Update ILAMBv2 dataset for v3: FLUXCOM

[nocollier]

There are a lot of choices, but this is what we did in v2:

Data generated by Artificial Neural Networks and forced with CRUNCEPv6 meteorological data and MODIS (RS+METEO)
ftp://ftp.bgc-jena.mpg.de/pub/outgoing/FluxCom/CarbonFluxes_v1_2017/RS+METEO/CRUNCEPv6/raw/monthly/
ftp://ftp.bgc-jena.mpg.de/pub/outgoing/FluxCom/EnergyFluxes/RS_METEO/member/CRUNCEP_v8/monthly/"

Files we used looked like:

LE.RS_METEO.EBC-NONE.MLM-ANN.METEO-CRUNCEP_v8.720_360.monthly.1981.nc

• we should include the cell measures as I did in v2 but also include them properly in the metadata
• we should be more clear in the metadata about which versions we picked