Skip to main content

I am copying Sentinel-2 L1C tiles from the SentinelHub S3 bucket to my own and using the tile ID to rename the file. A snippet of my code is as follows:


    wfs_iterator = WebFeatureService(
bbox,
date_range,
data_collection=data_collection,
maxcc=cloud_max,
config=config
)
return [(tile['properties']['id'], tile['properties']['path']) for tile in list(wfs_iterator)]

I’ve noticed for a few tiles, the date as indicated in the AWS path differs from the date in the Tile ID. For example, the tile stored at s3://sentinel-s2-l1c/tiles/18/N/YK/2015/10/19/0 has the Tile ID S2A_OPER_MSI_L1C_TL_MTI__20160410T011941_A001695_T18NYK_N02.01. Is there a reason for this discrepancy? Does the date in the Tile ID not necessarily correspond to the sensing date of the tile? I can’t seem to find a breakdown of the naming conventions for the Tile ID, so I can’t confirm exactly what the date is supposed to be. However, for most other tiles, the date in the AWS path matches the date in the Tile ID.

If you check S2 user guide, “Old format Naming Convention”, you will see that for granules, this is a creation time, not sensing time:


This is exactly what I was looking for, thanks! Couldn’t figure out the search terms to find the naming convention.


try “sentinel-2 granule naming convention”


Reply