Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation


Profile Information

  • Location

Recent Profile Visitors

912 profile views
  1. Not impossible, generally dew points are too high in GFS. So in reality I'd expect the LCL (convective cloud base) to be a fair bit higher than shown by GFS. A higher LCL reduces tornado risk.
  2. Elevated convection in the morning across the SW and Wales where the in the areas where mid level cloud abundant, here CAPE around 750Jkg, so lightning and heavy precipitation the main hazards. Although some mid-level cloud extends further east towards southeast England drier higher level air should prevent much elevated convection here. Pic 1: 10/0600 UTC Mid Level Clouds from EC Pic 2: SkewT to from GFS valid over mid Wales at 10/0600 UTC This elevated convection then extends north as the upper trough which assists in de-stabilising the zone pushes northwards, it is then the question whether the cap that exists to the E of this zone can be broken and release, always a close call in these events but a combination of surface heating and convergence (including from outflows left by the elevated convection earlier in the day) could combine to release some isolated but extremely violent thunderstorms (>2500Jkg of CAPE, with very large hail, strong winds, frequent lightning and incredible precipitation rates possible). Difficult to judge area at the highest risk from these....but I'd guess a zone from Wilshire towards Greater Manchester....and these could take well into the day early evening to fire (if they do at all). Pic 3: SkewT to from GFS valid over West Midlands at 10/1800 UTC
  3. Agreed it's very complex statistics. The main point I was trying to draw out of it is the broadscale synoptics (especially mid-upper troposphere) are the key measures via which NWP centres compare each other against. Generally NWP is not optimised for any one thing (surface temperatures, snowfall accumulations, tropical cyclones, snow depths, wind speeds) and we must look at the individual fields with that in mind. It's all a compromise, with the main aims of development being to perform well against others NWP models in the bench mark statistical skill scores (which correlates with give broadscale forecasting). Although I suspect the GFS may sacrifice this slightly to better resolved some of the non-linear features such as cold pools with severe convection, which is used for initialisation and boundary conditions for higher resolution models run across the USA....these often mess up the broadscale!
  4. It's amazing how much we expect often expect of NWP whatever the weather element of interest. Remember that these models are generally scored via official metrics with their performance we geopotential height, temperature and winds speeds at mid/upper levels, not by how well they forecast a maximum temperature over London....examples can be seen in the paper attached. Think of all the elements that models that cover a global domain have to forecast, from tropical cyclone tracks and intensities, precipitation accumulations from deep tropical convection to snowfall which is orographically enhanced by flow over high mountains. Temperatures from the hottest desert regions over sand surfaces by day, to the arctic with snow cover / permafrost in the dead of the winter night, to wind speeds and gusts over seas and mountains. These models also have to provide initial and ongoing boundary conditions for the limited area higher resolution models that nest inside them. As these are unified forecast systems (one model forecasting all variables) each forecast element is a trade off against something else. For example if you were to try and fix the low temperature bias during warm days in the mid-latitudes....increasing temperatures could ruin CAPE and convective precipitation forecasts...with this then potentially ruining synoptic patterns. GFS is often derided for its wildly varying synoptics which are often out of line with others across the UK. However from using it and the higher resolutions models that feed off it across the USA at short lead times I am confident that is performs better than others with several features of severe convection such as generating strong cold pools (and the hazards associated with these). I believe that this may be one of the models failings in synoptic forecasting as severe convection in the wrong location leads to large head aches for NWP (as we will no doubt see this week in the UK). Even the UKV is a trade-off between all these different weather elements temps, cloud cover, wind, gusts, precipitation accumulations and many others...fixing one can worsen others. It's all a big trade off. We're lucky they are often as good as they are! Progress in Forecast Skill at Three Leading Global Operational NWP Centers during 2015–17 as Seen in Summary Assessment Metrics (SAMs) | Weather and Forecasting | American Meteorological Society JOURNALS.AMETSOC.ORG
  5. Just to let you know that is not correct, the pixelated nature if the RADAR returns is due to range of the object from the RADAR head. It appears that the netweather RADAR only utilises the UK / Ireland Weather RADAR network including Jersey in the Channel Islands. So RADAR returns in NE France are being scanned at range from the RADAR at Thurnham in SE England and hence look more pixelated in output. In addition at this range even the lowest radar beam (I believe it's 0.5 degrees for Thurnham) is scanning storms over France at fairly high elevations (see the graph for WSR-88D), so will not pick up on showers produced by shallow surface based convection across NE France (tops below 8,000-10,000 FT), but will pick up deeper convection from surface based cells across NE France.
  6. You've yet to reach the temperature needed for cumulus formation. A weather balloon from Nottingham Watnall at 11:00 UTC, suggested a temperature of of around 20 C needed to initiate surface convection (in the air that sampled). Lack of cumulus on the satellite images suggests that is still the case E of the Pennines, surface observations suggest the temperature more like 18-19 C in the Leeds region, where it is low 20s C further W (across NW England) where convection is deeper, and probably more moisture at mid levels.
  7. Hi there, I've used both Weather Stations - Accessories and Parts - Greenfrog Scientific WWW.GREENFROGSCIENTIFIC.CO.UK and HamRadioStore for spares. The Fine Offset Station come under many brands (mine looks exactly like the image you shared and is branded as a Watson), but all the parts that you recognis from your station alreadye (despite the brand differences) work well with mine.
  • Create New...