Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

204 Excellent

Recent Profile Visitors

3,020 profile views
  1. Yes sorry, there are some missing files on the NCEP servers, they're coming very slowly. I try to get and draw the ones that are present. The files are correct though, no data corruption.
  2. The german server delivers new files every 9 mins, hence the small stops between updates.
  3. Seems to be a temporary bug because now I see already ENS files but they're blocked Hopefully they will be unblocked in 20 mins...
  4. Yes something is fishy there This shows that there is a delay for us, or they improved the computation speed a lot. I think we usually have a delay because the datetime of the files are the "usual ones". We will see if it's the same for ENS. Yesterday they changed their FTP server, so probably something didn't get implemented correctly. I think the delay will be readded when they will notice it has disppeared
  5. Yes weird I don't understand. The files just appeared out of nowhere with even a future date.. The data seems to be correct though. It doesn't look like today 00Z or yesterday 12Z.
  6. Yep, sorry about that, I'm regenerating charts.. But maybe all data won't be there. 72-144h data Z500-PRMSL is ok now. For some reason, I got data from last month...
  7. The information might be outdated in the app. GFS is displayed at 0.25° and GEFS at 0.5°. Moreover there is a difference between the actual resolution the models are run with, and the data resolution that we get to display. The models are not ran on traditional lat/lon grids. the GFS model is currently computed with a resolution of about 13km ( T1534 , spectral truncation ), and we get it as 0.25° in usual grib files. the GEFS model (control + 20 perturbations) is currently computed with a resolution of about 33km (T574, spectral truncation) until 192h and 55km (T382) till the end, and we get it as 0.5° or 1.0° in usual grib files.
  8. Hi Steve. Probably not. We don't have access freely to the 168 data. UKMO data is already very sparse. But if i ever find it, it will get on the site right away. Moreover it is weird, the MetOffice site mentions only forecasts up to 144h here : https://www.metoffice.gov.uk/research/modelling-systems/unified-model/weather-forecasting Sylvain
  9. There were some NCEP server issues tonight causing delay and weirdness like that. I will rebuild all now Not many perturbation data are faulty, Most is ok.
  10. Hello, sorry for the late answer : unfortunately we don't have access to this data freely (and it costs > 100000€ yearly) . Whenever it is available, it will be available on the site the day after
  11. By the way, you can use this zoom to see the UK snow extent ( resolution 1km ) : http://www.meteociel.fr/observations-meteo/neige-glace.php?region=uk (it is exactly the same data as the previous NOAA map you have shown)
  12. SylvainTV

    Model Output Discussions 12z 01/09/2016

    JMA stats are available for the 12Z runs, the 0Z run does not go far enough : http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/12Z/ Besides, I have added again the NCMRWF model, they now run the MetOffice Unified Model for their model (but not the same assimilation) once a day. Not many fields to look at though. http://www.meteociel.fr/modeles/ncmrwfe_cartes.php
  13. SylvainTV

    Model Output Discussions 12z 01/09/2016

    Hello, sorry can't do that, I have no access to parallel ECMWF data at all . It seems only customers have access to it. I think he forgot to put a restriction on this page, ECMWF parallel data is usually even more restricted than operationnal data so it shouldn't be accessible freely, ( and the Op is actually available there too ... ). Maybe it's just a sample
  14. SylvainTV

    Model output discussion pm 13/01/2016

    It is not the good GFS run data that is plotted on this Wetterzentrale graph (it was probably the 00Z run). Note that the location taken for London is "Lat : 51 Long : 0" which is 50km south of London. They use the nearest point of GEFS 1.0deg data. The Isle of Wight thing was probably when GEFS data had a 2.5deg resolution. Same thing for the GFS op resolution. It had a 2.5deg resolution in the > 192hours part of the run until the last GFS upgrade. (On Meteociel we now use GEFS 0.5deg / GFS 0.5deg data for the graphs and interpolate between grid points, so the values are supposedly more conform for London)
  15. SylvainTV

    Model Output Discussion - 05/01/2016 18z onwards

    Oops I have to fix that on Meteociel, it's so rare that I didn't even bother drawing iso lines after 1055hPa. Sorry about that ! However remember that the Greenland altitude is quite high, so that Pression at Mea Sea Level is quite "virtual" and does not mean much.