Jump to content
Will it snow? 7 day forecast Weather Radar Will it be cold?
Snow?
Local
Radar
Cold?

Recretos

Members
 View Profile  See their activity

  • Posts

    496

  • Joined

  • Last visited

  • Days Won

    5

 Content Type 

Everything posted by Recretos

  1. Recretos
    3D polar vortex, based on relative vorticity. From 100mb to 1mb. It is nicely seen how it is split in the lower strat. Data from GFS 12z at 384h.
  2. Recretos
    I was playing around with its files earlier, and its actually pushing for an SSW in Jan. So what you are seeing is actually SSW response.
  3. Recretos
    Well, the CMA GSM is actually based on a revision of a ECMWF spectral model, with initial data from NCEP GSI. Probably has one or two custom modifications. But still, it has configuration of T639, with 60 vertical levels and the top at 0.1mb. The output grid is 0.25, so that is why it looks high res compared to the 0.5/1 from GFS. It has a legit configuration for resolving stratosphere dynamics, AFAIK. Tho I am not familiar with any verifications for the strat.
  4. Recretos
    Actually the main trough axis does appear more neutral/positive tilted, with the low core slightly ahead of the main trough axis. I would call it neutral/positive tilt.
  5. Recretos
    Well. according to actual data for day6 and day8 forecasts, its 00z and 18z that are statistically the best so far. On the 12z multi model stats, the para GFS (PRX) was actually not that far behind ECM, and better than normal GFS.
  6. Recretos
    Actually, it looks/was weaker, because it has more meridional momentum component and lower general angular momentum. And as far as strat goes, it was also weaker, with a wave2/3 event in progress, ultimately down the line leading to the 1963 split SSW.
  7. Recretos
    Yes, its CFS, but lets face it, we all like to look or at least peek, at longer ranges.
  8. Recretos
    As I was also saying to Matt and Chiono on twitter, it was actually the FNMOC control run that was the first to spot the developing wave. GFS was next, and GEFS behind it. GEFS has weird temp. placement, but it is normal for its resolution. Regards
  9. Recretos
    Just to let you know, I now have a source to JMA Global spectral model data, which I will start plotting when/if this current warming gets into its range (264h). It has output up to 10mb, which is enough for basic monitoring. Otherwise the model has configuration 0.18° (TL959), with 100 vertical levels and the model top at 0.01mb. So it is quite on the level of ECM, and has a very good configuration as far as strat monitoring is concerned. Will be interesting to see how it performs. The testing I did for 240h, has it very similar to ECM Berlin output, so it does appear to be quite decent for now in these "peaceful" days. As far as models go, the more the merier. Regards
  10. Recretos
    Thats not even a warming, let alone a good location.
  11. Recretos
    Well, the para GFS is starting to play around with something, which could be just a random thought. We will see generally how and when will models pick up new dynamics, which will happen eventually in mid and late December. Since the current forecasts are mainly straight-forward, it will be interesting to see which model will start to pick up on any real new dynamics and how accurate and stable. Cheers.
  12. Recretos
    Matt has beaten me to it again, but I will finish this quick analysis nonetheless, since I was already mostly done. Despite the activity waning, which was to be expected, the damage has been done, and the full vortex power like last year, was denied or rejected or prevented, for the time being. As I have said a few times so far, the troposphere has the more dominant role this year, as the general stratospheric state is for now disorganized and not compact and "columnised". Due to all the wave activity, QBO phasing and strong tropospheric forcing, the vortex has a much lover angular momentum than last year and a weaker core. It has (had) more meridional momentum component and lower zonal momentum component. Due to weaker core in general and the pressure gradient force, the polar jet is weaker along with the pretty much entire vortex (or vice-versa, as you wish). With the vortex in such a state, the strong tropospheric forcing comes to more importance and basically dominates the lower strat in a ratio around 70-30 for the troposphere. Higher up, there is this "surf zone" in the mid strat, while the upper strat has had its own dynamics which never down-welled quite far. And that was what even the normal GFS was signalling in the mid range. Unlike last year with a very organised and complex vortex with great angular momentum and strong core, which downright dominated the entire atmospheric column, and we know what the outcome was. Ok, enough plain text for now. These are height and temperature anomalies for last year, for the 7/12, which is the date of current 240h forecasts. We wont go beyond that for known reasons. Very organised and focused anomalies, and this is only one level in the mid strat. Now if we look at the entire column comparison to last year. It is getting clear how the re-organisation of the vortex is more limited to mid strat in the first phase, as it takes a bit longer for the wave activity to back off in the higher levels. And it is more bound to the core region. So pretty much a normal mid strat reorganisation in this phase. The height and zonal wind comparisons speak the loudest how disorganized the vortex column is compared to last year. And the temperature of course. Nothing too wild in the strat for now, and neither to be expected in a reasonable range. I wont go beyond 240h, because that is where the model biases come more into play and that is a problem when doing day-to-day or year-to-year comparisons. 240h is a generally good range. So much for now. Regards. Edit: Adding the full vertical profile reanalysis from last year, from MERRA.
  13. Recretos
    Yes, I was kinda expecting something like this, as I was trying to say in my "mysterious" posts, as we move into December, for the vortex to try and re organize and start moving, but not achieving its full potential, and the wave 1 kinda present at all times, tho not as strong as in the current days. This is just another step on the road. If this transfer will slowly complete, (perhaps not so fast as ECM has it, but still), I expect new waves to slowly re-emerge, as I was saying then. I would also expect the Atlantic wave to slowly start emerging from its long slumber. Regards
  14. Recretos
    Well, yes. Tho the resolution is an important part of the model, its partially inferior to its physics. You see, you can have a global model with 5km resolution, but if its physics are a bit off, it wont do any better than some global model with 20km or 30km resolution with better physics. Its not how many points you have, but how quality calculations can you make on each point. Tho of course in this day and age the physics are comparable, but in our situation with the GFS, it got a resolution increase together with some physics change (grid from Eulerian to semi-Lagrangian), so the verification is not that straight forward, and it takes more time. As far as the strat goes, the para does seem to be generally less biased than GFS, at least at 240h. And later on, it is generally a bit better, but its hard to say without serious dynamics in the forecasted range. In summer time, the normal GFS was better for the strat, but in the winter time, at least last season and this year so far, it appears that para was a bit better. Tho I think the new (latest) cycle is PRX, not PRHW14, but its pretty much a similar thing. We will see in time. Regards
  15. Recretos
    I am quite busy in the last days, and I am not even at home, so I cant make any plots for you. But I called home and I gave my sister a crash course in modelling and plotting and she made a graphic, tho not as perfect, but still, the main thing is visible. This is 384h forecast of the para GFS, from 23.11 00z. its the 10mb GPH.
  16. Recretos
    Yep, and some more here:
  17. Recretos
    Not sure how to describe this, other than "chaos" or "mayhem". As fun and positive as it is, it is kinda starting to be a very slight thing of concern for me. Why? Im glad you asked. The whole stratosphere looks so badly "disorganized", if you also look at waves and fluxes, that its almost unreal. Its now kinda looking closer to the para GFS. The upper strat is really weak, and the whole column is out of sync. I would almost wish for the vortex to get a small part of its act together and step up to the game, because what this forecast has is pretty much a mess. The upper strat completely in its own world getting its face smashed in, the mid strat kinda holding as it should, acting like a "surf zone" between upper and lower strat, while the lower strat is getting dominated by the forcing of the trop patterns. My point is, that in such a state, you perhaps wouldnt have much luck with an SSW, since the downwelling could not be completed when there is nothing to grab onto in the column. Not to mention trop counter forcing. In such a state, all that remains is watching scandi. wave 2 breaking and smashing into low/mid strat. But hey, these are only forecasts and my speculation. Tho we should perhaps get ready for some scandi wave2 later on as we go on into December, as I was saying in my "mysterious" posts. Regards.
  18. Recretos
    The troposphere looks very "resistful" for now, also pulling low and some mid strat to its own gameplan. The upper strat and the waves better step up to the game.
  19. Recretos
    One in a hurry.
  20. Recretos
    Not really much to add to what Matt has said. This current upcoming split wont really do any special good, troposphericaly speaking, tho it is very welcome strat-wise. Later on, the main activity will be confined to the more upper layers of the stratosphere, and it really is very decent activity. Still, not much out of it for the troposphere (naturally), but as long as things are stirring up there on the top, the vortex cannot be one large single entity as it wants to be. There is also some slight disconnect between the upper/mid/low strat layers. The reason is that the troposphere is slightly more on the dominant side as far as lower strat goes. So we will need that upper strat activity to start taking over more than just the upper layers, combined with more constant dynamics. Just one wave of activity isnt going to cut it. We need constant battering, to put it that way, and some help from the trop. patterns. Now there are of course many positive points, like the temperature and wind profiles. Here is a comparison of the ECM forecast, and analysis of last year. I use the both hemisphere projection, because there is a little glitch in the software and I cant plot just one, but nonetheless it gives a nice perspective. Now for those of you that cant see those "positive points" right away on the graphics above, I have made three new graphics just for you. It is the difference between this year and last year, centred on 27.11 as far as ECM forecasts go and in a relatively good range of skill. It is the difference on this date