Jump to content
  • Member Statistics

    17,509
    Total Members
    7,904
    Most Online
    joxey
    Newest Member
    joxey
    Joined

The Psuhoffman Storm


Ji

Recommended Posts

  • Replies 6.9k
  • Created
  • Last Reply

... The surface looks more like an incoming clipper at hour 60 than a Miller A in the works..

Same with the 500 mb....

http://www.nco.ncep....fs_500_060m.gif

But the 500 has no where to go but south because of the big vortex over the maritimes. I doubt it will be as wet as the last two runs but it probably will be wetter than a straight clipper.

Link to comment
Share on other sites

Perhaps DTK can elaborate. I do know the NAM expanded its domain sometime in the recent past--but it still is initialized by older GFS runs along its outer domain as well (0Z NAM would be initialized with the 18Z GFS--if I understand it correctly)--another reason to give it less credence when the system in question is tracking in from the Aleutians.

I don't have much to add as I'm not even sure I fully understand why the NAM had (still has?) a drift in the large scale (though there are various hypotheses related to the model itself, various DA aspects including bias correction of satellite data, etc.). The NAM domain nowadays is actually quite large (see below)...so lateral boundary conditions shouldn't be an issue for most of the CONUS part of the domain for most of the integration.

I think that we've seen pretty big improvement since going to partial cycling instead of allowing the NAM to cycle on itself. I'm not in the meso-group, so I may not have this exactly right (I don't know exactly what what states they use and how far they go back for partial cycling...but basically before each analysis, the NAM is started from a previous GFS state (from 12 hrs ago all fields, not just the boundaries). Then, a series of analyses and forecasts of the NAM are run leading up to the actual initialization time, at which point the full 84 hour forecast is run (the lateral boundary conditions for this are updated using the previous GFS fcst, from 6 hours ago). I think the cycling is 3 hourly leading up to initialization time (in other words, it starts from a GFS forecast, assimilates obs, runs a short NAM forecast, assimilates more obs, runs another short NAM forecast, etc.). This allows the NAM to start from a large scale that is like the GFS, but try to resolve scales / features at its own resolution and consistent with its own dynamics. This essentially cuts off any possible "drift" in the large scale, at least relative to the GFS, since it's forced to do a complete restart from a global state very so often.

g110.12kmexp.jpg

Link to comment
Share on other sites

There's virtually no gulf moisture in the storm now. Last night had multiple inches of rain in Dixie. Dart board models again. What's the point of even tracking? This solution is nothing like the socialist models.

I have this copied for just these posts.

Yeah exactly. The inevitable "THE MODELS SUCK!" comments are going to come out in these patterns. The models are doing fine. It is a nod to human ingenuity that these models can even suggest these threats 5-7 days in advance. Kudos to them; no kudos to the people who believe models suck if they can't nail down threats a week in advance.

Link to comment
Share on other sites

Still way off the coast south of the Aleutians.

I'm pretty sure there will be Alaska and/or Hawaii based recon missions as part of WSR 2011 in the coming days. The flight they sent out of Japan this morning (obs made it into 12z) was technically targeted for an Alaskan event, but the more data the better :).

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...