Open ADCollard opened 2 years ago
@ADCollard what observation type(s) have you noticed this for? I'm a bit surprised at this to be honest, I wonder if it is a bug in the combine script rather than it not being written out by GSI.
@CoryMartin-NOAA We noticed this when looking at Praveen's sonde virtual temperature evaluation. I interpreted it as GSI not doing a geoval calculation for Tv/Tsens within the ps code if there is no ps observation. (We did produce a plot showing this correlation but I will need to reproduce it).
I bet this is true for Tv, but Tsen should be there (I think). I'm a bit on the fence about writing out Tv also, I feel like UFO should be able to convert Tsen + q to Tv, no?
When running the GSI to create geovals, the virtual and sensible temperatures are read from the surface pressure diag file. This results in the geovals only being reported where surface pressure is present.
We need to change the GSI and converter code so that virtual and sensible temperatures are output explicitly from setupt.f90 and used in the geoval creation.