Strange line shift anomaly in Sentinel-2 level 1C data

Hi guys, I have recently been working with Sentinel-2 data and specifically looking at cloud cover. In one of the images I downloaded over the Karakorum on October 1st, I am came a across a strange line. Does anyone know the cause of this or have some documentation I can read related to this, and if applicable a solution to mitigate this? My first guess was that it was a calibration error or something. It runs along the whole image and is best visible in the clouds. My guess is that it runs along a granule. Cheers.


I believe what you are seeing is the boundary between Detectors in the Tile. This is not a data quality issue, but is an artefact due to the registration of the image being aligned on ground, and is a combination of the neighbouring detectors ‘looking’ in opposite directions - one up the Track, the next one down the Track - and the difference in altidude between the top of the Clouds and the Ground. This effect is more noticeable in the 60 Metre Bands.



Jan Jackson
S2MPC Operations Manager