Latest land use timeseries output crashes fates LUH2 testing on izumi

[glemieux]

Cross posting ESCOMP/CTSM#2653 here.

The issue notes that the older timeseries file that was generated when this tool was part of the fates repo back on fates tag sci.1.68.3_api.31.0.0_tools.1.0.1. There have been some updates to the tool between now and then which may have caused an issue.

One avenue of investigation is to determine why the current output can not be converted to netcdf classic format, which was possible with the older fates tooling version.

This issue also highlights the need for regression testing as captured in #4.