between this year and last year. Data from this year is based on the Para GFS forecast (it is perhaps a bit different than ECM, but it gives a good general idea nonetheless), and data for last year is based on reanalysis from ECMWF ERA-Interim (accessible from the data portal me and Matt were talking about). It shows nicely just how weaker the mid-upper strat vortex really is, and as far as temperature goes, it has a pretty non-hospitable environment. It looks really cool (no pun intended), at least to me, and its a good indicator how different things are up there. It is interesting to note how the QBO phase also has its own temperature profile. Wind-wise, the QBO difference is visible right away, since the phase is different to last year (we are in a QBO-, so that is why the negative difference relative to last year, and the positive above it). The polar jet is quite weaker at this point, because of the wave onslaught. And the geopotential height difference, which also tells its own story of the vortex strength and the wave onslaught. I used the MERRA dataset for this one, because the ECMWF ERA gpm data is not coded in a compatible format with GFS, tho wind and temperature are. With this graphic you kinda get the fell what those wave graphics from FU-Berlin might mean. Here is also interesting to note how the QBO also has its own geopotential height profile. Not to mention that the ECM forecast has an even weaker vortex (slightly) and a warmer wave as the GFS. I really love playing around with data. Hope you like it. Regards.
  21. Recretos
    Interestingly, the Para GFS is a bit different than the ECM this time. Will be interesting to see who emerges victorious. And some wave 2/3 fun and games in the FI. Well you cannot really specify areas, but the files have specific parameters and vertical layers in them, so you just chose the right file. Every file has the full globe coverage. You can start here regarding files: http://www.nco.ncep.noaa.gov/pmb/products/ Regards
  22. Recretos
    Well if I explain this from experience, since I use and plot the raw data from the model:The parallel GFS does run at a higher resolution, 13km. The output grids are tho different. You have "a","b" and "full" grid outputs in GRIB2 format. a files are available in 0.25° (original), 0.5°, 1° and 2.5° grid outputs. The point is basically in the file size, with larger the grid spacing, smaller the file size, and higher speed, etc... Depends what details you need. For strat I use normally 1° grid or 2.5 if I am in a hurry. b files are available in the same output resolution, except there is no 2.5 grid output. Now the difference between a and b are the vertical levels and parameters. a output has certain layers from surface to 10mb, and b output has the missing ones that are not in a, and the upper strat from 10mb to 1mb. The point is again in lower file size if you need only specific parameters (b has less parameters than a) and specific vertical layers. Both a and b go to 384h in all degree grid outputs. The "full" output goes from 0 to 240h, and has full vertical layering and max resolution (0.25°) and all parameters. But the file sizes are of course bigger and the loading is slower if you run it on your website, etc... Its not a big deal if you only download specific files, since one file (each forecast hour has its own file) of the full output is around 190MB. The same a,b,full system is applied to op GFS and GEFS. Basically you can only use specific files or combinations and depending on the detail, parameters, etc required, you can use one file type that fita and you have faster loading with smaller file size, and NCEP saves sone if their bandwidth. And its practical for people like me that need specific vertical layers, specific times and specific parameters. I dont have to download full outputs and gigabytes of data. Meteociel by the look of it uses 1° grid for the FI in old GFS and 2.5° grid for the parallel GFS. That is quite a difference, but for a large scale like the stratosphere its not too big of a big deal for the FI. I hope I cleared a thing or two. Kind sunday regards, Andrew.
  23. Recretos
    Yea I forgot to mention, you have to register first, and all the data is in the gridded form only, so you have to make graphics yourself out of that data. Its a gold mine of reanalysis for me because that is pretty much what I love the most, playing around with data and making plots.
  24. Recretos
    Past years no, GFS has a one month archive for the full vertical profile up to 1mb as far as I see. But I am not sure for ECMWF. But there is reanalysis data for the full profile up to 1mb and some in the mesosphere from ECMWF. You also have the great MERRA dataset or many others. I dont really see the point of looking at past forecasts, because each season is a bit different and "tricks" the models in their own way, so there isnt much you cant get out of it, as far as skill is concerned.
  25. Recretos
    Too bad I dont have enough time to make full plots of the Para GFS 12z, but it is enough to just say that despite it looking horrible for the trop, in the upper parts of the strat it starts a wave 3 assault on the vortex in the final FI parts. Pacific W1, Atlantic W2 and Indian W3. Edit: Oh what the hell,... Dont have the time to edit text stuff, so just for the record until the new run removes it (or not). This is 10mb 384h forecast of the parallel 12 GFS. This could be a start of a good story if it would be to materialise. Oh well, the darn FI,... Pretty much the most defined wave 3 pattern example you can find far around.
×
×
  • Create New...