Cool, I like that sort of thing. Looks like a few percent for temperature (although the profile is probably pretty well known). I’d have to work out the expected variation over expected deviations from normal ocean salinity.