On Sat, Nov 24, 2018 at 5:07 PM Even Rouault <even.roua...@spatialys.com> wrote: > > On samedi 24 novembre 2018 16:27:33 CET Markus Metz wrote: > > I have a NetCDF file with west > east, gdalinfo reports > > Corner Coordinates: > > Upper Left ( 335.257, 70.000) > > Lower Left ( 335.257, 30.000) > > Upper Right ( 44.7424710, 69.9999970) > > Lower Right ( 44.7424710, 29.9999954) > > Center ( 190.000, 50.000) > > > > west = 335.257 > east = 44.7424710 > > > > The reason is that west has been wrapped such that it is in the range [0, > > 360], and GDAL thinks that columns need to be read from right to left > > instead of left to right. > > > > The fix is to convert west with west -= 360, then assign corrected extents > > with gdal_translate -a_ullr > > > > There is a config option GDAL_NETCDF_BOTTOMUP. Could something similar be > > implemented for west > east, something like GDAL_NETCDF_LEFTISLEFT? Or is > > it a corrupted file and manual correction is the only option? > > Couldn't find anything related mentionned in the CF convention, but that > doesn't make that file illegal. I guess that could be auto-corrected by > detecting there's a discontinuity in the values of the longitude variable.
This particular NetCDF file, PM10 analysis from CAMS http://www.regional.atmosphere.copernicus.eu/, does not have a longitude variable, also no subdatasets. Resolution should be 0.1 degrees, but GDAL reports Pixel Size = (-0.415021467959250,-0.100000003824258) because west > east full gdalinfo output without band info: Warning 1: No UNIDATA NC_GLOBAL:Conventions attribute Driver: netCDF/Network Common Data Format Files: W_fr-meteofrance,MODEL,ENSEMBLE+ANALYSIS+SURFACE+PM10+-24H-1H_C_LFPW_20181120000000.nc Size is 700, 400 Coordinate System is `' Origin = (335.257498526948382,69.999996950154312) Pixel Size = (-0.415021467959250,-0.100000003824258) Metadata: latitude#long_name=latitude latitude#units=degrees_north level#long_name=level level#units=m longitude#long_name=longitude longitude#units=degrees_east NC_GLOBAL#ANALYSIS=Europe, 20181119+[0H_23H] NC_GLOBAL#comment=Tools used: CDO V1.6.1rc6 and NCO V4.3.7 NC_GLOBAL#history=Model ENSEMBLE ANALYSIS NC_GLOBAL#institution=Data produced by Meteo France NC_GLOBAL#nco_openmp_thread_number=1 NC_GLOBAL#project=MACC-RAQ (http://macc-raq.gmes-atmosphere.eu) NC_GLOBAL#source=Data from ENSEMBLE model NC_GLOBAL#summary=ENSEMBLE model hourly ANALYSIS of PM10 concentration at the Surface from 20181119+[0H_23H] on Europe NC_GLOBAL#title=PM10 Air Pollutant ANALYSIS at the Surface NETCDF_DIM_EXTRA={time,level} NETCDF_DIM_level_DEF={1,5} NETCDF_DIM_level_VALUES=0 NETCDF_DIM_time_DEF={24,5} NETCDF_DIM_time_VALUES={0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23} pm10_conc#species=PM10 Aerosol pm10_conc#standard_name=mass_concentration_of_pm10_ambient_aerosol_in_air pm10_conc#units=µg/m3 pm10_conc#value=hourly values time#long_name=ANALYSIS time from 20181119 time#units=hours Corner Coordinates: Upper Left ( 335.257, 70.000) Lower Left ( 335.257, 30.000) Upper Right ( 44.7424710, 69.9999970) Lower Right ( 44.7424710, 29.9999954) Center ( 190.000, 50.000) Band 1 Block=700x1 Type=Float32, ColorInterp=Undefined ...
_______________________________________________ gdal-dev mailing list gdal-dev@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/gdal-dev
