[netcdf-java] GeoTiff conversion problems

• To: Java NetCDF <netcdf-java@xxxxxxxxxxxxxxxx>
• Subject: [netcdf-java] GeoTiff conversion problems
• From: Steve Ansari <Steve.Ansari@xxxxxxxx>
• Date: Wed, 13 Aug 2008 17:09:14 -0400

Hello John, Ethan, Yuan, Anyone!

I'm having problems converting from NetCDF to GeoTIFF using the
'GeotiffWriter' class.  I've attached some example code that will
convert each 'GridDatatype' in a NetCDF file to a GeoTIFF.  However, the
resulting 32-bit GeoTIFFs will not display in ArcGIS (no data is found)
or by GDAL/FWTools (listgeo.exe and gdalinfo.exe both crash).  Note that
the files I'm testing with don't have time or level dimensions.

I've tested with 2.2.20, 2.2.22 and 4.0 and get the same results.  I'm
positive that this used to work.  It actually does work in the Java
NEXRAD Tools, which uses an older version of NCJ 2.2.

I've tested the grayscale option, which works with ArcGIS, but still
bombs with GDAL (listgeo.exe).

Thanks,
Steve


import java.util.List;

import ucar.nc2.dt.GridDataset;
import ucar.nc2.dt.GridDatatype;
import ucar.nc2.geotiff.GeotiffWriter;


public class Nc2GTiff {

    /**
     * @param args
     */
    public static void main(String[] args) {
       
        try {
           
            if (args.length != 1 || ! args[0].endsWith(".nc")) {
                System.err.println("Input argument must be a NetCDF
(.nc) file!");
                return;
            }
           
            String infile = args[0];
           
            GridDataset dataset = ucar.nc2.dt.grid.GridDataset.open(infile);
            List<GridDatatype> grids = dataset.getGrids();
            for (GridDatatype grid : grids) {
               
                String gridName = grid.getName();
                String outfile = infile.substring(0,
infile.length()-3)+"_"+gridName+"-4.0.tif";
                GeotiffWriter gtWriter = new GeotiffWriter(outfile);
                gtWriter.writeGrid(dataset, grid, grid.readDataSlice(-1,
-1, -1, -1), false);
               
            }
            dataset.close();
           
        } catch (Exception e) {
            e.printStackTrace();
        }
       
    }

}