20001020: GEMPAK: Fwd: question about Q vector s component

David,
If you need to create terms of longer equations, the program to
use would be GDDIAG. you can store the terms in a grid file and
recall them later. 

You can define the output GRDNAM to be the name and level,cord,time
you need, for example GRDNAM=TERMA, will inherit the glevels and
times of the calculation...but if you need to specify the level rather than
a layer, you can set GRDNAM=TERMA@500 for example.

Here is a URL of a support question I answerd a while back on the use of
GDDIAG to give you some background.

http://www.unidata.ucar.edu/projects/coohl/mhonarc/MailArchives/gempak/msg02413.html
http://www.unidata.ucar.edu/projects/coohl/mhonarc/MailArchives/gempak/msg03828.html


Steve Chiswell
Unidata User Support



>From: David Wojtowicz <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200010210234.e9L2YQ424619

>Hi,
>
>  A user here as posed the following question to me concerning plotting
>Q vector components in GEMPAK.   (I'll let you read it first...)
>
>>Date: Fri, 20 Oct 2000 16:17:00 -0500 (CDT)
>>From: Mei Han <address@hidden>
>>To: address@hidden
>>cc: address@hidden
>>Subject: question about Q vector s component
>>
>>Hi, David,
>>
>>I have to ask you something about How to plot Q vector s component
>>using gempak.
>>
>>I tried a lot. But neither of them work.
>>
>>You might met such problem. If I write the whole expression of the Qs in
>>one variable,
>>
>>gvect=vasv[smul[4e9;qvcl[hght,geo]];smul[-1;kcrs[grad[ldf[hght]]]]]
>>
>>It's too long and need to separate the function. But if I define
>>temporary variable like,
>>gfunc=ldf[hght]//thickness in gdcntr, it cann't be used in the later
>>gdwind program.
>>
>>So I look for other two ways to solve this problem, but I didn't get them
>>work. I am wondering if I can get some idea from you.
>>I tries the two way:
>>
>>1. Use gddiag to construct a new grid named thickness and put it into a
>>new gempak data file. It said it need to separate the gvect function once
>>more.
>>    But actually it seems there is no other grids can be generated for
>>this function. Because gddiag just generate scalar grids.
>>
>>
>>2.I try to use gdplot, instead of gdcntr and gdwind, to plot Qs.
>>But it didn't work also.
>>
>>
>>Thank you very much!
>>
>>Mei
>
>
>We have run into this problem with the function being too long
>before in the context of Garp.  The solution there is to break
>the function into smaller pieces.  However the syntax is different
>here from the fdf files in Garp.
>
>I tried breaking it up as follows:
>
>gvect=smul(4e9;qvcl[hght,geo])  // A
>gvect=smul(-1,kcrs(grad(ldf(hght)))) // B
>gvect=vasv(A,B)
>
>However I only get:
>  [DG -25]  Vector field cannot be computed.
>
>
>The input data is from a eta211, lookalike grid output from MM5.
>You can plot the individual non-derived fields OK.
>glevel is set to 750:500
>
>
>Any ideas on how to accomplish what she wants.   (she'd use Garp where
>it works, but needs to be able to print it to postscript)
>
>Thanks!
>
>
>
>-- 
>--------------------------------------------------------
>  David Wojtowicz, Sr. Research Programmer
>  Department of Atmospheric Sciences Computer Services
>  University of Illinois at Urbana-Champaign
>  email: address@hidden  phone: (217)333-8390
>--------------------------------------------------------
>