Bug report #21822
Cannot load multivariable netcdf files
|Affected QGIS version:||3.4.6||Regression?:||Yes|
|Operating System:||ubuntu 18.04||Easy fix?:||No|
|Pull Request or Patch supplied:||No||Resolution:|
|Crashes QGIS or corrupts data:||No||Copied to github as #:||29637|
I upgraded QGis and I cannot load some netcdf anymore. All the file multivariable that I have cannot be read.
The netcdf that I can load do not trigger the window were to choose the variable as before but directly load the single variable present in the files.
When I try to load a multivariable netcdf Qgis report on the GUI that source is not valid and on terminal I am getting only warnings relative to the naming of dimensions
Warning 1: dimension #2 (x) is not a Longitude/X dimension.
Warning 1: dimension #1 (y) is not a Latitude/Y dimension.
Warning 1: dimension #0 (Year) is not a Time or Vertical dimension.
so no critical error.
When I am using gdalinfo is all fine.
when I am using gdal_translate to transform a single variable in geotiff is all fine and am getting a tiff file correctly read by qgis itself.
#1 Updated by saverio vicario about 2 years ago
when from my python script were I generate the netcdf I select a single variable from my mutivariable dataset and produce a new netcdf files, this is loaded no problem inside qgis.
So the problem is really that qgis 3.4.6 does not produce a pop up window were used can select the variable of interest to be loaded.
#3 Updated by Giovanni Manghi about 2 years ago
- Regression? changed from No to Yes
- Priority changed from Normal to High
- Subject changed from Cannot load multivariable netcdf files into QGIS 3.4.6 that were loading in QGIS 2.8 to Cannot load multivariable netcdf files
Also possibly duplicate of #21783
#5 Updated by saverio vicario about 2 years ago
- File Testy.nc added
I solved partially the problem.
I changed the name of dimensions from x to X, from y to Y, and from Years to Time and Qgis recognize my multivariable netcdf as something to be read by MDAL. So I got my visualization,albeit always with an error message. So I would like to change the report to "feature request": the error message should more clear and state if some variable are not recognized or at least report the warning of gdalinfo that guided me to make the changes. I attach a smaller data set so that you can read the error message that stay even if now the data set is read ( probably some other metadata are asked)