Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

132 Excellent

Profile Information

  • Gender
  • Location
    Wickford, Essex

Recent Profile Visitors

5,474 profile views
  1. Paul S Bumped into your mate Pete Scott at Stansted last night, he showed me a nice spot (thanks again, Pete) for shooting CG's against the ATC tower - nothing hit to our disappointment Still, nice to blow the dust off the camera and see some UK lightning. Not the best pic (but best of the night), I'm a bit rusty this year...
  2. Hi Paul Is your 4km running a CU scheme, and is it nested or a straight 4km grid?
  3. Some of the forecast soundings do show some impressive SRH values with some rather large looping hodographs. I'm assuming this is due to the low level south-easterly flow (possibly 35kts at 925mb on some output) with veering winds above. Its almost a shame, as the boundary layer looks quite stable and with these storms rooted above, but a surface based storm in that environment could be very interesting.
  4. They could well be two different model grids placed over different geographic regions, rather than one big 2KM grid. Probably a good example of how these models are quite sensitive to the initial conditions they are initialised with and their model configuration, physics, domain sizes etc.
  5. Modellzentrale's 4KM looks better for the south east, though if you follow that model closely it can get a little excitable with convection at times. Doesn't mean its necessarily wrong this time, but I guess we will find out later! The ManUniCast model would be worth checking when it rolls out.
  6. Couple of charts from an 18z 3km GFS init run. The model appears to destabilise the edge of the plume again, its been fairly consistent with this approach recently but its no more accurate than any other deterministic model run. It does put a bullseye over central southern areas at midnight - make of that how you will Note the big 'bite' missing from the northern edge of the elevated instability, this is likely to be where the storm is eating into the instability and it appears to ride along this moving north east out towards the Wash area. The supercell composite parameter is almost off the scale, though that could well be skewed as its a composite of a few different parameters and its also a fixed layer (not effective layer). Interesting to see that there is still disagreement between some of the higher res GFS models...Something, somewhere would be my guess
  7. Re: the Skipton storm. Max 2-5km updraught helicity product attached from a 3km 06z run this morning. You could argue that the convection was modelled too early but this does appear to suggest an environment that could support a rotating cell in Northern England this evening.
  8. There are a few other free to access WRF's ManUniCast 4km http://manunicast.seaes.manchester.ac.uk/index.php ModellZentralle 12 & 4km (just about covers south east England) http://www.modellzentrale.de/ RipWRF 3km https://ripwrf.info (this is my own project which has been in development for about a year) All ARW cores initialised from GFS (as are the NMM's at Meteociel/Netweather etc)
  9. There is some very interesting output, for tomorrow afternoon, modelled from the 06z 3km ripWRF. This has some notable surface gusts at times underneath the ongoing convection, the most striking is the 85mph surface gust modelled from the convection near to the Lake District. Would be interesting to see if that verifies - and fits nicely with the Estofex forecast.
  10. Couple of charts from some of the other WRF-ARW convection allowing models... The 3km ripWRF output for the overnight hours appears to destabilise the edge of the plume later this evening with elevated convection moving north-east from the Midlands. Decent signal on the max integrated column graupel product - suggestive of hail (n.b max possible value) The ManUniCast 4km WRF also looks quite similar with its simulated reflectivity, though both models are initiated from the GFS (different physics schemes however) - if thats wrong, well, these may well be too!
  11. Not exactly, the WRF model has two core resolvers available, one of which is the NMM core, the other the ARW core. The NMM core is favoured by some as its generally a lot quicker to run when compared to the ARW core, useful if your forecasts are time sensitive etc. The ARW might be considered to be more accurate but it can be slower to run. The US HRRR model uses the WRF with the ARW core, I think Meteociel use the WRF with the NMM core, last time I used the Netweather high res suite, that was using the NMM core. They are all run independently of each other and can all be ran with different physics options/at different resolutions/nests-no nests (the ARW core actually has a lot more options in this regard) and might not all necessarily 'agree'. Getting a bit off topic now but I never know which particular model people are referring to when they mention the NMM/WRF etc...could be one of many! Bit more info from the WRF site http://www.wrf-model.org/index.php
  12. My guess would be that's a model initialised from the GFS, the site is a bit sparse with regards model setup and configuration etc, though its free to access so can only assume its using GFS. Unless they have deep pockets to use the ECMWF data...who knows. If its GFS data, then that's not available for a WRF to initialise until approx. 4 hours after the GFS run (e.g 16z for the 12z GFS). The current charts on their site could well be run from the 12z GFS, it just takes that long for their model to run after the GFS data becomes available.
  13. Out of interest, which WRF variant are you referring to? Most of the WRF's are run at the mesocale with their initial and boundary conditions coming from a global model, such as the GFS. If its a model initialised from the GFS then its not unusual for the WRF to have similar output - depending on model resolution, physics etc.
  14. Have you had a look at WeeWX? I think that will run on a Raspberry Pi and it does list a Fine Offset 3080 for supported hardware (if that's your model). I've used it for a while through a virtual machine and it's been very reliable. http://www.weewx.com/
  • Create New...