regards. P.S.: Dont forget to spread the word that the strat is trying to send a deep trough into Europe.

If you are using GEFS for strat guidance, let me recommend the bias corrected version. That is pretty much the only GEFS version I ever use.

After almost liquefying my RAM, here is a new animation, showing strat dynamics of November and December. Top down levels are 1mb, 3mb, 10mb, 30mb, 150mb.

Could be an error by first estimate. I will check ECMWF datasets to see if its the same.

Well of course it does. It is all connected, tho the reflection is not 100% as you may know. But first, just to clarify that temperature question a bit more. Yes, generally and globally the temperature and temperature gradients do run the weather more or less, but in the strat where the changes are relatively fast and "localized" different rules apply. Besides, The pressure gradient force drives the wind in the first place. The warming could not reduce the wind when the pressure pattern doesn't favour it. If you would understand what me and Chiono replied to 22nov, regarding his question on wind reversal, you would not be questioning why the zonal mean was not reversed in the first place. As far as you second question on the trop pattern preventing an SSW. I made an animation of December 2014. 150mb level on the bottom, 30mb above, and the -40C isosurface. Not much to say, except how the troposphere seemingly drove the 30mb plane or the low-mid strat, like I was saying the whole time. The feedback from the strat was the reorganisation of the strat vortex which connected downward with the trop around the 12th. Around the 18th-20th you can see in the background how the pacific ridge intensifies, and with some lag the pacific wave 1 at 30mb starts intensifying. The troposphere went later with the help of this ridge creating an Atlantic ridge into a stretched mode by the emergence of the Atlantic ridge, around 27.12, which with some lag resulted in a wave 2 response at 30mb. Which is the Atlantic wave2 i was expecting for so long to do the ugly work for us, but the vortex was too sneaky. Now this is only for December and only shows the 30mb strat plane. http://youtu.be/r4keByIvhLs And next is a multi layer geopotential height analysis, which also tends to show how the mid strat was more in run by the trop., while the top strat has its own fun. The layers top down are 3mb, 10mb, 30mb and 150mb. http://youtu.be/53TkfxeWuD8 Regards.

As far as major SSW goes, anything less then a mean reversal of the zonal wind component is not accepted by me. Chio, where have you registered? I would like to try too.

Wind is bound to the pressure gradient force (among some other stuff, but thats besides the point at the moment). And that brings me to a specific point. I am on here for the third season, and even tho I am trying to make this specific point every year (and one page back actually ), some still dont get it. Write it down, print it, and put it on the fridge or a wall, as long as you see it enough times to remember it: "Winds reflect the geopotential height pattern, NOT the temperature pattern." Enough said

I assume this is the guy that usually gives the interpretations of the WSI charts. What he is trying to say, is that this warming is a 3 sigma event, which means the warming anomaly is 3 times the standard deviation. The reason it is so high, is because usually or climatologicaly speaking, there is the vortex core over that region with temps from -65 to -80 or less.

