Skip to main content

In this S3 bucket sentinel-s2-l2a/tiles/36/N/UH/2021/9/26 I see two directories (0/ and 1/) which according to the file naming documentation refer to two images for the same date.


How is this possible?

I’ve plotted the images, and they seem to contain the same spatial patterns, but with different value ranges. Were they processed differently?

There are many possible reasons (if you look in the rest of the metadata, you will find what it is):


  • two ground stations processing same product, with somehow different versions of the application

  • two overlapping orbits

  • one product was reprocessed with new processing software (new baseline)



Thanks for the prompt reply!


Is there a way to systematically select the “best” or “recommended” image from a set of multiple images from the same date?

In my case, I’m selecting all images for a given tile, so it would be a painful process to have to read each single image’s metadata to understand what I should use.


There is no simple way to abstract the complexity of 10PB+ of data…

Perhaps you should try Sentinel Hub services, which do exactly that:

b14942c50ca9dbe37634328b3d02223eec1f3ad7.pngsentinel-hub.com
57b10d77cdf19cda03dbe73c52e6d5854ef7620d.png


Sentinel Hub



Sentinel Hub - Cloud API for Satellite Imagery








Reply