Automatic front drawing alorithm

[Sestak, Dr. Michael]

Since several people have requested this program, I will post it to the list
in a few days.  First I need to put together some background information.
The most important is to make it clear that this program does not plot nice
red and blue fronts with appropriate shaped bumps on them.  In fact it
doesn't plot anything.  From gridded fields of 1000 mb and 700 mb heights,
it calculates an index based on gradients (gradients of gradients, actually)
of the average temperature through that layer, which do a pretty good job of
indicating front locations.  While (as someone else mentioned) this
technique was developed in the ancient days of 2.5 degree global models (or
even earlier in concept), it gives excellent results with high resolution
limited area models (at least they look good, I have only done limited
verification and that would depend on the quality of the model in any case).


You have to write your own program to convert the index to front lines.  I
haven't done this except in the crudest form.  However, the authors of the
paper mentioned earlier have apparently developed a more elaborate front
drawing system (though I don't know if it uses grads).  Meteorological
Applications is a journal from Cambridge University Press which charges an
amazing price for reprints.  Probably better to contact the authors Donald W
McCann and James P Whistler, Aviation Weather Center, 7220 NW 101st Terr.,
Kansas City, MO 64153, USA

Michael Sestak
Fleet Numerical Meteorology and Oceanography Center
Monterey, CA 93940





In Meteorological Applications Volume 8 (2001), pages 193 to 203 is an
article "Problems and solutions for drawing fronts objectively" by Donald W.
McCann and James P Whistler which describes one way to do automatic frontal
location.  It is based on an algorithm called ggtheta developed by Leo
Clarke many years ago here at Fleet Numerical.  I have a fortran routine
that implements the basic technique, but I have not made the changes
suggested in the article.  Let me know if you are interested.

Michael Sestak
Fleet Numerical Meteorology and Oceanography Center
Monterey, CA

-----Original Message-----
From: Evgeny Himmelreich [mailto:evgenyh at IMS.GOV.IL]
Sent: Tuesday, December 28, 2004 5:23 AM
To: GRADSUSR at LIST.CINECA.IT
Subject: Automatic front drawing alorithm


Hi everybody!

Does anyone knows the algorithm (or have an GRADS script) of automatic front
lines determining ?
I tried to do it with Bracknell High resolution model using combination of
surface wind vorticity
and temperature advections but it not so good.

Maybe somebody can point me to publications in this area or give me working
algorithm.

Thanks

==========================
           Evgeny Himmelreich
   Israel Meteorological Service
 P.O.Box 25, Bet-Dagan, ISRAEL
        Tel: +972-3-9682157
         http://www.ims.gov.il
 ==========================