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

[IDV #KQC-444299]: I probably don't have my NetCDF file setup correctly...

Hi Tim:

ok, the sample file now has:

netcdf D:/work/efdcout.nc {
 dimensions:
   xi_rho = 259;
   eta_rho = 121;
   nv = 4;
   ocean_time = UNLIMITED;   // (24 currently)
 variables:
   float mask_rho(eta_rho=121, xi_rho=259);
     :long_name = "mask on RHO-points";
     :option_0 = "land";
     :option_1 = "water";
     :coordinates = "lat_rho lon_rho";
   float zeta(ocean_time=24, eta_rho=121, xi_rho=259);
     :standard_name = "height_above_reference_ellipsoid";
     :units = "m";
     :coordinates = "lat_rho lon_rho";
     :valid_min = -10.0f; // float
   float lat_rho_bounds(eta_rho=121, xi_rho=259, nv=4);
   float lon_rho_bounds(eta_rho=121, xi_rho=259, nv=4);
   int xi_rho(xi_rho=259);
   int eta_rho(eta_rho=121);
   float lat_rho(eta_rho=121, xi_rho=259);
     :long_name = "latitude";
     :units = "degrees_north";
     :standard_name = "latitude";
     :bounds = "lat_rho_bounds";
     :_CoordinateAxisType = "Lat";
   float lon_rho(eta_rho=121, xi_rho=259);
     :long_name = "longitude";
     :units = "degrees_east";
     :standard_name = "longitude";
     :bounds = "lon_rho_bounds";
     :_CoordinateAxisType = "Lon";
   float ocean_time(ocean_time=24);
     :standard_name = "time";
     :units = "days since 1995-1-1 0:0:0";
     :_CoordinateAxisType = "Time";

 :Conventions = "CF-1.4";
}

the lon_rho, la_rho are all zero, however.
the xi_rho, eta_rho just have 1,2.3...

if you want to use these as coordinate variables, you should use  

     :standard_name = "projection_y_coordinate";
     :standard_name = "projection_x_coordinate";

on xi_rho, eta_rho. and make sure they have the right values for the UTM 
projection (which i think you previously had there)

also have a look at

http://www.unidata.ucar.edu/software/netcdf-java/reference/StandardCoordinateTransforms.html


> I am confused and I apologize.  I thought that I had sent you a better 
> version than that.
> 
> I just replaced the old one on our ftp site with my latest, which is a lot 
> farther along.
> 
> The 'zeta' is actually the water elevation data and is what I have been using 
> as my test.
> 
> I project the UTM coordinates into lat/lon outside of NetCDF with a 
> subroutine that I added to EFDC and makes debugging NetCDF a little easier.
> 
> My latest version displays in IDV!  Mostly.  It is missing the border cells 
> and any single column or row.  I tried 'bounds=' to actually put in the cell 
> boundaries, but IDV doesn't seem to notice them.  'Color Shaded Plan View' 
> doesn't work, I think because I only have the lat/lon of the wet cells.  In 
> all displays, because I have 0 lat/lon the initial display field is huge and 
> you have to really zoom into the top left corner.
> 
> I have tried dozens of things to get the border cells to display, with the 
> latest being to set a 'valid_min' value for zeta - which didn't work.
> 
> Ncview has also been helpful and even though it only plots the computational 
> domain, it does report the correct I,J and X,Y when you run the cursor over 
> the grid cells.
> 
> Kindest regards,
> Tim


Ticket Details
===================
Ticket ID: KQC-444299
Department: Support netCDF Java
Priority: Normal
Status: Open