Suggestions on auto-mapping using GPT

Be careful what you wish for!

With free and open source software there are often problems of creeping feature overload without the sustained effort to maintain features. This is a problem for new users who see a feature mentioned on the internet but can’t get it to work because it has not been maintained. In the long run, sticking to core functions makes complex systems easier to maintain and use. This does leave a problem of identifying 3rd party tools to fill in the gaps.

SNAP and GPT allow you to export mapped images to NetCDF4-CF format grids. There are many 3rd party tools that can make high-quality maps with annotations, colorbars, vector overlays, etc. If you convert you template to an image with transparent areas for the data, you and composite the template with with an image derived from a NetCDF4-CF format grid (e.g., using ImageMagick). I’d prefer that the SNAP developers focus on capabilities that aren’t available from general purpose mapping and image processing tools. In practice, SNAP/GPT do take advantage of 3rd party libraries, so in some cases, useful features should be added by the library developers. Again, this does create a problem for users to understand which limitations are inherited from 3rd party libraries.

NASA SeaDAS 7 is based on BEAM, but did add an improved color manager combined with metadata in standard NASA OBPG products so the default behaviour provides consistent color scales across a series of images. I’m not sure if the CF conventions address color scales so such metadata would be used by external tools. This is an example of
an enhancement that would be nice to see in 3rd party tools and is best be done thru metadata conventions.

1 Like