Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by Recretos

  1. There is not much hard data to compare the two as far as I know. There likely is data in existence, that was used during the development, testing and fine tuning phases for the new FV3. But those are probably mostly internal. At least I have not came across any comparison data. On a general note, I kinda feel an anti-FV3 agenda in this thread. Not sure why, because FV3 is a very capable solver. It is actually appropriate for the strat, as a dynamic core. Has anyone noticed when looking at the strat charts, they look much more realistic? Especially the temperature forecast. You can actuall
  2. Legacy is the GEFS v10 version. As it says on the site, it is the pre-December (v11) version. That means it is the pre-upgrade version. It has 42 vertical levels and model top at 2mb. The v11 version (december 2015-present) has 64 vertical levels and model top at 0.2mb. So the legacy version is underqualfied to be a useful strat forecasting tool. At least in the modern times.
  3. Just a side note: The GFS//GEFS suite is running without the Aura Ozone Monitoring Instrument (OMI) data in the inputs since 15th 06z run. I have seen a warm bias in the FI before when this dataset was not in the initials. Not saying that is why we are seeing the warming. The warming is likely legit, and already seems good on the 12z as well. I am just adding a note that there is a lack of ozone data in the initials for future reference.
  4. For all that watch the zonal winds. Let me urge you to look at the geopotential heights more. At least as far as weakening/strengthening trends go. Because as the polar vortex cries for help, you might gonna see an increase in zonal winds, or at least no weakening, while the vortex is actually starting to lose the battle. It was seen on many occasions. This goes for GFS and GEFS. As the vortex goes under heavy load, via wave 1 or 2, it gets contracted, so one side of the vortex can get "compressed" which can amplify the wind speed as the pressure gradients tighten. If one looks at a 60N lat U/
  5. Of course. Trop-Strat coupling is not such an easy matter for the medium range, harder for the extended range, and very hard for the seasonal range.
  6. Well, I can share the fresh ECMWF seasonals for strat, which I just made. It does see the warming, and even escalates it towards January! For December it has the wave 2 pattern, with the warming we are seeing now in the FI GFS. January tho, it drops the bomb. Not sure it has an actual SSW, but this is definitely a weak vortex event. Not much recovery is seen in February, but that is a long lead time. UKMO Glosea5, is quite different as far as geopotential heights go. It does not have an obvious weak vortex event. These are monthly averages, but the ECM forecast i
  7. Yes, I was aware of early 3D animations, which served as a great guideline. I was not aware of this specific paper (which has great looking 3D plots), but all I ever found, was in research papers. Which is why, as I said above, I consider to be one of the first (not necessarily THE first), that did it on an operational/forecast basis, rather than research-case study examples. ?
  8. Well, any type of copyright would not be very scientific. ? Besides, the data is not mine, and also the software is not mine, so its not really possible to copyright just the method, even if I would want to. On a side note, I just noticed that after the last upgrade of the GFS to the dynamical core FV3, the model levels in the output grids got an addition of the 0.4mb level. Not a lot of forecasting power in it, but t will be interesting to see where the "lid" of the warming is. The graphic I made shows the 16-day change in temperature, a zonal mean. The warming is obvious and anomalous,
  9. Well, I did extensive research back then, and true 3D simulations of the polar vortex were nowhere to be found. There were images of a 3D polar vortex, but in a 2D (cartoonish) style, and specific to research papers individually. That was the very reason I decided to research this field (3D simulation), and essentially (by my knowledge at least) being among the first (if not the first) to create operational 3D simulations of the polar vortex in 3 (4) dimensions (time) from operational model forecast data and reanalysis data (ECMWF ERA-Interim, NASA MERRA-2). That was already in 2013 I think, s
  10. As I also stated in my article on another site, as long as the pattern configuration with the Siberian high/Aleutian low persits, there should be more of this on the way, increasing the potential for a proper SSW by the end of the year. On a seasonal note, I was in a hurry, but managed to make the 10mb temperature anomaly charts for december, from ECM and Glosea5, from October runs. Will be interesting to see the November runs. Its amazing how different they are despite pushing a similar tropospheric agenda. Tho these maps have no practical use in forecasting, it is interesting to se
  11. The circulation pattern overall is supportive for further wave activity in the stratosphere. At least if one trusts the likes of weekly CFS and JMA ensemble models. The Siberian high is a good sight to see for now, if we can also get decent lows in the North Pacific. I have gathered more thoughts in another place. An unusually strong early attack on the Polar Vortex expected! Sudden Stratospheric Warming event is not excluded in the next 30-60 days – Severe Weather Europe WWW.SEVERE-WEATHER.EU -spread the word- We often hear the term “p
  12. My new animation sequence, showing the nice wave2 pressure from GFS in late terms, which are strongly tropospheric forced, by the Alaskan and Atlantic ridges.
