Inconsistent IGBP types

[jaideep777]

The IGBP_veg_short field has inconsistent info, i.e., it sometimes has full forms fo the vegetation type and sometimes short forms.

Proposed fix:

Use the following code (or something to that effect) to remap the full-forms while generating the NetCDF files

IGBP_veg_short = dplyr::case_match(IGBP_veg_short,
                                 "Woody Savannas" ~ "WSA",
                                 "Savannas" ~ "SAV",
                                 "Permanent Wetlands" ~ "WET",
                                 "Open Shrublands" ~ "OSH",
                                 "Closed Shrublands" ~ "CSH",
                                 "Grasslands" ~ "GRA",
                                 "Evergreen Needleleaf Forest" ~ "ENF",
                                 "Evergreen Broadleaf Forest" ~ "EBF",
                                 "Deciduous Broadleaf Forest" ~ "DBF",
                                 "Mixed Forest" ~ "MF",
                                 "Croplands" ~ "CRO",
                                 "Cropland/Natural Vegetation Mosaic" ~ "MF",
                                 "Urban and Built-Up" ~ NA,
                                 .default = IGBP_veg_short)

[stineb]

We have code in ingestr for preparing site-level meta data here. This may be adopted to solve this issue. It takes information from {amerifluxr} and from Falge et al.. We don't need all of these meta info, but I recommend the following:

• Make sure to use IGBP vegetation type provided as meta-info for the site (available through {ameriflux} and Falge et al.), and not extracted from global maps
• Koeppen-Geiger climate zone, to extract from global map by Beck et al. (file contained in {ingestr}, see here)
• Complement missing information on elevation using ETOPO1
• Revise start and end year based on available data (and not file name)

We will need more info for the following points. Ask Yanghui Kang (yanghuikang@berkeley.edu) and Trevor Keenan (trevorkeenan@berkeley.edu). They collected more. What do they have? We definitely need:

• Presence of C4 grasses (logical)

[khufkens]

• Add script to clean up meta-data
• don't run whole routine again
• ..

[stineb]

Resolved by https://github.com/geco-bern/FluxDataKit/pull/72