Interitus mentioned 3 examples above. 2012, 1986 and 1991. Lets have a look, shall we. 1986; A quick overview: The strat warming was there, but the real effect was nothing more than a North America ridge, to be exact. Tho that ridge had its role down the line, the real kicker in this animation comes from the E Europe ridge that is tropospherically driven, and that kicks up really strongly and even almost instantly leaves a good wave bumb at 30mb, and a bit later even at 10mb (not shown on this animation, but I did check it out). So via the feedback from this wave, the mid strat vortex bounces and aides in creating the cold spell down in the trop, where the vortex core moved more over Scandinavia via the help of the N pacific ridge. Near the end of the animation you can also see the trop vortex splitting via the new Atlantic/Azore ridge, and with some delay the 30mb level goes into wave 2 mode. All in all, this was a period of more tropospheric dominance rather than stratospheric. http://youtu.be/5O_rB99toYM 1991: Well. A big mess, to put it simply. On a quick glance again, the warming was again on the outskirts of the vortex, on the Asian side into Pacific, aiding a ridge formation over NA/Alaska. Meanwhile you can see a small ridge bumb being created at 30mb over Atlantic on 17. and 24.01, created by the Atlantic ridges a few days earlier. The Atlantic ridge persisted as the natural wave response, ultimately splitting trop vortex and soon at 30mb via wave2 attack. The Atlantic side of the strat vortex, free from the wave bounds quickly re-intensified, sending the response down to the trop, creating a strong cyclonic area, and via wave response helping to further intensify a menacing Atlantic/Azore ridge which bumped up and completely splited the trop vortex and up into the strat and the game continued. The rest is history. Of note is how the vortex "stabbed" itself in the beginning of February with that Azore ridge. http://youtu.be/XKeb00iUDoo 2012: Boy is this one an even bigger mess. To note before start, the vortex both in the strat in trop is more organised, and in at 30mb a bit less prone to effects from tropospheric Atl. ridging like the two examples above, which were constantly affecting it with every cycle almost. Starting out notable is the warming combined with strong pacific strat high, creating a ridge over NA, and via wave response an Atlantic/Azore ridge. The game continued with re-intensified strat Pacific highs pressuring the vortex and in the process pressuring and elongating the trop vortex aswell, and creating NA ridging, and down the line more Atlantic/Azores ridges. With the aid of further warming and this ridging game, the trop vortex broke into a wave 3 pattern, partially "disconnecting" from the strat vortex, but also soon reflected in the strat with a wave 3 pattern, which gave a feedback down again when the core in the strat resettled and briefly intensified and reconnected with the trop core, also in the process stabbed itself with an emergent Azore ridge created in the troposphere that went fast back up into strat, and also bringing the frigid temps into central Europe. After this, the vortex was driven by the trop processes. http://youtu.be/NZhGq9WGHwA summa summarum Pretty much a game of back and forward, like it always is, between the trop and the strat. These animations just simply show it more nicely. I created a brand new color table for the last video, to better show ridging. So please give me some feedback, which color table looks better. The one on the top two videos or the last one. Data is from the ECMWF ERA-Interim dataset. Regards. Edit: Added zonal height means to each year, which show, like the animations, that 2012 was the only year with a stronger strat involvement.

Chiono was short and precise. But I will also add my 2 cents. I am no physician by any means, so I can only try to explain it in layman terms. I am going to extend this so it also might be a simple crash course in strat wind zonality for others too. It is something I explained before, but oh well, I have a few minutes to spare. In laymans terms: The wind reversal simply follows the change in geopotential heights. Basically you have to look at geostrophic wind since the pressure gradient force is being changed due to the warming and geopotential waves which alter geostrophic balance, and you need to look at ageostrophic (wind) flux fields since Coriolis is in play, and ageostrophic flux is also induced by wave-breaking. In more laymans terms: When we say reversal of the wind, we do not mean exact 180° reversal from 100% west flow, to 100% east. In even more laymans terms: Every wind has momentum components. They are zonal (W-E) and meridional (N-S) component. For example, an exact SW wind has the same zonal and meridional component. This graphic show it nicely. The v is the meridional component, and u is zonal component. This means that this SW wind basically moves towards east and north at the same time, having a west and north momentum to it. And that movement is in meters per second. Which means that the wind that moves Xm/s towards NE, moves a certain %ofX towards east and a certain %ofX towards north. So the more straight that the wind blows, the stronger the zonal component. Or the more N or south it blows, stronger is the meridional component. So basically when we are talking about zonal winds in the stratosphere, or the SSW reversal, we are not talking about winds blowing in a straight line from west to east, or east to west. We are talking about zonal momentum component of the wind fields. Or more exact, the zonal average along the 60N lat line (average zonal component). In even more and more laymans terms: If we look at this graphic, we can see the zonal wind component at 10mb. I am using a random GFS run and fcst time. This is the average zonal component (what we look at also on FU Berlin) that corresponds to the map above. It basically shows the average zonal component along the latitude lines. We can see the reduction of average positive zonal component at 10mb between 30N and 45N. This does not mean the wind there is weaker in general, but if you look at the chart above, you can see there is a blob of negative zonal component (easterly), which affects the average along that longitude. Now why is there a blob of negative zonal momentum? Well, we can answer that by looking at the geopotential heights and zonal component combined. The answer is a stratospheric high, or more precisely a wave 2 high. We all know in what direction the highs and lows turn. You can see the zonal component is weaker where the isohypses are turning more N-S and stronger where they are straight. Note that this is not the actual wind speed, but only shows how strong is the zonal (west-east) component of the actual wind speed. And this is the actual wind and I added wind vectors: Next in line is ageostrophic wind flux, pretty much induced by wave breaking processes. Notable is the tropical wave breaking / wave dissipation. Now. Why is all this important? Well, because now we are going to look at an actual SSW and analyse it. And no better example as the 2009 SSW. The obvious is the zonal components and geopotential height at 10mb. The polar high is established, and there is naturally now the negative zonal momentum around it, since the wind blows clockwise around an anticyclone. And that is pretty much why the winds get reversed. In zonal average, the north hemisphere was nicely reversed up into the mesosphere. And that concludes this lesson on zonal wind. Regards p.s.: Since you are a future physicist, let me give you an advice on physics as a total amateur / non-physicist: At least by my experience, physics (or at least meteo physics) are 80-90% (if not 100%) pure logic. So just try and look at it in a simple way and dont overcomplicate it (like I sometimes do when I try to explain something ). I cant solve equations, but I understand why some things happen the way they do in weather. I am the kind of guy that has a big imagination, and that sees force vectors in a lot of things when doing certain stuff or observing something that involves physics (pretty much anything).

