Understand Landsat mask layer classification


I’m using Sentinel hub py in order to mask clouds from landsat 8 images.
I am using the cloud confidence layer for this image:

in the documentation , it says that vaues 1-2-3 indicate cloud confidence (for a certain level), and 0 means no cloud or not determined. However, when I print the cloud layer this is what I get:

as it can be seen from the colorbar, seems like the cloud was detected, however, it gets value 0 , meaning no confidence at all that this is cloud?
I’m a bit confused, why the cloud classified as 0? is this mistake of the classification or I underrstood that layer wrong?



Hi Reut,

Zero doesn’t necessarily mean “no confidence” but can also mean that the algorithm did not determine the status of this condition. See page 56 of the data users handbook.