global_energy_fixer postblock (gen2) only supports explicit up/down fluxes — no net-flux option, and docs are gen1-only

[kjmayer]

Summary

The gen1 global energy fixer postblock (credit/postblock/gen1.py::GlobalEnergyFixer) accepts net TOA and surface radiation fluxes (e.g. FSNT, FLNT, FSNS, FLNS), matching what's documented at https://miles-credit.readthedocs.io/en/latest/postblock.html. The gen2 implementation (credit/postblock/conservation.py::GlobalEnergyFixerUpDown) only supports explicit up/down flux decomposition (toa_up_solar_var, toa_up_olr_var, surf_down_solar_var, surf_up_solar_var, surf_down_lw_var, surf_up_lw_var, etc.) — there's no net-flux equivalent registered for gen2.

This means:

1. Users can't use net-flux output with the gen2 energy fixer without manually decomposing into up/down components first.
2. The registry key global_energy_fixer in credit/postblock/init.py is aliased to GlobalEnergyFixerUpDown, so a gen1-style config using that key (with net-flux variable names) will fail or silently misconfigure rather than clearly erroring

Proposed fix

• Port a net-flux-compatible energy fixer (GlobalEnergyFixer, mirroring gen1's class) into credit/postblock/conservation.py for gen2, alongside the existing GlobalEnergyFixerUpDown.
• Register it under a distinct key (e.g. global_energy_fixer_net) so it doesn't collide with the up/down variant, and consider renaming/aliasing the existing global_energy_fixer key more explicitly

Docs

The postblock readthedocs page currently only shows the gen1/net-flux config example under "Global Energy Fixer." Once gen2 supports both methods, the docs should be updated to show both config examples (net-flux vs. up/down) and clarify which config keys map to which method for gen2.

Environment

• Repo: NCAR/miles-credit, main branch
• Relevant files: credit/postblock/gen1.py, credit/postblock/conservation.py, credit/postblock/init.py