10 day temperature change. Impressive. Its meant to say 28/12/14, not 15. A typo.

If there are any more significant interactive effects of the trop/strat activities, I will be (or should be) able to track it down in my 3D wider spectrum analysis.

Ok. or he can just PM me the info. @John: It takes some effort to find the model info which is usually scattered. Vertical stuff is usually the hardest to find, especially model tops. But its not impossible to find. EDIT: Decided to add seasonal model info to main centers (besides ECMWF): added GloSea5 info to UKMO category edit2: Added CFSv2 to NCEP category. Edit3: Added JMA seasonal EPS model to JMA category.

@Chio: Thanks. I can add tge info if you give me the configurations? I am not familiar with the NetWx model, assuming its a house model? @ba: By my latest info it has only one leg with fixed resolution throughout the forecast period.

Lorenzo did a very nice listing of models and their configuration a while back. The list as good as it is, was getting a bit outdated. So since I have a minute to spare, I decided to update it, and create a new topic since the old one was, well, old. Here it is, my ultimate model configuration listing, updated in January 2015, so its the latest info of the lates. --------------------------------------------------------------------------------------------------------------------------------------------------------- UKMO - United Kingdom met office - UK - Global Model 17km Resolution, 70 Vertical Levels, model top at 0.01mb. Forecast duration: 0-144h for 00z & 12z. Run times: 00z, 06z, 12z, 18z. - UKV high resolution model 1.5-4km Resolution, 70 Vertical Levels, model top at 2mb. Forecast duration: 0-36h for run times: 03z, 09z, 15z, 21z. - MOGREPS Global ensemble forecast 33km Resolution, 70 Vertical Levels, model top at 0.01mb. Forecast duration: 0-7 days (12 members) at run times: 00z, 06z, 12z, 18z. + 0-9 hrs forecast (33 members for hybrid analysis). - MOGREPS medium-range ensemble forecast 33km Resolution, 85 Vertical Levels, model top at 0.01mb. Forecast duration: 0-15 days (24 members) at run times: 00z, 12z. - Global Seasonal Ensemble Prediction System - GloSea5 N216, 0.5°, 50km Resolution, 85 Vertical Levels, model top at 0.01mb. Forecast duration: 0–6 months. Run times: Once per month. 42 ensemble members 　 ECMWF - European center for medium range weather forecast - UK - Global Deterministic High resolution Model T1279, 0.125°, 16km Resolution, 137 Vertical Levels, model top at 0.01mb. Forecast duration: 0-240h for 00z & 12z runs. Run times: 00z, 06z, 12z, 18z. -Global Ensemble Prediction System (EPS) T639, 0.25°, 32km Resolution, 91 Vertical Levels, model top at 0.01mb. Forecast duration: 0-240h for 00z & 12z runs. 51 members (50+control run) -Global Ensemble Prediction System (EPS) - LEG B T319, 0.5°, 64km Resolution, 91 Vertical Levels, model top at 0.01mb. Forecast duration: 240-360h for 00z & 12z runs. 51 members (50+control run) -Global monthly Ensemble Prediction System - ECM32 T319, 0.5°, 64km Resolution, 91 Vertical Levels, model top at 0.01mb. Forecast duration: 0-32 days (768h). Run times: 00z every Monday and Thursday. 51 members (50+control run) -Global Seasonal Ensemble Prediction System - ECM SEAS T255, 1°, 80km Resolution, 91 Vertical Levels, model top at 0.01mb. Forecast duration: 0–7 months (monthly), 0-13 months (annual range). Run times: Once per month. 51 members (50+control run) NCEP - National center for environmental prediction - USA - Global Model - GFS (latest update) T1534, 0.25°, 13km Resolution, 64 Vertical Levels (planned to increase later to 128 levels), model top at 0.3mb. Forecast duration: 0-240h for 00z, 06z, 12z, 18z run times. - Global Model - GFS (latest update) - LEG B T574, 0.5°, 35km Resolution, 64 Vertical Levels (planned to increase later to 128 levels), model top at 0.3mb. Forecast duration: 240-384h for 00z, 06z, 12z, 18z run times. - Global Ensemble forecast model - GEFS T254, 0.5°, 55km Resolution, 42 Vertical Levels, model top at 2mb. Forecast duration: 0-192h for 00z, 06z, 12z, 18z run times. 