Add new 'cloud_mask' param to load_ard

[robbibt]

Proposed changes

A recent client enquiry revealed a limitation of the current load_ard implementation of multiple cloud masks: at the moment the choice of whether to use Fmask or s2cloudless is tied to the same param (mask_pixel_quality) that decides whether to actually mask out cloudy pixels as NaN​. Users may instead just want to use s2cloudless for filtering to non-cloudy scenes using min_gooddata, without actually applying the mask.

This PR separates the choice of cloud mask from the choice of whether to actually mask out clouds as NaN​ pixels by removing the choice of cloud mask from the mask_pixel_quality param, and introducing a new cloud_mask param that controls the choice of cloud mask. This allows users to e.g. filter to non-cloudy cloudy scenes using min_gooddata + s2cloudless, but not actually have to apply the cloud mask to their data.

Have updated the Using_load_ard.ipynb and DEA_Sentinel2_Surface_Reflectance.ipynb notebooks to update them for this change.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB