Automatically, of course as part of the pipeline. You just upload a container script or whatever to the super-computing center at the NCHC. Then, you automatically have access to lots and lots of the environmental data without having to download it, because it’s impractical.