When comparing the same Sentinel-2 image that was downloaded from AWS S2 OpenData bucket to the same image that was downloaded from Google Earth Engine I see great radiometric difference in the DN values on all bands B1-B12 (on the order of hundreds values).
Anyone know what is the source of this difference and is there any way to compensate for this difference.
Not sure which of the AWS S2 buckets you are looking at, but the source of the difference is almost certainly related to new baseline, which introduced an offset (description was available here until recently, but is no longer available).
The “main” AWS S2 bucket keeps data as they are produced by ESA
Thank you.
Yes, I was referring to the bucket you linked to in your response.
Could you please elaborate or give different source to the offset you refer to since the link is broken? or did you refer to the difference mentioned here?
Yes, this is the same thing - GEE does “shift range by 1000”, whereas on AWS the data are stored as produced by ESA.
But is the data stored on AWS is the L2A product?
This one indeed is:
arn:aws:s3:::sentinel-s2-l2a
Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.