Hi Terry, I did some digging and found a related issue on the GDAL Trac site [1]. It seems like the problem is the "elevation:_ChunkSizes" attribute: in the original dataset it is "1, 946" and in the NCSS version it is "2631, 946". The ticket indicates that GDAL prefers the former. As it turns out, a recent change to THREDDS addresses this issue by attempting to maintain the old chunk sizes when writing new NetCDF-4 data. As a test, I fired up the most recent development version of THREDDS (4.5.1) [2], and ran your original dataset through NCSS. I've attached the result to this email (busselton-final-ncss4.5.1.nc4). Can you run your code on it to see if the issue persists? Cheers, Christian Ward-Garrison [1] https://trac.osgeo.org/gdal/ticket/5082 [2] http://www.unidata.ucar.edu/software/thredds/current/tds/TDS.html On Thu, Jul 24, 2014 at 9:54 PM, <Terry.Rankine@xxxxxxxx> wrote: > Hi Christian > > > > *I think *Thredds has destroyed my PROJCS data which doesn’t allow my > code to rewarp it into a different projection. > > > > busselton-final-meta.nc-thredds-ncss.nc4 -> subset service “request all” > > busselton-final-meta.nc -> original file from HTTP service. > > > > > http://vhirl-dev.csiro.au/thredds/catalog/dem/catalog.html?dataset=testDatasetScan/busselton-final-meta.nc > > > > > > > > ran110@vhirl-dev:~/testing$ gdalwarp > busselton-final-meta.nc-thredds-ncss.nc4 -t_srs "+proj=latlong > +datum=WGS84" -co "COMPRESS=DEFLATE" -r cubic output.wgs84.cubic.nc -wo > OPTIMIZE_SIZE=TRUE -of netCDF -dstnodata -9999 > > Creating output file that is 1152P x 2597L. > > Warning 5: NOTICE: Format set to NC4C because compression is set to > DEFLATE. > > Processing input file busselton-final-meta.nc-thredds-ncss.nc4. > > Using internal nodata values (e.g. -9999) for image > busselton-final-meta.nc-thredds-ncss.nc4. > > ERROR 1: nBlockYSize = 2631, only 1 supported when reading bottom-up > dataset > > ERROR 1: busselton-final-meta.nc-thredds-ncss.nc4, band 1: IReadBlock > failed at X offset 0, Y offset 0 > > ERROR 1: GetBlockRef failed at X block offset 0, Y block offset 0 > > > > ran110@vhirl-dev:~/testing$ gdalwarp busselton-final-meta.nc -t_srs > "+proj=latlong +datum=WGS84" -co "COMPRESS=DEFLATE" -r cubic > output.wgs84.cubic.nc -wo OPTIMIZE_SIZE=TRUE -of netCDF -dstnodata -9999 > > Creating output file that is 1152P x 2597L. > > Warning 5: NOTICE: Format set to NC4C because compression is set to > DEFLATE. > > Processing input file busselton-final-meta.nc. > > Using internal nodata values (e.g. -9999) for image > busselton-final-meta.nc. > > 0...10...20...30...40...50...60...70...80...90...100 - done. > > ran110@vhirl-dev:~/testing$ > > > > > > HEADER INFO > > > > ran110@vhirl-dev:~/testing$ gdalinfo > busselton-final-meta.nc-thredds-ncss.nc4 > > Driver: netCDF/Network Common Data Format > > Files: busselton-final-meta.nc-thredds-ncss.nc4 > > Size is 946, 2631 > > Coordinate System is: > > PROJCS["unnamed", > > GEOGCS["unknown", > > DATUM["unknown", > > SPHEROID["Spheroid",6378137,298.257223563]], > > PRIMEM["Greenwich",0], > > UNIT["degree",0.0174532925199433]], > > PROJECTION["Transverse_Mercator"], > > PARAMETER["latitude_of_origin",0], > > PARAMETER["central_meridian",117], > > PARAMETER["scale_factor",0.9996], > > PARAMETER["false_easting",500], > > PARAMETER["false_northing",10000], > > UNIT["kilometre",1000, > > AUTHORITY["EPSG","9036"]]] > > Origin = (189.659982684810018,6757.539391574399815) > > Pixel Size = (0.250000000000000,-0.250000000000000) > > > > > > VS > > > > ran110@vhirl-dev:~/testing$ gdalinfo busselton-final-meta.nc > > Driver: netCDF/Network Common Data Format > > Files: busselton-final-meta.nc > > Size is 946, 2631 > > Coordinate System is: > > PROJCS["UTM Zone 50, Southern Hemisphere", > > GEOGCS["WGS 84", > > DATUM["WGS_1984", > > SPHEROID["WGS 84",6378137,298.257223563, > > AUTHORITY["EPSG","7030"]], > > TOWGS84[0,0,0,0,0,0,0], > > AUTHORITY["EPSG","6326"]], > > PRIMEM["Greenwich",0, > > AUTHORITY["EPSG","8901"]], > > UNIT["degree",0.0174532925199433, > > AUTHORITY["EPSG","9108"]], > > AUTHORITY["EPSG","4326"]], > > PROJECTION["Transverse_Mercator"], > > PARAMETER["latitude_of_origin",0], > > PARAMETER["central_meridian",117], > > PARAMETER["scale_factor",0.9996], > > PARAMETER["false_easting",500000], > > PARAMETER["false_northing",10000000], > > UNIT["METERS",1]] > > Origin = (189659.982684810005594,6757539.391574399545789) > > Pixel Size = (250.000000000000000,-250.000000000000000) > > > > > > *From:* Christian Ward-Garrison [mailto:cwardgar@xxxxxxxx] > *Sent:* Friday, 25 July 2014 7:49 AM > *To:* Rankine, Terry (Mineral Resources, Kensington) > *Cc:* thredds@xxxxxxxxxxxxxxxx > *Subject:* Re: [thredds] NetCDF Subset service > > > > Hi Terry, > > You might expect that when you request all of the data in a dataset, NCSS > will simply return the original dataset (no subsetting required). That's > not how NCSS works; instead, it will always create a new dataset, and copy > over the requested data (which could potentially be everything) from the > original. > > In this scenario, you might notice that the original and new datasets are > *structurally* different (i.e. different attributes, scale factors, etc), > but they are *semantically* equal (i.e. they contain the same data, though > it might be represented differently). > > As far as I can tell, your 2 example files are semantically equal. Are you > struggling with structural equality issues? > > Cheers, > > Christian > > > > On Mon, Jul 21, 2014 at 12:46 AM, <Terry.Rankine@xxxxxxxx> wrote: > > Hi Guys > > > > Is there any other users currently using the netCDF subset service? I am > having quite a few issues with it and would like to check ‘it’s not just me’ > > > > This: > > http://vhirl-dev.csiro.au/thredds/fileServer/dem/busselton-final.nc > > > > is very very different to this: > > > http://vhirl-dev.csiro.au/thredds/ncss/grid/dem/busselton-final.nc?var=elevation&north=-29.2729&west=113.5915&east=116.2396&south=-35.2422&horizStride=1&accept=netcdf4 > > > > How are others handling THREDDS and CRS changes? > > > > Regards > > Terry > > > _______________________________________________ > thredds mailing list > thredds@xxxxxxxxxxxxxxxx > For list information or to unsubscribe, visit: > http://www.unidata.ucar.edu/mailing_lists/ > > >
Attachment:
busselton-final-ncss4.5.1.nc4
Description: Binary data
thredds
archives: