Fix ERA5 DATM shortwave band handling.

[ezhilsabareesh8]

ERA5 ssrd is an accumulated energy field in J m-2, while MOM/CICE require shortwave band fluxes in W m-2. The previous ERA5 datamode incorrectly treated albedo-like stream fields as shortwave flux bands, producing near-zero shortwave reaching MOM/CICE.

This PR:

• Converts ERA5 ssrd to flux with / 3600.
• Constructs Faxa_swvdr, Faxa_swndr, Faxa_swvdf, and Faxa_swndf from fixed physical partitions of ssrd_flux.
• Removes unused ERA5 albedo-like shortwave stream dependencies from the datamode.

Specific notes

CDEPS Issues Fixed (include github issue #): ACCESS-NRI/access-om3-configs#548 (comment)

Any User Interface Changes (namelist or namelist defaults changes): Changes made in datm.streams.xml and generate_xml_datm_era5.py

[anton-seaice]

@NickSzapiro-NOAA - does UFS use the ERA5 datamode in CDEPS ?

We've found these lines to calculate sw in vis/ifrared/direct/diffuse bands don't work and the resulting incoming shortwave is much too low.

https://github.com/ESCOMP/CDEPS/blob/648807485bfc4924eeeb9c45f2b62753611f4dc9/datm/datm_datamode_era5_mod.F90#L415-L420

[NickSzapiro-NOAA]

Yes ... but I think those lines are essentially unused as variables are overwritten in the ! convert J/m^2 to W/m^2 below that.
Ufuk commented some extra context at ESCOMP#376 (comment)

The changes here look similar to ESCOMP#396 , fyi