I need to think about this. For some reason I remember one of the test cases being just like this although with not so many bands. As the cobwebs clear I believe this was started but not completed. There's even a BANDSITEM member of the layer object defined in the source. This would allow definition of a column to map a temporal value to a particular band within a larger image. I'm thinking it's probably not far from functioning.
Steve >>> On 2/19/2008 at 10:08 AM, in message <[EMAIL PROTECTED]>, Frank Warmerdam <[EMAIL PROTECTED]> wrote: > vincent wrote: >> Hello, >> >> I would like to create Web Coverage Service from one netCDF file in >> following http://mapserver.gis.umn.edu/docs/howto/WCSServerFormatHowTo >> >> My netCDF store rainfall matrix every five minutes in the "pluie" >> variable. This the CDL : >> >> dimensions: >> >> time = UNLIMITED; // (496 currently) // (has coord.var) >> x = 50; // (has coord.var) >> y = 50; // (has coord.var) >> variables: >> int Lambert_Conformal; >> :grid_mapping_name = "lambert_conformal_conic"; >> :standard_parallel = 43.20317, 44.99683; // double >> :longitude_of_central_meridian = 2.337229; // double >> :latitude_of_projection_origin = 44.1; // double >> :false_easting = 600.0; // double >> :false_northing = 3200.0; // double >> :_CoordinateTransformType = "Projection"; >> :_CoordinateSystems = "ProjectionCoordinateSystem"; >> :_CoordinateAxes = "y x"; >> :_CoordinateAxisTypes = "GeoX GeoY"; >> * double pluie(time=496, y=50, x=50);* >> * :units = "mm/h";* >> * :long_name = "Intensité de pluie";* >> * :coordinates = "lat lon";* >> * :grid_mapping = "Lambert_Conformal";* >> * :_CoordinateSystems = "ProjectionCoordinateSystem > LatLonCoordinateSystem";* > > Vincent, > > Because GDAL treats this array as a single dataset having 496 bands, I don't > believe there is a way to properly create a tileindex with one entry per > time slice, referring back directly to that band though this is something > that > I think was discussed at one point. > > You *can* serve the pluie array, but it will be represented as having 496 > bands along a dimension called Band instead of time. > > Hopefully Steve will correct me if I am wrong. What would be nice > is a way to have WCS (and raster rendering in general) potentially > override the BANDS= processing option with information read from a > tile index, and then update the gdaltindex program accordingly to > generate such results. This shouldn't - in theory - be too hard to > implement. > > Best regards, > -- > ---------------------------------------+------------------------------------ > -- > I set the clouds in motion - turn up | Frank Warmerdam, > [EMAIL PROTECTED] > light and sound - activate the windows | http://pobox.com/~warmerdam > and watch the world go round - Rush | President OSGeo, http://osgeo.org > > _______________________________________________ > mapserver-users mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/mapserver-users _______________________________________________ mapserver-users mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/mapserver-users
