[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

19990913: trajectories with GEMPAK

Donna,

I included my contributed trajectory program that I use to create 2D
trajectories such as are shown at:

http://www.unidata.ucar.edu/packages/gempak/examples/
(see the Forward and Back forecast trajectories link).

The trajectories are 2D, eg they follow a constant model level.
I typically use the 0:30 MB above ground level wind from the AVN for
a 72 hour boundary layer trajectory. You could do a 700mb level
trajectory as well. If you wanted to do an isentropic trajectory,
then you would first need to use gdvint to provide the U and V wind
grids on the theta level that you wanted. The point being that
the trajectories are kinematic following whateve level and vertical
coordinate you start on.

I repacked the program and accompanying scripts in
~gbuddy/nawips-5.4/contrib/trajectory.tar.Z

(you should already have $NAWIPS/unidata/programs/trajectory
in the current gempak release, but the scripts have been updated
in the new tarfile).

You can unpack and build the program with:

cd $NAWIPS
zcat trajectory.tar.Z | tar xvf -

cd $NAWIPS/unidata/programs/trajectory
make
make install

Building will install the programs backtraj_gem and trajectory_gem
into the $GEMEXE directory and copy traj.pack into $GEMTBL.

The scripts I use to generate the trajectories and produce gif and
ntrans meta files are in $NAWIPS/unidata/programs/trajectory/scripts.


The scripts are as follows:

trajectory.csh
--------------
This script dumps out the lat, lon, U and V wind grids for each model
time (though 72 hours for AVN or 48 hours for ETA, any model can be
used if you define the HOURSm GLEVEN and GVCORD fields).
The script also expects a working directory defined as TRAJECTORY
to be available to dump out the fields.

After dumping out the needed fields, the script calls:
create_forw.gem
plot_forward.gem
create_back.gem
plot_back.gem
to create the gempak trajectory files from the grids that were dumped
out and plot the trajectories respectively (the scripts are set
up to plot gif files....to plot meta files, just comment out those
plot commands and use the plot_forward.meta script for example).

The "create_xxx.gem" scripts use the programs that were build 
in the previous step to generate forward and backward trajectories
as ship format surface files. The scripts currently iterate for
every 10 degrees longitude and 5 degrees latitude to create
a grid of trajectories...but you could modify this to create
trajectories from any lat/lon point desired.

A forward trajectory starts at the starting point and runs through
the forecast hours and determines where the point will end. Eg,
where wil the parcel at this point now be in 72 hours.

A backward trajectory assumes the initial point is where you
will end and determines where the point must start from at
f000 to arrive at that point. Eg, in 72 hours, where will our
parcel have originated that ends here.

The plot scripts plot the locations for all points produced in the
ship file (yellow dots), then place special markers at start, and
12,24, 36,48, 60 and 72 hours. A vector (SFPARM=arrw) is placed at the
end of the trajectory with magnitude relative to the speed.
You will want to change the "mv" command at the end that moves the gif files
to my www directory.

Let me know if this will be help you in what you need. There are other
program such as Hysplit that people use to compute trajectories
that take into account vertical motion etc. I initially used the program
to track boundary layer air parcels for the ACE experiment where the
assumption was that boundary layer air remained in the boundary layer.
Using the model grids available at each time was a first approximation
for the kinematic trajectory rather than trying to model the
processes involved in a parcel.

If you need help in setting these up, let me know.

Steve Chiswell
Unidata User Support


>From: address@hidden
>Organization: .
>Keywords: 199909131930.NAA03677

>I sent a message a couple of weeks ago to which I did not get a reply.  The
>substance of the message was that I am trying to find out if GEMPAK has a
>facility to do trajectories either from soundings or gridded fields.  I
>am willing to use a contributed program or add on if one exists.
>
>Donna Tucker                      http://chinook.phsx.ukans.edu/tucker.html
>address@hidden      Department of Physics and Astronomy  
>(785) 864-4738 (new area code!)    University of Kansas                 
>(785) 864-5262 (fax)               Lawrence, KS  66045-2151                   
>  
>
>From address@hidden  Tue Aug 31 12:48:18 1999
>Received: from phoenix.phsx.ukans.edu (phoenix.phsx.ukans.edu [129.237.97.64])
>       by unidata.ucar.edu (8.8.8/8.8.8) with SMTP id MAA17680
>       for <address@hidden>; Tue, 31 Aug 1999 12:48:18 -0600 (MDT)
>From: address@hidden
>Organization: .
>Keywords: 199908311848.MAA17680
>Received: by phoenix.phsx.ukans.edu (SMI-8.6/SMI-SVR4)
>       id NAA00866; Tue, 31 Aug 1999 13:46:02 -0500
>Date: Tue, 31 Aug 1999 13:46:02 -0500
>Message-Id: <address@hidden>
>To: address@hidden
>Subject: trajectories using GEMPAK?
>X-Sun-Charset: US-ASCII
>
>Is there a program in GEMPAK or related contributed software which 
>computes trajectories either from soundings or model data? 
>
>Donna Tucker                      http://chinook.phsx.ukans.edu/tucker.html
>address@hidden      Department of Physics and Astronomy  
>(785) 864-4738 (new area code!)    University of Kansas                 
>(785) 864-5262 (fax)               Lawrence, KS  66045-2151                   
>  
>