BB4CMIP: test robustness of (a) harmonization estimate and (b) patternscaling spatial grid - using 10 or 30 (or other) means

[jkikstra]

Some things to explore (using GFED4.1s now, or ideally BB4CMIP later), for the harmonized sectors:

• how much, and where (spatially), for which gases, is there a difference for 10yr vs 30yr mean

need to think more about how to exactly create useful diagnostic indicators to constrain what the best pattern would be to use for CMIP7.

could for instance do:

• recording diff between many 10 yr means, then between many 20 yr means, etc.