  13. I think it should be around 10-12th, and similar for UKMO.
  14. Great job Lorenzo! If I can make a suggestion: Reduce the amount of lon/lat lines to 90x30 interval, since for this type of data it is not as important and can sometimes even be a distraction. Also, you have an important flaw in the color scaling. You dont have zero centered colour scale, which is critical when plotting anomalies. That means that the colour scaling is centered on zero, so all negatives are on one side and positives on the other. Basically you must have 0 value in the white middle. One plot has colour bar from -80 to 60 for example, which means that the zero lands in
  15. Not really. The configurations and systematics of the CFSv2 model, are not really that appropriate for stratospheric forecasting beyond the normal operational range (15-16 days). Anything beyond that should be ensemble (breeding) only.
  16. Yes its relative vorticity, but I am not sure about the value, I think around 18? Since it was done on the fly... Tho for SSW reanalysis from ECMWF ERA data, I rather use absolute vorticity. Its much more clean and gives a better impression. But the new ERA-5 data only has relative vorticity for now, so I will try that out since it has really nice resolution for a reanalysis model.
  17. I dusted off my 3D plotter today and did a first new season test of the 3D vortex structure, from todays 00z GFS data, for 384h forecast. This is the whole vortex core 1mb-150mb, so you can track how the whole vortex is tilted and where it connects to the troposphere. It is just to show how the GFS develops the core and links it downward, since this is deep FI, and beyond 240h nothing is really used as a forecasting tool.
  18. Here is my updated plot with SSW markers included. The SSW markers on top are colour coded by their respective ENSO phase.
  19. Not to be a party breaker, but 5K is like a normal core oscillation. Can hardly call it a warming event really, let alone count on it to bring colder trop weather. And this is a 50mb level, so its even less significant.
  20. Thank you for your work Lorenzo. I am tho a bit confused with the first plot. The enso regime from 1990 to 2001 seems quite off, being all in negative? I did one a while ago, I guess I should update it and add SSW dates. Black line is ENSO.
  21. And to further elaborate, below is my own version of this graphic from @Interitus , after I apply my "Recretos' 30-second full-HD optimisation procedure". It really is that quick once you know what to do. Colour is of course just a matter of taste.
  22. Yes, the monthly mean of 50-members at a 3-5 month lead time is diluted enough to show near climatological pattern to the untrained eye. Anomalies are needed to reveal actual forecasted deviation, which must be calculated from the model hindcast climatology. Panoply is a great tool. It needs some getting used to, but its a great begginners tool. I use it on a daily basis for all kinds of plots. It takes some basic trial-and-error, to learn how to optimize the graphics and to create strat-specific plots. Tho it is not for mass production of maps. I use a whole different beast for that
  23. As far as the stratosphete goes, the Atlantic has not gone wild because there is nothing wild in the strat to begin with. 25m/s is a good start, but in mid winter that speed would be considered a weak/very weak vortex. The vortex core 10mb pressure height dropped below 3000dm just a few days ago, so its not yet that wild beast that we all know and love. Or hate, for that matter. In the near future the trop. Atlantic wave will help to push the core around a bit, keeping a lot of the vortex' core energy away from the Atlantic sector, pushing it more into the Siberian region.
  24. As promised, I finally managed to complete the website with ECMWF seasonal forecasts, including the stratosphere. I will add Glosea5 around its November update, so it will be fresh when it goes live. http://www.severe-weather.eu/ecmwf/
  25. I noticed talks of QBO, and decided to look at it via GFS, and it is in the positive values already, so October mean might end up near or at positive values. I made one HD graphic of QBO progress at 10mb and 30mb from GFS data.
  • Create New...