21 members (20+control run) + Bias correction - Global Ensemble forecast model - GEFS - LEG B T190, 0.5°, 70km Resolution, 42 Vertical Levels, model top at 2mb. Forecast duration: 192-384h for 00z, 06z, 12z, 18z run times. 21 members (20+control run) + Bias correction -Global seasonal forecast system - CFSv2 T126, 1.0°, 100km Resolution, 64 Vertical Levels, model top at 0.2mb. Forecast duration: One control member 0-9 months run at 00z, 06z, 12z, 18z. Three control members 0-123 days (1 season) at 00z. Three control members 0-45 days at 06z, 12z, 18z. Total of 16 CFS runs initiate every day, of which 4 runs go out to 9 months, 3 runs (3 members in one run) go out to 1 season and 9 runs (3 members x 3 runs) go out to 45 days. FNMOC - Fleet Numerical Meteorology and Oceanography Center - USA - Global Model - NAVGEM T359, 0.5°, 50km Resolution, 50 Vertical Levels, model top at 0.04mb. Forecast duration: 0-180h for 00z, 12z / 0-144h for 06z, 18z run times. - Global Ensemble forecast model - FENS T119, 1.0°, 90km Resolution, 30 Vertical Levels, model top at 1mb. Forecast duration: 0-384h for 00z, 12z. run times. 21 members (20+control run) + Bias correction JMA - Japan Meteorological Agency - Japan - Global Spectral Model - JMA GSM T959, 0.187°, 20km Resolution, 100 Vertical Levels, model top at 0.01mb. Forecast duration: 0-84h for 00z, 06z, 18z run times / 0-240h for 12z run. - Global Ensemble forecast model - JMA EPS T479, 0.375°, 40km Resolution, 60 Vertical Levels, model top at 0.1mb. Forecast duration: 0-264h for 00z, 12z run times. 27 members. -Global monthly Ensemble Prediction System - JMA Weeklies T319, 0.562°, 60km Resolution, 60 Vertical Levels, model top at 0.1mb. Forecast duration: 0-35 days (840h), run at 12z once per month. 51 members (50+control run). - Global Seasonal Ensemble Predistion system - JMA Seasonal EPS T96, 1.875°, 180km Resolution, 40 Vertical Levels, model top at 0.4mb. Forecast duration: 0-210 days, run once a month at 00z. 51 members (50+control run). ---------------------------------------------------------------------------------------------------- Adding new GFS model topography for the end: So much for now. I will leave out GEM model, since I dont have complete data. Besides, who even uses GEM these days,... Dont ask how long it took to gather all the info. Feel free to point out if you find different information somewhere else. If you are interested in any specific model and you need info, just say it and I will do my best to find information. Regards

Model topography of the new GFS: A little bit of an improvement, looking at the old gfs, dont you agree?

Hehe well, all of you are doing a very good job monitoring, and if there would be anything really really worthy of posting, I would post it. So no, she is not getting in the way. Noteworthy is the 06z GFS with its wave 2 attack return. It is also on the para, but less intense, as one would expect. It has a vigorous wave amp. Note that this is similar to those FU Berlin wave graphs, but it is a slightly different "formula" that I use. It is interpreted quite similarly tho. The Scandi ridge in the trop actually connects with the Atlantic high all up to the upper strat. So this GFS run has at least one direct connection throughout the troposphere and stratosphere. And thats why this run is noteworthy. Its actual verification or probability of occurrence is besides the point. It also nicely shows the connection between the trop and strat. http://youtu.be/mcqbbn1A0rk Every graphic usually has its source written on the graphic. Either a logo or a website or something indicating its source. And yes, my graphics are privately made by me, using freely sourced data or some by private sources, like JMA strat output to 264h. Speaking of JMA, its still going quite in tandem with ECM, for now reassuring its position as the second best strat model. Regards