Changed valid range of alt_refrac and geop_refrac [1 h]

[Ian Culverwell]

https://trac.romsaf.org/ropp/changeset/5427/ropp_src/branches/dev/Share/dmi_trunk_9.0/ropp_io/ropp/ropp_io_types.f90

Original changesets:

Changed valid range of alt_refrac and geop_refrac: https://trac.romsaf.org/ropp/changeset/4921/ropp_src/branches/dev/Share/dmi_trunk_8.1/ropp_io/ropp/ropp_io_types.f90

[Ian Culverwell]

This doesn't cause any downstream problems in ropp_io-dependent modules, so commit change at r5937. (No need to update ROPP IO user guide as it already claims alt_refrac and geop_refrac range from -1 to 150 km. Oops.) Closing ticket.

[Ian Culverwell]

Actually, it must have caused problems in the ropp_io tests, which I evidently didn't notice at the time. Testing of #585 shows that ucar2ropp fails because the new and reference netCDF files have different valid_ranges of alt_refrac, geop_refrac and shum_sigma (the latter being introduced at r5919 in #551).

[Ian Culverwell]

Actually, it's the valid range of shum_sigma, increased to [0, 50] g/kg at r5934 in #586, that the new tests have picked up. (The reference dataset was correctly updated as part of r5919.)

[Ian Culverwell]

This also afftects the reference files for gfz2ropp and ropp2bufr. Update all of them at r5947.

[Ian Culverwell]

It's concerning that I didn't spot these changes before. Perhaps I failed to rebuild ropp_io from scatch, so the changes to ropp_io/ropp/ropp_io_types.f90 weren't picked up. Anyway, it looks OK now, so keep the ticket closed.