Default gdirs should use fixed_geometry_spinup_yr when no spinup is used

[fmaussion]

All is in the title. IT would allow to avoid relying on the longstanding and tedious statistical reconstruction that was used until now

[pat-schmitt]

This is already the case, if I understand the request correctly. The runs with filesuffix '_spinup_historical' fallback to the fixed_geometry_spinup_yr. Therefor, all gdirs historical runs (with no other error) start at 1979.

Or do you have an example where this is not the case?

However, we could think of deleting the '_historical' run from the gdirs. These are the runs which use always the fixed_geometry_spinup, and were the default prior to v1.6, and I only kept those for comparison if the dynamic spinup is useful at all.

[fmaussion]

The _historical files in the glacier directories do not use the fixed geometry spinup I think? I wanted to compare spinup with historical and I couldn't without re-running historical. This is the line where we would need to use it:

oggm/oggm/cli/prepro_levels.py

Lines 706 to 709 in 5ac2eea

	workflow.execute_entity_task(tasks.run_from_climate_data, gdirs,
	min_ys=y0, ye=ye,
	output_filesuffix='_historical')
	# Now compile the output