Preprocessor for ESA Landcover data

[tommylees112]

Preprocessing the landcover data (there is an issue with the regridding on the ouce_server)

    Process:
    -------
    * chop region of interset (ROI)
    * regrid to same spatial grid as a reference dataset (`regrid`)
    * create new dataset with these dimensions
    * assign time stamp
    * Save the output file to new folder

[tommylees112]

TODO:

• Need to copy the array forwards JUST FOR THAT YEAR
• use the src.utils.create_shape_aligned_climatology or something similar to copy the array for the start of the year forwards in time
• default behaviour for upsampling is nearest which is fine but it means the years are split. Do we prefer that behaviour?

In [84]: d1 = d.sel(time = '1995-01-31')
In [82]: d2 = d.sel(time = '1995-12-31')

In [85]: (d1 == d2).mean()
Out[85]:
<xarray.Dataset>
Dimensions:   ()
Data variables:
    lc_class  float64 0.9986