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

Re: problems with multiple grids in McIDAS files

Robb-
You'll need to add in the McIDASGridServiceProvider through the nj22Config.xml to be able to use it. It's not in NetcdfFile (I'm waiting for this issue to get fixed). Here's what I use: 


<?xml version="1.0"?>
<nj22Config>
<ioServiceProvider  class="ucar.nc2.iosp.gempak.GempakSurfaceIOSP"/>
</nj22Config>
for toolsUI (in my ~/.unidata directory). You'll need the fixes I checked in this morning in iosp.mcidas or else you won't be able to open GEMPAK files once you add the above. 

Unless you are changing GridIndex, this shouldn't be a problem.
There are multiple projections in the file, but each is in it's
own group.  I think the netCDF file specs are okay, but that the
CoordSysBuilder is picking up the projection type from the first
one in the file.  It looks like they all end up as Lambert Conformal.
For a single projection in a file, it works fine.  I don't have
any samples though. This has been a long standing problem (although it used to be worse), so I don't think it's anything you've changed. 


Don

Robb Kambic wrote:

On Thu, 18 Jun 2009, Don Murray wrote:


John/Robb-

I'm trying to track down an issue with the projections in the
McIDAS Grid IOSP.  The Lambert Conformal grids seem to be
okay, but the PolarStereoGraphic and Mercator projections in
the same file are not correct in the IDV or ToolsUI GridViewer.
So, it's either a bad definition or a problem with the CoordSysBuilder.

I'm comparing the 49x35 PS grids in testdata/grid/mcidas/GRID5510 with


Don,
I can't open any of the mcidas files in that directory. Since I'm working on the new index, I thought maybe that was the problem but I can open the files in the grid/gempak directory ok. Also, the NAM_ALaska-95km opens ok too. I'll have to do more research to see if the problem is my code, but I've been working to make all the index codings backwards compatiable. One thought is that you have multiple projections in the same file, that could cause problems too. Do any of the files in the mcidas directory have just one projection? Are the projections correct? 

Robb...



the NAM Alaska 95 km grid (e.g.:
dods://motherlode.ucar.edu:8080/thredds/dodsC/model/NCEP/NAM/Alaska_95km/NAM_Alaska_95km_20090618_1200.grib1
The projection varaibles seem to have the same parameters that are used in GridHorizCoordSys, although the McIDAS projection has more info (La2, Lo2, Latin1, Latin2, etc), but that is not used in GridHorizCoordSys.makePS: 

GRID5510:

    char PS_X:49_Y:35;
      :grid_mapping_name = "polar_stereographic";
      :longitude_of_projection_origin = -150.0; // double
      :straight_vertical_longitude_from_pole = -150.0; // double
      :scale_factor_at_projection_origin = 0.933; // double
      :latitude_of_projection_origin = 90.0; // double
      :earth_shape = "Spherical";
      :GDS_param_Dx = 95250.0; // double
      :GDS_param_Dy = 95250.0; // double
      :GDS_param_GDSkey = "PS X:49 Y:35";
      :GDS_param_La1 = 42.07661981300264; // double
      :GDS_param_La2 = 63.970746163829524; // double
      :GDS_param_Latin1 = 60.0; // double
      :GDS_param_Latin2 = 60.0; // double
      :GDS_param_Lo1 = -175.6410058243053; // double
      :GDS_param_Lo2 = -93.69006752597979; // double
      :GDS_param_LoV = -150.0; // double
      :GDS_param_Nx = 49; // int
      :GDS_param_Ny = 35; // int
      :GDS_param_Proj = 2; // int
      :GDS_param_ProjFlag = "PS";
      :_CoordinateTransformType = "Projection";
      :_CoordinateAxisTypes = "GeoX GeoY";

Alaska 95 km:

  String Polar_Stereographic_projection_Grid;
    :grid_mapping_name = "polar_stereographic";
    :longitude_of_projection_origin = -150.0; // double
    :straight_vertical_longitude_from_pole = -150.0; // double
    :scale_factor_at_projection_origin = 0.933; // double
    :latitude_of_projection_origin = 90.0; // double
    :earth_shape = "spherical";
    :GRIB_param_Dx = 95250.0; // double
    :GRIB_param_Dy = 95250.0; // double
    :GRIB_param_GDSkey = 464455425; // int
    :GRIB_param_La1 = 42.085; // double
    :GRIB_param_Lo1 = -175.641; // double
    :GRIB_param_LoV = -150.0; // double
    :GRIB_param_NpProj = "true";
    :GRIB_param_Nx = 49; // int
    :GRIB_param_Ny = 35; // int
    :GRIB_param_ProjFlag = 0; // int
    :GRIB_param_ResCompFlag = 8; // int
    :GRIB_param_Winds = "Relative";
    :GRIB_param_grid_name = "Polar Stereographic projection Grid";
    :GRIB_param_grid_radius_spherical_earth = 6367.47; // double
    :GRIB_param_grid_shape = "spherical";
    :GRIB_param_grid_shape_code = 0; // int
    :GRIB_param_grid_type = 5; // int
    :GRIB_param_grid_units = "m";
    :_CoordinateTransformType = "Projection";
    :_CoordinateAxisTypes = "GeoX GeoY";

The big difference is that the McIDAS grid has groups.

Don
*************************************************************
Don Murray                               UCAR Unidata Program
address@hidden                        P.O. Box 3000
(303) 497-8628                              Boulder, CO 80307
http://www.unidata.ucar.edu/staff/donm
*************************************************************
=============================================================================== 
Robb Kambic                       Unidata Program Center
Software Engineer III               Univ. Corp for Atmospheric Research
address@hidden           WWW: http://www.unidata.ucar.edu/
=============================================================================== 

--
*************************************************************
Don Murray                               UCAR Unidata Program
address@hidden                        P.O. Box 3000
(303) 497-8628                              Boulder, CO 80307
http://www.unidata.ucar.edu/staff/donm
*************************************************************