20030613: 20030611: Calculating Cape from gridded model data?..follow u p

Lewis,

You could calculate a surface height, with varying accuracy.
Assuming Z_1000 would only be reasonable at lower elevations.

The good news is that hght@0%none would be constant, so once
you have it, you can just copy the grid to your data set.
One possibility would be to use the global GFS from NCEP
and interpolate to your grid domain (The 1x1 degree grid
iis available, and you could interpolate it with gdbiint. 
Soon, NCEP will be producing the GFS output on the .5x.5 degree (grid #004)
global grid).

The Linux binary distribution for GEMPAK 5.6.J is:
http://my.unidata.ucar.edu/downloads/81/gempak_upc5.6.j_linux.tar.gz

Steve Chiswell


>From: "Poulin,Lewis [CMC]" <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200306131241.h5DCfQLd022126

>Thank you very much Steve. As usual..very valuable and helpful information..
>
>
>If I understand correctly: 
>
>(1)     since we are using the gempak from the source file:
>gempak_upc5.6.h_linux.tar.gz 
>        from :
>http://my.unidata.ucar.edu/content/software/gempak/index.html
>
>        We would first have to update our 5.6.h to 5.6.j for Linux?
>        I assume it exists for Linux?
>        Which would then include the Cape function you mentioned. 
>
>(2) At this time our gridded data:
>    does not include Z_surface (hght@0%none)
>
>    Could I then as an alternative: 
>
>    (a) use Z_1000mb instead? (hght@1000%pres)
>or
>    (b) since Psfc=Pmsl exp(-Z_surface/Scale_height)
>        first compute Z_surface = -Scale_height * Ln(Psfc/Pmsl)
>        Add this Z_surface value to my gempak grid
>        Then do my Cape calculation? 
>        I know it's not the best but it would work I think..
>
>Thanks again
>
>Lewis
>
>
>
>-----Original Message-----
>From: Unidata Support [mailto:address@hidden]
>Sent: 12 June, 2003 17:01
>To: Poulin,Lewis [CMC]
>Cc: 'address@hidden'
>Subject: 20030611: Can we calculate Cape from a gridded model data? 
>
>
>
>Lewis,
>
>I created a CAPE routine in GEMPAK 5.6.J that allows the computation of
>CAPE Grids using pressure levels for data sets that don't have CAPE computed
>
>already.
>
>The usage is:
>GFUNC    = cape(P_bottom,Z_bottom,T_bottom,Td_bottom)
>
>The 4 input variables are grids used to define the bottom of integration.
>Generally, they are at the surface, but could be elevated if desired.
>The integration uses data on pressure levels found in the data set and
>computes THTA_parcel - THTA_environment areas. 
>
>For example, the NGM data set we receive doesn't have CAPE. It does have
>surface pressure and surface elevation grids (PRES@0%NONE and HGHT@0%NONE),
>and Temperature and Relative Humidity at 2 meters (Dewpoint can
>be calculated from temperature and relative humidity internally within
>GEMPAK),
>so I can calculate CAPE with:
>
>GLEVEL = 2
>GVCORD = hght
>GFUNC    = cape(pres@0%none,hght@0%none,tmpc,dwpc)

>The quality of result depends on how many pressure levels are in your data
>set
>(eg, if you only have 1000, 700, 500, 300 levels, the result is likely to
>differ from a value that is computed within the model). I get very good
>comparison with data at 50 mb intervals when compared against the value
>output from the ETA model we have access to.
>
>In general, we don't have all 4 parcel bottom variables on the same level
>and
>coordinate, so I typically have to use teamperature and dewpoint on levels
>near the surface (like 2 meter, or .9950 sigma).
>
>Steve Chiswell
>
>
>
>>From: "Poulin,Lewis [CMC]" <address@hidden>
>>Organization: UCAR/Unidata
>>Keywords: 200306111853.h5BIrfLd012747
>
>>Hi Unidata Support..
>>
>>I have successfully converted our Canadian Models into Gempak format and
>can
>>view them in Garp and Nmap.
>>
>>The models I am viewing are not sounding data but rather actually model
>data
>>on various grids. Some grids are local to north america, other grids are
>>global in coverage. 
>>
>>I have successfully done addtional calculations on those grids to add
>>addtional fields to the gempak file like Thta level calcs and precipitation
>>interval calcs.
>>
>>Our model output dataset does not include CAPE as the American models do.
>>We do include a wide range of Height, Temperature and Dewpoint depression
>>data for a large number of levels from which I believe it would be possible
>>to calculate CAPE.
>>
>>I was researching a bit on how to use Gempak to calculate CAPE but my
>>interpretation of the info suggests that we need sounding data on which to
>>do the calculation?
>>
>>I was curious if there was a way I could generate CAPE values from our 3-D
>>gridded model data sets. Ideally it would be a Cape value for each grid
>>point which I could then 
>>contour in Gempak. 
>>
>>Any advice you would have that could point me in the right direction would
>>be greatly appreciated. 
>>
>>Many thanks..
>>
>>Lewis Poulin
>>CMC-CMOI.
>>
>
>****************************************************************************
><
>Unidata User Support                                    UCAR Unidata Program
><
>(303)497-8643                                                  P.O. Box 3000
><
>address@hidden                                   Boulder, CO 80307
><
>----------------------------------------------------------------------------
><
>Unidata WWW Service              http://my.unidata.ucar.edu/content/support
><
>****************************************************************************
><
>