Re: [cf-satellite] Attaching coordinate variables of different sizefrom band dimensions

• To: cf-satellite <cf-satellite@xxxxxxxxxxxxxxxx>
• Subject: Re: [cf-satellite] Attaching coordinate variables of different sizefrom band dimensions
• From: ghansham sangar <ghanshamsangar@xxxxxxxxx>
• Date: Tue, 17 Jan 2012 21:37:32 +0530

R/Sir

Well I have an idea for this one.
Should it not be like that:

we can add an attribute to coordinate variables and the data variables. Let
us call it resolution.
and if that resolution variable is defined, it should be able to figure out
how to correlate the
lat/lon values with data values.
For example
dimensions:
   xc = 1400;
   yc = 1400;
   xc1 = 5600;
   yc1= 5600;
variables:
   float xc(xc):
       xc: axis = X
       xc: long_name = "x-coordinate in Cartesian system"
   float yc(yc):
       yc: axis = Y
       yc: long_name = "y-coordinate in Cartesian system"
   float xc1(xc1):
       xc1: axis = X
       xc1: long_name = "x-coordinate in Cartesian system"
   float yc1(yc1):
       yc1: axis = Y
       yc1: long_name = "y-coordinate in Cartesian system"
  float lon(yc, xc):
        lon:long_name="longitude"
        lon:units="degrees_east"
        lon:resolution="8000" //e.g.
  float lat(yc, xc):
        lon:long_name="latitude"
        lat:units="degress_north"
        lat:resoultion="8000" //e.g.

float TIR(yc,xc) ;
    TIR:long_name = "Thermal Infrared;
    TIR:units = "mw.m-2.sr-2" ;
    TIR:coordinates = "lon lat" ;
    TIR:resolution = "8000" //e.g.
float VIS(xc1,yc1):
     VIS:long_name="Visible"
     VIS:units="mw.m-2.sr-2";
     TIR:coordinates="lon lat"
     TIR:resolution="2000" //e.g.
float WV(xc,yc):
     WV:long_name="Water Vapor"
     WV:units="mw.m-2.str-2"
     WV:coordinates="lon lat"
     WV:resolution = "8000"

I think some correlation should get established between data variable VIS
band and coordinate variables lon/lat using resolution attribute like here.

if (resolution_coordinate_variable>resolution_data_variable) {
        n = resolution_coordinate_variable/resolution_data_variable
        //For every  n values of data variable, there is one value of
coordinate variable.
} else {
         relative_resolution =
resolution_coordinate_variable/resolution_data_variable
        //For every n values of coordinate variable, there is one value of
data variable.
}

I think this is demostrable too.

regards
Ghansham


On Mon, Dec 26, 2011 at 11:14 PM, Russ Rew <russ@xxxxxxxxxxxxxxxx> wrote:

> > Can fill data be used in coordinate variables ?
>
> Not for CF-compliance currently, because section 1.2 (Terminology) of
> the CF Conventions says "Missing values are not allowed in coordinate
> variables".
>
> Also the NetCDF User Guide implies that coordinate variable values
> should be monotonically increasing or decreasing, which implies they
> should not include fill data.
>
> --Russ
>
> > ---------- Original Message ----------------------------------
> > From: ghansham sangar <ghanshamsangar@xxxxxxxxx>
> > Date:  Fri, 23 Dec 2011 12:31:18 +0530
> >
> > >Yeah you are very much right for every 4x4 pixel box there one lat/lon
> > >value. Its bilinear interpolation.
> > >And I think we should have such arrangement, the reason is that I have
> seen
> > >in some modis HDF4 products, lat/lon information is given in this way
> only.
> > >
> > >
> > >regards
> > >Ghansham
> > >
> > >
> > >On Thu, Dec 22, 2011 at 10:18 PM, john caron <caron@xxxxxxxxxxxxxxxx>
> wrote:
> > >
> > >> On 12/22/2011 12:40 AM, ghansham sangar wrote:
> > >>
> > >>> R/Sir
> > >>>
> > >>>
> > >>> Is there anyway to attach Coordinate variables of different dimension
> > >>> with bands of different dimension.
> > >>>
> > >>> For example
> > >>> data is of dimension 5600x5600
> > >>> And 2D lat/lon coordinate variables are of size 1400x1400 each.
> > >>>
> > >>> Is there a way to describe such datasets using CF conventions.
> > >>>
> > >>
> > >> there is no current way to do this in CF. i assume you have
> geolocation
> > >> for every 4th point? what is the standard way to interpolate to the
> other
> > >> points?
> > >>
> > >> John
> > >>
> > >> ______________________________**_________________
> > >> cf-satellite mailing list
> > >> cf-satellite@xxxxxxxxxxxxxxxx
> > >> For list information or to unsubscribe, visit:
> > >>
> > http://www.unidata.ucar.edu/**mailing_lists/<
> http://www.unidata.ucar.edu/mail
> > in
> > g_lists/>
> > >>
> > >
> > >
> > >
> > >
> >
> >
> >
> > ..............End of Message ...............................-->
> >
> >
> >
> >
> >
> > _______________________________________________
> > cf-satellite mailing list
> > cf-satellite@xxxxxxxxxxxxxxxx
> > For list information or to unsubscribe, visit:
> http://www.unidata.ucar.edu/ma
> > iling_lists/
>