Get scl_cloud_high_proba & scl_cloud_medium_proba with GPT

Optical Toolbox
#1

Hi everyone,
I have an application which extract the scl_cloud_high_proba and scl_cloud_medium_proba from the Sentinel 2 L2A images using Python, and stores them as GeoTIFF. However, now I am interested in getting these masks with GPT also in GeoTIFF
Does anybody know how the XML should be, because these masks cannot be obtained with the subset operator?
Thanks in advance

#2

Hi Borkadd
if you are looking for spatial-masks for the images: the cloud cover is encoded in the SCL(Scene classification map) image - as can be seen as white and grey areas in the EO Browser her

BTW the vegetation/non vegetation classification is fairly arbitrary while the cloud and cloud shadow are quite good.

#3

You can create a GPT graph with a ‘BandMaths’ operator that store the bands of your interest in a new band. The graph should contain the following operators:

  • Read
  • BandMaths
  • Write

I have attached a graph example that creates a raster mask in GeoTIFF format combining ‘scl_cloud_high_proba’ or ‘scl_cloud_medium_proba’ masks from Sentinel-2 L2A products.
Hope this fulfill your needs.S2_L2A_cloud_mask.xml (1.7 KB)

Federico