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

Re: 960729: Maps in NETCDF.

>To: address@hidden
>From: "johan.karlsson" <address@hidden>
>Subject: Maps in NETCDF.
>Keywords: 199607291403.AA13733

Hi Johan,

> I'm writing to you from the Swedish metereological and hydrological
> institute. I'm currently working with a system called AFPS. This
> system requires data in NETCDF-format, both the weatherdata and the
> backgroundinformations(maps) should be in NETCDF. We are on our way
> with the weatherdata(at least we think so?), but we don't really
> know how to deal with the background. Do you maybe have a set of
> maps in NETCDF, or do you know where we could obtain such files.
> Another solution  would be if there is a software that produces
> NETCDF-files from a "latitude-longitude-file". 
> 
> If you have the answer to our problem, please let me know.

We don't have any examples of representing map data in netCDF form, and
I don't know anyone else who has done this either.  I think netCDF may
not be particularly appropriate for representing map data, except as
arrays of latitude/longitude pairs representing coast lines and regions.
Each such array would be a separate variable with a fixed dimension
size, since they are all different sizes.  For example, you might
representing coast lines of continents by something like:

    dimensions:
       latlon = 2;       // latitude, longitude
       dimAfrica = 3000;
       dimEurope = 2000;
       ...
    variables:
       float AfricaCoast(dimAfrica, latlon);
       float EuropeCoast(dimEurope, latlon);
       ...

where, for example, this representation assumes the coast of Africa is
represented by the 3000 (latitude, longitude) pairs

   (AfricaCoast(0, 0), AfricaCoast(0, 1))
   (AfricaCoast(1, 0), AfricaCoast(1, 1))
    ...
   (AfricaCoast(2999, 0), AfricaCoast(2999, 1))

By the way, if the "AFPS" you refer to is the AWIPS Forecast Preparation
System <URL:http://www-md.fsl.noaa.gov/eft/papers/brochure.html>, then
you might address your question to the AFPS developers at
address@hidden.

--Russ

________________________________________________________________________

Russ Rew                                            UCAR Unidata Program
address@hidden                        http://www.unidata.ucar.edu