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

19990716: working with sgma


Matt, 

I looked at your script and things output fine.

You should have STBC at sgma levels in your gdinfo like
 LEVL1 LEVL2  VCORD PARM
  9925         SGMA STBC
  9875         SGMA STBC
  9825         SGMA STBC
etc.

To answer your question, quo(ldf(A),ldf(B)) is equivalent to dA/dB.
When viewing these values in Garp, you should select vertical
coordinate SGMA and ensure that the radio button for "level" is
selected and "Scalar" is set. Clicking on the "Level 1" button should
show you all the SGMA levels to choose from. It sounds like you have gotten this
far. After typing in your parameter name STBC in the gfunc bar, you
probably need to go to the "more" to set the scale factor to something like "3",
or "999" to let Garp pick its own scaling.

Comparing dT/dp with dT/dsigma, since your sgma layer depths from 9500 to 9000
would represent pressure values from 950 to 900mb, so you would expect 
at least dSGMA to be an order of magnitude larger than dPRES, so you will have 
to
increase the scale factor otherwise, the denominator will be so large, your
values will apear like 0. (Remember, that Gempak is scaling sigma by 10000 to
get the SGMA levels (since GEMPAK has no way of specifying decimal points
in GLEVEL values).

I had designed the script I sent you to specify GRDNAM using
a single level, instead of outputing the computation using
LEVEL1:LEVEL2 since it is easier to display levels in gdprof-
howver, if you want to use the STBC as a layer quantity, then
you can just remove the @GLEVA in the GRDNAM specification.

Once STBC is output as a layer rather than a level, then you
will have to specify LEVEL1 and LEVEl2 when plotting the function.
I has the script calculate GLEVA as the average of the top and
bottom level- eg the value in the middle of the 2 levels.

Regarding Mixing Depth, I don't know of an easy way to calculate this from the 
gridded data.
Unless you estimate this from the LCL, it requires integrating the grid levels.
If you know how you want to estimate this parameter, I can help you try to
do the calculation.

Steve CHiswell
Unidata User Support




>From: Matt Fearon <address@hidden>
>Organization: Desert Research Institute
>Keywords: 199907162317.RAA10915

>Steve,
>
>I am trying perform stability calculations again.. Only this
>time using sgma coordinates. I modified the script you gave
>me to calculate stability using pres and staying above ground.
>I attached it below. The script seems to run fine with gempak
>- gddiag and shows up at calculated levels when viewing the
>gdfile in gdinfo,but when I try to view the results of stbc in garp,
>I receive values of zero for contours.
>
>Is quo(ldf(tmpc),ldf(sgma)) equivalent to dT/dsigma?
>I have also tried stab(tmpc) using sgma and I always
>receive an error. Is stab available w/ sgma?
>
>I know dT/dsigma is not usually looked at in terms of stability,
>but I am trying to determine what sort of values of dT/dsigma
>correspond to values of dT/dZ. I would like to do this in layer
>sections through the troposphere, e.g. 9950:9900 , 9900:8950
>up to 3500:3000 sgma.
>
>One last question, do you know of any gempak scripts available
>that calculate mixing height for a grid?
>
>Thank you for all your help,
>Matt Fearon
>address@hidden
>
>
>
>#!/bin/csh
>
>set GDFILE=/data1/gempak/grids/99061412_eta.gem
>
>
>set GLEV1 = 9950
>while($GLEV1 > 9500)
>  @ GLEV2 = $GLEV1 - 50
>  @ GLEVA = $GLEV1 + $GLEV2
>  @ GLEVA = $GLEVA / 2
>
>  $GEMEXE/gddiag << EOF1
>  gdfile  = $GDFILE
>  gdoutf  = $GDFILE
>  gfunc   = quo(ldf(tmpc),ldf(sgma))
>  gdattim = f00
>  glevel  = ${GLEV1}:${GLEV2}
>  gvcord  = sgma
>  grdnam  = stbc@${GLEVA}
>  gpack   =
>  r
>
>  e
>EOF1
>  @ GLEV1 = $GLEV1 - 50
>
>end   #end while GLEV
>#end  #foreach FHR
>
>
>set GLEV1 = 9500
>while($GLEV1 > 3000)
>  @ GLEV2 = $GLEV1 - 500
>  @ GLEVA = $GLEV1 + $GLEV2
>  @ GLEVA = $GLEVA / 2
>
>  $GEMEXE/gddiag << EOF2
>  gdfile  = $GDFILE
>  gdoutf  = $GDFILE
>  gfunc   = quo(ldf(tmpc),ldf(sgma))
>  gdattim = f00
>  glevel  = ${GLEV1}:${GLEV2}
>  gvcord  = sgma
>  grdnam  = stbc@${GLEVA}
>  gpack   =
>  r
>
>  e
>EOF2
>  @ GLEV1 = $GLEV1 - 500
>
>end  #end while GLEV
>
>$GEMEXE/gpend
>rm *.nts
>
>