Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
Hi Frank, If my memory serves me right the current code allows to deal with different 2D grids of the same type (let's say multiple set of GeoX/GeoY grids), provided that multiple dimensions exist for them, in order to define the coordinates variables. Your case is different and your lon/lat variables will be handled as standard variables (not a coordinate variable, as X and Y will do instead). So I think that the culprit on your case is still the WKT not being properly setup on the initial dataset since it uses quotes instead of double quotes, which result in the reader being unable to recognize the coordinate reference system for the underlying variables. WKT spatial reference definition uses double quotes for texts. Since I was documenting myself on that, I have also looked for some formal documentation and I have found the specification: https://portal.opengeospatial.org/files/12-063r5 Chapter 6.1 of the WKT specification reports: *EXAMPLE KEYWORD1[attribute1,KEYWORD2[attribute2,attribute3]] * *Attributes may be from an enumeration, be numbers or be text. Text is enclosed in double quotes * Indeed, the classic: https://spatialreference.org/ref/epsg/4326/prettywkt/ GEOGCS["WGS 84", DATUM["WGS_1984", SPHEROID["WGS 84",6378137,298.257223563, AUTHORITY["EPSG","7030"]], AUTHORITY["EPSG","6326"]], PRIMEM["Greenwich",0, AUTHORITY["EPSG","8901"]], UNIT["degree",0.01745329251994328, AUTHORITY["EPSG","9122"]], AUTHORITY["EPSG","4326"]] And same for custom CRS defined in NetCDF. This is the output from a netcdf sample provided in GDAL datasets ( https://github.com/OSGeo/gdal/raw/master/autotest/gdrivers/data/cf_lcc2sp.nc ): PROJCS["unnamed",GEOGCS["unknown",DATUM["unknown",SPHEROID["Sphere",6371200,0]],PRIMEM["Greenwich",0],UNIT["degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic_1SP"],PARAMETER["latitude_of_origin",25],PARAMETER["central_meridian",265],PARAMETER["scale_factor",1],PARAMETER["false_easting",0],PARAMETER["false_northing",0],PARAMETER["standard_parallel_1",25],UNIT["kilometre",1000,AUTHORITY["EPSG","9036"]]] And this is how spatial_ref from your sample data is defined PROJCS['unnamed',GEOGCS['GRS 1980(IUGG,1980)',DATUM['unknown',SPHEROID['GRS80',6378137,298.257222101],TOWGS84[0,0,0,0,0,0,0]],PRIMEM['Greenwich',0],UNIT['degree',0.0174532925199433]],PROJECTION['Transverse_Mercator'],PARAMETER['latitude_of_origin',0],PARAMETER['central_meridian',18],PARAMETER['scale_factor',1],PARAMETER['false_easting',15],PARAMETER['false_northing',0],UNIT['Meter',1]] which uses single quotes instead of required double quotes. Hope this helps. Regards, Daniele On Tue, Aug 27, 2019 at 8:51 AM Van Der Stelt Frank < frank.vanderst...@smhi.se> wrote: > Hi Daniele, > > When I open ToolsUI I get five axisTypes, one for *time* and the other > four for location as you suspected; l*on, lat, GeoX *and* GeoY.* > You mention that there is code to deal with this case, is that already > implemented in the NetCDF-plugin? > > Thank you for your help! > > Kind regards, > > Frank > > -- > *Från:* Daniele Romagnoli [daniele.romagn...@geo-solutions.it] > *Skickat:* den 26 augusti 2019 16:39 > *Till:* Van Der Stelt Frank > *Kopia:* Alexander Petkov; geoserver-users@lists.sourceforge.net > *Ämne:* Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not > list layers > > Hi Frank, > not 100% sure about the longitude variable issue. > I'm wondering if, for any reason, the datasets looks like having 4 2D Axis > (both longitude,latitude and x,y projected coordinate axis) so it's like > having both meter and degree coordinates being defined and supported (some > very peculiar datasets contain multiple 2D grids definitions and multiple > 2D dimensions and there is code to deal with that case). > If you open the dataset with ToolsUI and go through the "CoordSys" tab, > how many "axisType" rows is reporting on the bottom table? > > The alternative hypothesis I have in mind is that being the dataset a > projected one, the lon and lat variables are handled as normal 2D variables > (and not as coordinates since there is no dimensions for them). Being "lon" > the first variable listed in your NetCDF, it's the first one spotting the > WKT parsing problem when looking for the associated coordinate reference > system during the variables scan. > > Please, let us know. > Regards, > Daniele > > > > On Thu, Aug 22, 2019 at 10:57 PM Van Der Stelt Frank < >
Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
Hi Daniele,
When I open ToolsUI I get five axisTypes, one for time and the other four for
location as you suspected; lon, lat, GeoX and GeoY.
You mention that there is code to deal with this case, is that already
implemented in the NetCDF-plugin?
Thank you for your help!
Kind regards,
Frank
Från: Daniele Romagnoli [daniele.romagn...@geo-solutions.it]
Skickat: den 26 augusti 2019 16:39
Till: Van Der Stelt Frank
Kopia: Alexander Petkov; geoserver-users@lists.sourceforge.net
Ämne: Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
Hi Frank,
not 100% sure about the longitude variable issue.
I'm wondering if, for any reason, the datasets looks like having 4 2D Axis
(both longitude,latitude and x,y projected coordinate axis) so it's like having
both meter and degree coordinates being defined and supported (some very
peculiar datasets contain multiple 2D grids definitions and multiple 2D
dimensions and there is code to deal with that case).
If you open the dataset with ToolsUI and go through the "CoordSys" tab, how
many "axisType" rows is reporting on the bottom table?
The alternative hypothesis I have in mind is that being the dataset a projected
one, the lon and lat variables are handled as normal 2D variables (and not as
coordinates since there is no dimensions for them). Being "lon" the first
variable listed in your NetCDF, it's the first one spotting the WKT parsing
problem when looking for the associated coordinate reference system during the
variables scan.
Please, let us know.
Regards,
Daniele
On Thu, Aug 22, 2019 at 10:57 PM Van Der Stelt Frank
mailto:frank.vanderst...@smhi.se>> wrote:
Thank you for your answer. I looked at this and found that this is a problem in
geotools. When the WKT string is parsed the single quotes are changed to double
quotes and the CRS factory doesn't like that. So I patched this part of
geotools to get around this problem. The question remains why the NetCDF plugin
looks at the lon variable at all. That one is included in the crun variable
that has the projection defined.
Kind regards,
Frank van der Stelt
Från: Alexander Petkov [green...@gmail.com<mailto:green...@gmail.com>]
Skickat: den 22 augusti 2019 19:48
Till: Van Der Stelt Frank
Kopia:
geoserver-users@lists.sourceforge.net<mailto:geoserver-users@lists.sourceforge.net>
Ämne: Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
First, I would trace why this is happening:
org.geotools.coverage.io<http://org.geotools.coverage.io>.netcdf.crs.NetCDFProjection$CRSParser
parseWKT
WARNING: Unable to setup a CRS from the specified WKT:
PROJCS['unnamed',GEOGCS['GRS 1980(IUGG,
1980)',DATUM['unknown',SPHEROID['GRS80',6378137,298.257222101],TOWGS84[0,0,0,0,0,0,0]],PRIMEM['Greenwich',0],UNIT['degree',0.0174532925199433]],PROJECTION['Transverse_Mercator'],PARAMETER['latitude_of_origin',0],PARAMETER['central_meridian',18],PARAMETER['scale_factor',1],PARAMETER['false_easting',15],PARAMETER['false_northing',0],UNIT['Meter',1]]
Then take it from there.
Kind regards,
Alex
On 8/22/19, Van Der Stelt Frank
mailto:frank.vanderst...@smhi.se>> wrote:
> Hi again,
>
> I'll try to rephrase the problem. When uploading a NetCDF file I get the
> "Could not list layers for this store" problem. The file follows the CF 1.6
> convention and can be read by qgis and other NetCDF-readers. I uploaded the
> file here.
