You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Current ARD products include a top-level property called tile_id, this maps to SceneId in case of Landsat imagery and to GranuleIdentifier (also referred as TILE_ID in some xml metadata files).
I think tile_id is a confusing name, my first guess was that it's a unique id for a spatial region that this dataset covers, like path,row in case of Landsat or MGRS code for Sentinel. It is not that however.
What this field contains is data provider supplied string describing a specific observation on a given day, it includes sensor information, location and time information, level of processing and any other information provider chose to include. The format is provider specific. For example Landsat Scene Id looks like this LT50970681987310ASA00, example for sentinel data: S2A_MSIL1C_20170105T013442_N0204_R031_T53NMJ_20170105T013443
This data is lineage related, essentially it's a unique identifier for a source dataset from which this ARD dataset was constructed. So I propose we put it somewhere under lineage subtree.
Two questions:
Do we need a unified name, or can we live with provider specific name
lineage.scene_id for Landasat
lineage.granule_id for Sentinel
If we need unified name, what should it be, some suggestions
lineage.provider_reference_code
lineage.provider_dataset_id
Names are really hard to change once they are backed in, tile_id has lot's of room for confusion and I would very much prefer we don't use that name.
Here is a definition from wikipedia:
A tessellation of a flat surface is the tiling of a plane using one or more geometric shapes, called tiles, with no overlaps and no gaps.
Current ARD products include a top-level property called
tile_id
, this maps toSceneId
in case of Landsat imagery and toGranuleIdentifier
(also referred asTILE_ID
in some xml metadata files).I think
tile_id
is a confusing name, my first guess was that it's a unique id for a spatial region that this dataset covers, likepath,row
in case of Landsat or MGRS code for Sentinel. It is not that however.What this field contains is data provider supplied string describing a specific observation on a given day, it includes sensor information, location and time information, level of processing and any other information provider chose to include. The format is provider specific. For example Landsat Scene Id looks like this
LT50970681987310ASA00
, example for sentinel data:S2A_MSIL1C_20170105T013442_N0204_R031_T53NMJ_20170105T013443
This data is
lineage
related, essentially it's a unique identifier for a source dataset from which this ARD dataset was constructed. So I propose we put it somewhere underlineage
subtree.Two questions:
Do we need a unified name, or can we live with provider specific name
lineage.scene_id
for Landasatlineage.granule_id
for SentinelIf we need unified name, what should it be, some suggestions
lineage.provider_reference_code
lineage.provider_dataset_id
Names are really hard to change once they are backed in,
tile_id
has lot's of room for confusion and I would very much prefer we don't use that name.Here is a definition from wikipedia:
Is there notation from STAC that we can borrow?
@simonaoliver @jeremyh @omad @ASVincent @GypsyBojangles
The text was updated successfully, but these errors were encountered: