Jump to content

radarman

Members
  • Posts

    13,100
  • Joined

  • Last visited

Posts posted by radarman

  1. 18 minutes ago, OceanStWx said:

    It's pretty comical how thin the supplies are sometimes. Like the whole radar service life extension program (SLEP). We had like two extra pedestals for the whole country, so we leapfrogged teams around the nation refurbishing. Give CAR the spare pedestal, refurbish theirs, give it to GYX, refurbish GYX's old pedestal, give it to BTV, etc. 

    It's a smart move as long as the holes are all drilled straight.  I've been waiting 12 mos for a refurbishment, with no spare

  2. 2 hours ago, HoarfrostHubb said:

    A few mangled flakes here this afternoon 

    Several rounds of convective showers along 495 and the pike this afternoon.  Temps were close to 50 though so no dice on mangled flakes.  Low 40s I think we could have pulled it off.  

    Gorgeous day on the Cape beforehand.  Mid Novie is a great time to be there.

    20231119_131013_resize_48.jpg.e8f415512e2f75b0116522b45b9588f1.jpg

    • Like 5
  3. 37 minutes ago, bwt3650 said:

    2018-2019 was phenomenal at for NNE if i remember?

    incredible... best ever...

    I believe Mt Ellen had something like 90" for Novie.  And even further south, Magic was 100% open with 3-4' of pow by 11/29.

    • Like 1
    • Thanks 1
  4. Our 1 minute update radar data has a ton of potential for training ML models.  Have hundreds of terabytes of it.  Starting to look into it now.  

    Recently had a paper accepted that uses a non ML approach, run in parallel with parameterizations tweaked to produce an ensemble nowcast and converted to gridded risk for stochastic routing purposes.  Every minute produces an updated grid of risk for every minute from 1-20 mins.  (About 2 mins latent)   But a single ML based deterministic nowcast would still be better it if were accurate.

    • Like 2
  5. 2 hours ago, DavisStraight said:

    Has there been any discussion of this new model? First I hear or read about it.

     

    Why your weather forecasts may soon become more accurate (yahoo.com)

    I'd like them to work more on their nowcasting.  Some results are promising but it still has a quite ways to go.  The benefits if they could quickly generate a highly accurate 20-30 minute nowcast would be immense.  AI based methods *might* be able to pick up on convective initiation even without true data assimilation.  Non AI nowcasts cannot.  But AFAIK non AI methods are still outperforming AI methods right now.

    • Like 2
  6. 19 minutes ago, dendrite said:

    Halloween 11. 

    Then there was the 59-0 Pats game on 10/18/09 in the accumulating snow in his area.

    IIRC 10/18/09 was rain further inland too... good lift was highly localized right in that part of SEMA

  7. 28 minutes ago, dryslot said:

    You have skiers at a ski resort skiing on snow and worried about falling on a snowy walk?.............:huh:

    We could be hucking cliffs and steep trees and ice waterfalls all day but never declare victory until after we navigate the stairs to the lodge and the trek across the parking lot.  

    • Haha 1
  8. 2 hours ago, dendrite said:

    Was that for sure a marten or an ermine?

     

    1 hour ago, tamarack said:

    Never heard of an albino marten before, but most mammals produce very rare albino offspring so it's certainly possible.  A long-tailed weasel is about 2/3 the size of a marten and always turns white, just like it's much smaller short-tailed weasel.  I think both species are called ermine when white.

    Yep you are right.  An Ermine not a Marten.  A big one.  I guess they are not particularly rare after all.

    • Like 1
  9. Big southerly flow and high dews and rain are what takes out man made this time of year.  A couple hours of 50s or even 60s on low sun angle is of negligible effect on your typical early season trail offerings.  The bigger issue is the pause in expansion.

    • Like 2
×
×
  • Create New...