Hi I have been using the land sea mask after importing a vector shape file of lake shapes then using it to mask out the land. This works fine in SNAP desktop but under the gpt I get this error
“Error: [NodeId: Import-Vector] The Product ‘A_MSIL1C_20170404T115401_N0204_R023_T29UPB_20170404T115356’ already contains a band with the name ‘latitude’.”
The full error is much more extensive but I think the last line definitely outlines the problem. The input file is the standard .nc output from the polymer atmospheric correction and does contain a latitude band and a longitude band but when masking in the desktop this causes no issues is there any option that I am missing in the gpt to allow it to import the vector file or is this a limitation of the gpt. I have thousands of images to process so using the desktop is not a option. Any advice would be greatly appreciated.
This is indeed a bug in the Import-Vector operator.
At the moment there is only workaround I could think of.
You first need to reproject the product, e.g. to WGS84 or UTM. Afterwards it should be possible to import the shapefile. But this is probably not what you want.
To release a fix for this might take some weeks.
Thank you for that marpet. I had attempted the fix using reproject but it is very slow the reproject function does not make use of all available resources or much more than 10% across cpu ram and disk. That is with altering the vm options so it was taking around 6day to go through 250 images. My new fix is to subset from the .nc into geotiff but subsetting out the lat lon bands removes the error but retains the positioning and from there perform my process chain then reconvert back into .nc for matlab. Even with the file type changes the time is around 2min per image.
I have another very curious problem when I go to mosaic but I will create a new topic for that.