>
> https://www.dropbox.com/s/ssanmgozpg16g5p/stockholm_110_historic_crun-monthly_mean.nc?dl=0
>
> The geoserver log-file says:
>
> Caused by: java.io.IOException: java.lang.IllegalArgumentException: Unable
> to find a CRS for the provided variable: lon
>
> I added the WKT-string in a file under the user_projections catalogue.
> Geoserver can find my CRS but for the lon-variable he still insists on the
> projection.
>
> Does anyone have any clue?
>
> Here is the NetCDF-header:
>
>
> netcdf stockholm_110_historic_crun-monthly_mean {
> dimensions:
> x = 110 ;
> y = 110 ;
> time = UNLIMITED ; // (60 currently)
> bnds = 2 ;
> variables:
> float lon(y, x) ;
> lon:standard_name = "longitude" ;
> lon:long_name = "longitude" ;
> lon:units = "degrees_east" ;
> lon:_CoordinateAxisType = "Lon" ;
> lon:grid_mapping = "projection" ;
> float lat(y, x) ;
> lat:standard_name = "latitude" ;
> lat:long_name = "latitude" ;
> lat:units = "degrees_north" ;
> lat:_CoordinateAxisType = "Lat" ;
> lat:grid_mapping = &q
Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
Hi Frank,
not 100% sure about the longitude variable issue.
I'm wondering if, for any reason, the datasets looks like having 4 2D Axis
(both longitude,latitude and x,y projected coordinate axis) so it's like
having both meter and degree coordinates being defined and supported (some
very peculiar datasets contain multiple 2D grids definitions and multiple
2D dimensions and there is code to deal with that case).
If you open the dataset with ToolsUI and go through the "CoordSys" tab, how
many "axisType" rows is reporting on the bottom table?
The alternative hypothesis I have in mind is that being the dataset a
projected one, the lon and lat variables are handled as normal 2D variables
(and not as coordinates since there is no dimensions for them). Being "lon"
the first variable listed in your NetCDF, it's the first one spotting the
WKT parsing problem when looking for the associated coordinate reference
system during the variables scan.
Please, let us know.
Regards,
Daniele
On Thu, Aug 22, 2019 at 10:57 PM Van Der Stelt Frank <
frank.vanderst...@smhi.se> wrote:
> Thank you for your answer. I looked at this and found that this is a
> problem in geotools. When the WKT string is parsed the single quotes are
> changed to double quotes and the CRS factory doesn't like that. So I
> patched this part of geotools to get around this problem. The question
> remains why the NetCDF plugin looks at the lon variable at all. That one is
> included in the crun variable that has the projection defined.
>
> Kind regards,
>
> Frank van der Stelt
>
>
>
>
> Från: Alexander Petkov [green...@gmail.com]
> Skickat: den 22 augusti 2019 19:48
> Till: Van Der Stelt Frank
> Kopia: geoserver-users@lists.sourceforge.net
> Ämne: Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list
> layers
>
> First, I would trace why this is happening:
>
> org.geotools.coverage.io.netcdf.crs.NetCDFProjection$CRSParser parseWKT
> WARNING: Unable to setup a CRS from the specified WKT:
> PROJCS['unnamed',GEOGCS['GRS 1980(IUGG,
>
> 1980)',DATUM['unknown',SPHEROID['GRS80',6378137,298.257222101],TOWGS84[0,0,0,0,0,0,0]],PRIMEM['Greenwich',0],UNIT['degree',0.0174532925199433]],PROJECTION['Transverse_Mercator'],PARAMETER['latitude_of_origin',0],PARAMETER['central_meridian',18],PARAMETER['scale_factor',1],PARAMETER['false_easting',15],PARAMETER['false_northing',0],UNIT['Meter',1]]
>
> Then take it from there.
>
> Kind regards,
> Alex
>
>
> On 8/22/19, Van Der Stelt Frank wrote:
> > Hi again,
> >
> > I'll try to rephrase the problem. When uploading a NetCDF file I get the
> > "Could not list layers for this store" problem. The file follows the CF
> 1.6
> > convention and can be read by qgis and other NetCDF-readers. I uploaded
> the
> > file here.
