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

[GEMPAK #NGX-103614]: gdvint thta levels specification

> Hello, Steve:
> I want to use gdvint to interpolate RUC grids to
> isentropic levels. Using glevels=1000-100-25 this
> produced coarse resolution in THTA:

John,

The GLEVEL values should be the levels you want to interpolate to, as in
the first "phelp gdvint" example. I used GLEVEL = 260-440-4

 GDFILE   = $MODEL/ruc/2006103012_ruc.gem
 GDOUTF   = ruc_thta.gem
 GDATTIM  = f000
 GVCORD   = pres/thta
 GLEVEL   = 268-436-4
 MAXGRD   = 10000
 GAREA    = grid
 VCOORD   = none
 GEMPAK-GDVINT>r
 PRES         found. 
 TMPK         found. 
 HGHT         found. 
 RELH         found. 
 UREL         found. 
 VREL         found. 
 OMEG         found. 
 AVOR         found.

The initial list of "found" is just a list of variables found on the input 
coordinate system.



> 
> NUM       TIME1              TIME2           LEVL1 LEVL2  VCORD PARM
> 12     060902/2300F000                        425         THTA UREL
> 
> 18     060902/2300F000                        400         THTA UREL
> 
> 24     060902/2300F000                        375         THTA UREL
> 
> 30     060902/2300F000                        350         THTA UREL
> 
> 36     060902/2300F000                        325         THTA UREL
> 
> 42     060902/2300F000                        300         THTA UREL
> 
> (this report from gdlist). Is there a way to specify higher
> THTA resolution? The terminal output from the run contains some
> information I don't fully understand:
> 
> GEMPAK-GDVINT>run
> PRES         found.
> TMPK         found.
> HGHT         found.
> RELH         found.
> UREL         found.
> VREL         found.
> OMEG         found.
> AVOR         found.
> HGHT         not interpolated to level =   100
> HGHT         not interpolated to level =   125
> HGHT         not interpolated to level =   150
> HGHT         not interpolated to level =   175
> HGHT         not interpolated to level =   200

Those levels eg 100 Kelvin (in THTA coordinates are outside the range of vlues 
that
the input pressure levels can generate. The 260-440-4 I used above only issued 
warnings about 
260, 264 and 440K. The rest produced levels.



> 
> A number of variables are 'not interpolated' to levels 100...
> 1000; what does this mean, I wonder? UREL and VREL are listed,
> but there are fields at THTA values; HGHT and RELH are listed,
> and while they don't appear, arguably corresponding variables
> PSYM and SPFH do appear. I assume that this variable 'swap' is
> a convention, and I'm certainly familiar with the significance
> of the Montgomery Potential in THTA coordinates.

SPFH is the specific humidity. It is interpolated rather than relative
humidity since RH varies with both temperature and moisture (which leads to
an extra degree of freedom in interpreting your results!), whereas
SPFH can be calculated at the available input levels and interpolated using 
only the
pressure. 

Montgomery sream is as you mention a useful quantity for analyzing vertical 
motion
on THTA surfaces. The CpT + gZ quantity is scaled by 100 (which you will be
able to see by the output values). You could get back to Z of course.


> 
> I assusme that I can change gfunc from the
> ALL specification I have to a limited set, omitting variables
> in which I have no interest in this application

There is no GFUNC, but the VCOORD variable is used to just copy over variables 
on
other than your input coordinates (such as TROP, NONE, etc) if you want.

> 
> Also, when I've got an isentropic grid file I want to calculate
> PV; this would be the venerable Ertel PV to an adequate approximation.
> I see that gdvint isn't interpolating AVOR, but I assume that
> use of the PVOR(?,wnd) gparm value will yield this. But what is
> the scalar variable in this case, I wonder?

Your scalar will be the isentropic PRES grids you produce. (eg use s=THTA
in pressure coordinates, and s=PRES in isentropic coordinates).

> 
> Thanks for any help you can provide on this.   John M.
> 
> 

I used the above input to gdvint and then was able to run GDCROSS with:
 GDATTIM  = f000
 GVCORD   = thta
 GFUNC    = pvor(pres,wnd)
 PTYPE    = lin
 CINT     = 0
 SCALE    = 999

Let me know if you have further questions,

Steve Chiswell
Unidata User Support

Ticket Details
===================
Ticket ID: NGX-103614
Department: Support GEMPAK
Priority: Normal
Status: Closed