NetCDF test for quantization and compression

[harry-shepherd]

NetCDF quantization, along with gzip compression provides an interesting approach to reducing the data output from models. This is not currently implemented in XIOS. This issue aims to construct a small standalone test to demonstrate the functionality in NetCDF. The aim is to:

• Write out the data without any processing
• Write out the quantized data
• Write out the quantized and compressed data

As there are three different quantization methods; bit grooming, granular bit round, and bit round, and I will explore all three.

https://docs.unidata.ucar.edu/netcdf-c/current/md__media_psf_Home_Desktop_netcdf_releases_v4_9_2_release_netcdf_c_docs_quantize.html#:~:text=The%20quantize%20algorithms%20assist%20with%20lossy%20compression%20by,built%20into%20netCDF-4%2C%20to%20better%20compress%20the%20data.

[harry-shepherd]

The example will be written in C, as this is more directly relevant to the C++ implementation in XIOS.