> >
> >
> https://www.dropbox.com/s/ssanmgozpg16g5p/stockholm_110_historic_crun-monthly_mean.nc?dl=0
> >
> > The geoserver log-file says:
> >
> > Caused by: java.io.IOException: java.lang.IllegalArgumentException:
> Unable
> > to find a CRS for the provided variable: lon
> >
> > I added the WKT-string in a file under the user_projections catalogue.
> > Geoserver can find my CRS but for the lon-variable he still insists on
> the
> > projection.
> >
> > Does anyone have any clue?
> >
> > Here is the NetCDF-header:
> >
> >
> > netcdf stockholm_110_historic_crun-monthly_mean {
> > dimensions:
> > x = 110 ;
> > y = 110 ;
> > time = UNLIMITED ; // (60 currently)
> > bnds = 2 ;
> > variables:
> > float lon(y, x) ;
> > lon:standard_name = "longitude" ;
> > lon:long_name = "longitude" ;
> > lon:units = "degrees_east" ;
> > lon:_CoordinateAxisType = "Lon" ;
> > lon:grid_mapping = "projection" ;
> > float lat(y, x) ;
> > lat:standard_name = "latitude" ;
> > lat:long_name = "latitude" ;
> > lat:units = "degrees_north" ;
> > lat:_CoordinateAxisType = "Lat" ;
> > lat:grid_mapping = "projection" ;
> > float x(x) ;
> > x:standard_name = "projection_x_coordinate" ;
> > x:long_name = "x coordinate of projection" ;
> > x:units = "m" ;
> > x:axis = "X" ;
> > float y(y) ;
> > y:standard_n
Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
Thank you for your answer. I looked at this and found that this is a problem in
geotools. When the WKT string is parsed the single quotes are changed to double
quotes and the CRS factory doesn't like that. So I patched this part of
geotools to get around this problem. The question remains why the NetCDF plugin
looks at the lon variable at all. That one is included in the crun variable
that has the projection defined.
Kind regards,
Frank van der Stelt
Från: Alexander Petkov [green...@gmail.com]
Skickat: den 22 augusti 2019 19:48
Till: Van Der Stelt Frank
Kopia: geoserver-users@lists.sourceforge.net
Ämne: Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
First, I would trace why this is happening:
org.geotools.coverage.io.netcdf.crs.NetCDFProjection$CRSParser parseWKT
WARNING: Unable to setup a CRS from the specified WKT:
PROJCS['unnamed',GEOGCS['GRS 1980(IUGG,
1980)',DATUM['unknown',SPHEROID['GRS80',6378137,298.257222101],TOWGS84[0,0,0,0,0,0,0]],PRIMEM['Greenwich',0],UNIT['degree',0.0174532925199433]],PROJECTION['Transverse_Mercator'],PARAMETER['latitude_of_origin',0],PARAMETER['central_meridian',18],PARAMETER['scale_factor',1],PARAMETER['false_easting',15],PARAMETER['false_northing',0],UNIT['Meter',1]]
Then take it from there.
Kind regards,
Alex
On 8/22/19, Van Der Stelt Frank wrote:
> Hi again,
>
> I'll try to rephrase the problem. When uploading a NetCDF file I get the
> "Could not list layers for this store" problem. The file follows the CF 1.6
> convention and can be read by qgis and other NetCDF-readers. I uploaded the
> file here.
>
> https://www.dropbox.com/s/ssanmgozpg16g5p/stockholm_110_historic_crun-monthly_mean.nc?dl=0
>
> The geoserver log-file says:
>
> Caused by: java.io.IOException: java.lang.IllegalArgumentException: Unable
> to find a CRS for the provided variable: lon
>
> I added the WKT-string in a file under the user_projections catalogue.
> Geoserver can find my CRS but for the lon-variable he still insists on the
> projection.
>
> Does anyone have any clue?
