For Weather Quest (and in general) it would be useful to have 2 meter air temperature as an output diagnostic. We should compute this using MO similarity theory using surface fluxes.
We should also not use ERA5 surface pressure to extrapolate to other pressures, instead we should start from a known geopotential height and pressure level and work from there.