Closed danielm09 closed 3 months ago
Thanks, Daniel! There are similar issues showing up in other use cases/demos where display tiles are incorrectly masked. I think it is not with the dataset. We're looking into it.
The notebook/operation is working as expected again after a server-side roll back. New tests should help catch these issues.
Thanks, Justin.
I can confirm it is working fine here.
Hi.
There is some issue with the cloud masking process described in this tutorial. I still can't pin point what is the exact cause, but the effect is that the masking is not working properly - it seems that all pixels are being masked. When executing this tutorial on colab, we cannot see the cloud mask and the cloud free composite generated in the last cell displayed on the map.
Upon running
s2_sr_cld_col.map(add_cld_shdw_mask).map(apply_cld_shdw_mask).getRegion(geometry=AOI, scale=10).getInfo()
we get as a result:[['id', 'longitude', 'latitude', 'time', 'B1', 'B2', 'B3', 'B4', 'B5', 'B6', 'B7', 'B8', 'B8A', 'B9', 'B11', 'B12'], ['20200601T185919_20200601T190551_T10TER', -122.26901958504497, 45.701026511589156, 1591038717370, None, None, None, None, None, None, None, None, None, None, None, None], ['20200619T190921_20200619T191722_T10TER', -122.26901958504497, 45.701026511589156, 1592594517900, None, None, None, None, None, None, None, None, None, None, None, None], and so on...]
Indicating that maybe the function is masking all pixels somehow.
Everything was working fine a few days ago.
I wonder if there is something wrong with the
COPERNICUS/S2_CLOUD_PROBABILITY
collection.