>
> Here is the NetCDF-header:
>
>
> netcdf stockholm_110_historic_crun-monthly_mean {
> dimensions:
> x = 110 ;
> y = 110 ;
> time = UNLIMITED ; // (60 currently)
> bnds = 2 ;
> variables:
> float lon(y, x) ;
> lon:standard_name = "longitude" ;
> lon:long_name = "longitude" ;
> lon:units = "degrees_east" ;
> lon:_CoordinateAxisType = "Lon" ;
> lon:grid_mapping = "projection" ;
> float lat(y, x) ;
> lat:standard_name = "latitude" ;
> lat:long_name = "latitude" ;
> lat:units = "degrees_north" ;
> lat:_CoordinateAxisType = "Lat" ;
> lat:grid_mapping = "projection" ;
> float x(x) ;
> x:standard_name = "projection_x_coordinate" ;
> x:long_name = "x coordinate of projection" ;
> x:units = "m" ;
> x:axis = "X" ;
> float y(y) ;
> y:standard_name = "projection_y_coordinate" ;
> y:long_name = "y coordinate of projection" ;
> y:units = "m" ;
> y:axis = "Y" ;
> int projection ;
> projection:false_easting = 15. ;
> projection:false_northing = 0. ;
> projection:grid_mapping_name = "projection" ;
> projection:latitude_of_projection_origin = 0. ;
> projection:longitude_of_central_meridian = 18. ;
> projection:proj4 = "+proj=tmerc +lat_0=0 +lon_0=18 +k=1
> +x_0=15 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs" ;
> projection:scale_factor_at_central_meridian = 1. ;
> projection:semi_major_axis = 6378137. ;
> projection:inverse_flattening = 298.257222101 ;
> projection:spatial_ref = "PROJCS[\"SWEREF99 18
> 00\",GEOGCS[\"SWEREF99\",DATUM[\"SWEREF99\",SPHEROID[\"GRS 1980\",
> 6378137.0, 298.257222101, AUTHORITY[\"EPSG\",\"7019\"]],TOWGS84[0.0, 0.0,
> 0.0, 0\
> .0, 0.0, 0.0, 0.0],AUTHORITY[\"EPSG\",\"6619\"]],PRIMEM[\"Greenwich\", 0.0,
> AUTHORITY[\"EPSG\",\"8901\"]],UNIT[\"degree\",
> 0.017453292519943295],AXIS[\"Geodetic latitude\", NORTH],AXIS[\"Geodetic
> longitude\",
Re: [Geoserver-users] Geoserver Store NetCDF issue: Could not list layers
First, I would trace why this is happening:
org.geotools.coverage.io.netcdf.crs.NetCDFProjection$CRSParser parseWKT
WARNING: Unable to setup a CRS from the specified WKT:
PROJCS['unnamed',GEOGCS['GRS 1980(IUGG,
1980)',DATUM['unknown',SPHEROID['GRS80',6378137,298.257222101],TOWGS84[0,0,0,0,0,0,0]],PRIMEM['Greenwich',0],UNIT['degree',0.0174532925199433]],PROJECTION['Transverse_Mercator'],PARAMETER['latitude_of_origin',0],PARAMETER['central_meridian',18],PARAMETER['scale_factor',1],PARAMETER['false_easting',15],PARAMETER['false_northing',0],UNIT['Meter',1]]
Then take it from there.
Kind regards,
Alex
On 8/22/19, Van Der Stelt Frank wrote:
> Hi again,
>
> I'll try to rephrase the problem. When uploading a NetCDF file I get the
> "Could not list layers for this store" problem. The file follows the CF 1.6
> convention and can be read by qgis and other NetCDF-readers. I uploaded the
> file here.
>
> https://www.dropbox.com/s/ssanmgozpg16g5p/stockholm_110_historic_crun-monthly_mean.nc?dl=0
>
> The geoserver log-file says:
>
> Caused by: java.io.IOException: java.lang.IllegalArgumentException: Unable
> to find a CRS for the provided variable: lon
>
> I added the WKT-string in a file under the user_projections catalogue.
> Geoserver can find my CRS but for the lon-variable he still insists on the
> projection.
>
> Does anyone have any clue?
>
> Here is the NetCDF-header:
>
>
> netcdf stockholm_110_historic_crun-monthly_mean {
> dimensions:
> x = 110 ;
> y = 110 ;
> time = UNLIMITED ; // (60 currently)
> bnds = 2 ;
> variables:
> float lon(y, x) ;
> lon:standard_name = "longitude" ;
> lon:long_name = "longitude" ;
> lon:units = "degrees_east" ;
> lon:_CoordinateAxisType = "Lon" ;
> lon:grid_mapping = "projection" ;
> float lat(y, x) ;
> lat:standard_name = "latitude" ;
> lat:long_name = "latitude" ;
> lat:units = "degrees_north" ;
> lat:_CoordinateAxisType = "Lat" ;
> lat:grid_mapping = "projection" ;
> float x(x) ;
> x:standard_name = "projection_x_coordinate" ;
> x:long_name = "x coordinate of projection" ;
> x:units = "m" ;
> x:axis = "X" ;
> float y(y) ;
> y:standard_name = "projection_y_coordinate" ;
> y:long_name = "y coordinate of projection" ;
> y:units = "m" ;
> y:axis = "Y" ;
> int projection ;
> projection:false_easting = 15. ;
> projection:false_northing = 0. ;
> projection:grid_mapping_name = "projection" ;
> projection:latitude_of_projection_origin = 0. ;
> projection:longitude_of_central_meridian = 18. ;
> projection:proj4 = "+proj=tmerc +lat_0=0 +lon_0=18 +k=1
> +x_0=15 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs" ;
> projection:scale_factor_at_central_meridian = 1. ;
> projection:semi_major_axis = 6378137. ;
> projection:inverse_flattening = 298.257222101 ;
> projection:spatial_ref = "PROJCS[\"SWEREF99 18
> 00\",GEOGCS[\"SWEREF99\",DATUM[\"SWEREF99\",SPHEROID[\"GRS 1980\",
> 6378137.0, 298.257222101, AUTHORITY[\"EPSG\",\"7019\"]],TOWGS84[0.0, 0.0,
> 0.0, 0\
> .0, 0.0, 0.0, 0.0],AUTHORITY[\"EPSG\",\"6619\"]],PRIMEM[\"Greenwich\", 0.0,
> AUTHORITY[\"EPSG\",\"8901\"]],UNIT[\"degree\",
> 0.017453292519943295],AXIS[\"Geodetic latitude\", NORTH],AXIS[\"Geodetic
> longitude\", E\
> AST],AUTHORITY[\"EPSG\",\"4619\"]],PROJECTION[\"Transverse_Mercator\",
> AUTHORITY[\"EPSG\",\"9807\"]],PARAMETER[\"central_meridian\",
> 18.0],PARAMETER[\"latitude_of_origin\", 0.0],PARAMETER[\"scale_factor\",
> 1.0]\
> ,PARAMETER[\"false_easting\", 15.0],PARAMETER[\"false_northing\",
> 0.0],UNIT[\"m\", 1.0],AXIS[\"Northing\", NORTH],AXIS[\"Easting\",
> EAST],AUTHORITY[\"EPSG\",\"3011\"]]" ;
> float time(time) ;
> time:standard_name = "time" ;
> time:long_name = "time" ;
> time:bounds = "time_bnds" ;
>time:units = "hours since 2006-01-01 00:00" ;
> time:calendar = "standard" ;
> time:axis = "T" ;
> double time_bnds(time, bnds) ;
> float crun(time, y, x) ;
> crun:long_name = "runoff" ;
> crun:units = "mm h-1" ;
> crun:grid_mapping = "projection" ;
> crun:coordinates = "lat lon" ;
> crun:_FillValue = -1.e+20f ;
> crun:missing_value = -1.e+20f ;
> crun:ModelAttributes = "{\'Hydrological Model\': \'HYPE\'}"
> ;
> crun:DataAttributes = "{\"ECV\": \"monthly mean\",
> \"ECV_calculated_on\": \"2017-10-25\"}" ;
>
> // global attributes:
>
