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

Devonshire

Members
 View Profile  See their activity

  • Posts

    122

  • Joined

  • Last visited

 Content Type 

Posts posted by Devonshire

    The Hunt for Cold continues 02/01/19

    in Spring Weather Discussion

    Posted

    14 minutes ago, nick sussex said:

    Current rankings at day 6 using anomaly correlation for 12 hrs runs .

    ECM 0.878

    UKMO 0.855

    FV3  GFS para 0.840

    GEM 0.838

    GFS 0.824

    JMA 0.813

    NAVGEM 0.812

    Some observations/questions Nick:

    - what part of the globe is that for? (our little country is rather small in the scheme of things!)

    - and how is 'success' measured for the sake of those performance stats (rained in London when it said it would?)?

    Also, if those are correlations (predicted/actual something), then 77% of variance explained by ECM compared to 66% for NAVGEM is not really so huge a difference - and how does that translate into snow/no snow in UK?!;

    Add to that that the SSW is likely to stretch these models to different degress (and the stats won't have been achieved in those conditions in any case!)

    CORE MESSAGE: stats (especially of 'fuzzy' measurements) can appear to be more authoritative than they actually are for any given purpose. I would think those figures are about as useful as a proverbial chocolate teapot at the moment. Happy to be convinced otherwise

    • Like 2
    • Thanks 1

    The Hunt for Cold continues 02/01/19

    in Spring Weather Discussion

    Posted

    13 minutes ago, weirpig said:

    No matter what time of year  it never has the edge over the Ecm.  For me no model has covered itself in glory this winter but the GFS in particular has been woeful  it comes out 4 times a day which is why most look at it  but for me its ranked fourth in weather consistency.    

    I may be wrong, but I get the impression that the GFS parallel/FV3, or whatever it is best called, has been well-received of late - or is that just it has been providing the best-looking charts rather then most accurate?

    Hunt For Cold: Model Discussion - Heading Into Christmas

    in Spring Weather Discussion

    Posted

    31 minutes ago, karlos1983 said:

    Happy New Year everyone!

     I’m giving this place a miss for the rest of the day before I get dragged into this extremely bizarre negativity!  

    The clues are there! Wild swings in the output, GEFS members showing Hight rises to the North, SSWE unfolding as we speak! But whatever  :oldrolleyes:

    Don't let the moaners get you down! (no snow flurries today, but a flurry of newly-ignored members getting me close to 22 pages of them now)! MODS - maybe 'the ignored' can be notified each time someone sets them to that status - or at least given stats as to how many people are ignoring them?!

    I actually think the last couple weeks and upcoming couple weeks of model watching are the most interesting for a long while - looking to see how the different models handle the SSW and indeed how the SSW will play out. I particularly look forward to the posts of the several forum members who take time to put together thoughtful posts, we all know who they are: the few who nobody sets to 'ignore'!!!

    I also agree, it is bizarre how some posters are nor only verging on calling winter over, but denying any taste of winter so far - even down in Devon this winter we have had several frosts, with snowfall up the road on higher Dartmoor, and several more frosts to come this week. Other more northerly/easterly members have had a proper taste.

    Happy New Year - and may it be a frosty, snowy, chilly one in the not too far off future! I am most definitely not going to be lowering my expectations!

    • Like 5

    Hunt For Cold: Model Discussion - Heading Into Christmas

    in Spring Weather Discussion

    Posted

    11 minutes ago, Stuie W said:

    Sorry not being grumpy but if we are talking about the models then unless someone shows the effect combined with a model chart then it`s a hard watch in here.

    I guess the ssw graphics do come from models - I quite like to see them as context to operational runs. I do feel that when ssw graphics are posted, though, more often than not they could do with a little more meat in terms of interpretation viz our hunt for cold (as often happens with other model graphics too)

    • Like 1

    Hunt For Cold: Model Discussion - Heading Into Christmas

    in Spring Weather Discussion

    Posted

    16 minutes ago, Walsall Wood Snow said:

    Am I to assume cold just in time for Christmas or even between Christmas and New Year is looking a lot less likely now then?

    Impossible to answer that question as you don't provide a reference point (much less likely than it was when?). Perhaps I will assume you mean since August this year - in which case I would answer 'no - it is much more likely now'! (chin-up)

    • Thanks 1

    The hunt for cold - Model discussion (late November)

    in Spring Weather Discussion

    Posted

    2 minutes ago, booferking said:

    To much negativity in here just wait until the turn of the year everything is aligning nicely signals are there ssw glosea etc the met are on board what more do you want???

    I suspect the comment related to there being a lot of babble on here without any model evidence or discussion - I feel the same. And just as frustrated when a chart is posted with a one liner like 'extraordinary', without saying why, or just '18z' etc. All of which happen now and then

    • Like 2

    The hunt for cold - Model discussion (late November)

    in Spring Weather Discussion

    Posted

    1 hour ago, johnholmes said:

    Not comparing run to run at those time scales but like with like 00Z or whichever run you want to use. Do this down to about 144 then yes run to run and comparisons between the 'big' 3, GFS, ECMWF, UK Met. 

    It will still give the enjoyment of watching the models but is almost always a less 'up and down' ride.

    John, I can see the logic of this in terms of expectation management (and emotional resilience!) - but this begs the question - that has probably been asked (and answered) before somewhere;  are the 0, 6, 12, and 18z runs (e.g. for GFS) different in the assumptions, algorithms, no. of data points built in (or anything else) - or is it simply a case of different starting data being put in? (notwithstanding weather balloons avoiding Santa's sleigh over Christmas and related myths)

    • Like 2

    The hunt for cold - Model discussion (late November)

    in Spring Weather Discussion

    Posted

    7 minutes ago, Ed Stone said:

    That's the beauty of the weather, coldie...Though people and models are, to varying extents, always flapping in the wind, the evolution of the weather itself is 'set in stone'. If it weren't, and we were not in a deterministic universe (quantum effects notwithstanding) numerical weather-prediction devices simply wouldn't work..

    ummm - Chaos Theory? Not sure if you are being ironic - if so, apologies - I am a little' emoji-autistic'.

    • Like 1

    Model output discussion - The Hunt For Cold

    in Forecast Model Discussion

    Posted

    7 minutes ago, Quicksilver1989 said:

    Yup that is correct, the GFS parallel is also included. I'd love for ECM ensembles to appear on WZ! As that would be really valuable information but they aren't available there unfortunately.

    Ah, ok - I wonder if there is an easy way of distinguishing the parallel run from the others - indicating which category it is in, to see its trends alongside the other trends?

    Model output discussion - The Hunt For Cold

    in Forecast Model Discussion

    Posted

    19 minutes ago, Quicksilver1989 said:

    image.thumb.png.ac23b8594e49d2d54ddfa04f24c2544c.png

    06z swingometers tell you all you need to know about the output. Downgrades all round with temperatures pretty close to average (hence the description of non-descript for temperatures). I'd say only two or three of the ens members look capable of producing a decent northerly looking at the output on November 25th.

    I really do like visual presentations of trends - such as this. Am I correct in thinking this shows the distribution of gfs ensembles into your categories? If so, I would wager that a few on here might be interested in such presentations for other models - especially the gfs parallel! I guess similar information can also be gleaned from cluster images that are posted on a run by run basis, but I like this presentation of trends.

×
×
  • Create New...