Re-gridding CABLE fluxes to GEOS-Chem compatible files
As flagged in the CABLE page, CABLE fluxes are currently provided for land points only (makes sense), but without obvious mapping to lat/long co-ordinates. Before working with them I will re-grid to something more reasonable (or somethings... not sure yet whether to go for multiple options, or aggregate later).
The routine 'fill_nan.ncl' doesn't actually appear to do anything on the current files. Presumably it could be useful if the NaN values had to be set specifically depending on the program used to read the files.
The lat and lon that correspond to each value of the array in the file (i.e. the indices of the spatial variable that corresponds to the land points) are contained in cable_latlon.csv. We'll use this to read the existing files, and re-write them to a regular grid.