Re: [ferret_users] height to pressure inversion

[Paul Young <paul.young@xxxxxxxxxxxxxxxx>]

Sorry to ask this again, but it fits in with the previous ZAXREPLACE
post:

--------------------
Hello all,

I would like to plot my atmospheric model data zonal mean using
sigma height co-ordinates instead of the sigma-pressure levels. If
possible, I would like to add a 2nd y-axis that has log pressures on it.

So far, I've produced a 4D array that contains the sigma height (km)
levels for each model point, converted from the sigma pres levels. This
allows me to re-grid my data onto height levels using ZAXREPLACE:

        !new sigma height axis
        DEFINE AXIS/Z=0:20:1/UNITS=km newz

        !get sigma heights
        LET sigma_ht = sigma_km[D=1]    
        
        !re-grid my data
        new_var = ZAXREPLACE(old_var,sigma_ht,newz)

        !plot Jan zonal mean
        FILL/L=1 new_var[x=@ave]

THE PROBLEM: Above works ok, but I would like to add data about the
level boundaries so that the plot starts at 0km (rather than at the
mid-point of the lowest sigma level)

I have tried using different definitions of the Z-AXIS, such as using
the /BOUNDS qualifier:

        !Define sigma ht bounds axis
        DEFINE AXIS/UNITS=km/BOUNDS/DEPTH newz = \  
                {0.02,0.18,..etc..,37.75}, \
                {0.0,0.04, 0.04,0.31, ..etc.., 32.33,53.30}

but this plots upsidedown (doesn't work at all without \DEPTH). I can't
get the /EDGES qualifier to work either (by just providing the
coordinates of the level edges). 

Am I using the /BOUNDS (or /EDGES) AXIS definitions incorrectly?

Also, I can't get the second (right-hand) y-axis to display a
log (sigma) pressure coordinate at all! 

Hope this is clear. Thanks for any help, and apologies again if I am
just being stupid about it.

Cheers,

Paul

On Wed, 2007-01-31 at 19:04, Ansley Manke wrote:
> Hi Halldor,
> The ZAXREPLACE functions will do this. The function takes a variable
> on one Z axis, a function relating the new Z axis to the old one; and
> it makes the conversion. The documentation contains examples using
> sigma coordinates, but the process can be used for any two monotonic Z
> axes.  See ZAXREPLACE in the Users Guide documentation.
> 
> Ansley
> 
> 
> 
> halldor bjornsson wrote: 
> > Dear Ferreters,
> > 
> > I am working with NCEP height data
> > ftp://ftp.cdc.noaa.gov/Datasets/ncep.reanalysis.derived/pressure/hgt.mon.ltm.nc 
> > 
> > This data has height in terms of pressure, i.e. Z(p).
> > I want to invert this into pressure as a function of height, i.e.
> > P(z).
> > 
> > I have been looking at the sigma regridding go files, but these are
> > fine to  change the 
> > vertical coordinate from p to z, but do not invert the variable from
> > Z to P, i.e. do not give me the pressure.
> > 
> > Does anyone know how to do a conversion such as this one?
> > 
> > Sincerely,
> > Halldór
> > 
> > -- 
> > Halldór Björnsson
> > Deildarstj. Ranns. & þróun
> > Veðursvið Veðurstofu Íslands
> > 
> > Halldór Bjornsson
> > Weatherservice R & D
> > Icelandic Met. Office
> 
> 
> -- 
> 
> ~>   ~>   ~>   ~>   ~>   ~>   ~>   ~>   ~>   ~>   ~> 
> 
> Ansley Manke, NOAA/PMEL   ansley.b.manke@xxxxxxxx
> 7600 Sand Point Way NE, Seattle WA
> Phone 206-526-6246,  FAX 206-526-6744
>