How to deal with cloudy points?


#1

Dear Sir,

I am working on Sentinel-3A data to visualise the variable LST. I found in some days when there is clouds covered, the LST values are negative upto negative (-) 20 deg C (Normally it is never below 15 deg C). I hope this is the temperature of the clouds. How to remove such cloudy points and assign right values before reprojection?

Regards,
Jitendra


#2

Hi Jitendra,

beside the LST you also have the uncertainty band. I had a quick look and high uncertainty value (>0.18) matched the low LST values pretty well. You can use it to mask out the values. In addition, you can evaluate the flag information which is given in the cloud_in band. Check the flags _cloud_in_gross_cloud, cloud_in_thin_cirrus, cloud_in_medium_high. With those you can mask your cloudy pixels too.