Apr 6, 2021 Junchang Ju & Brian Freitag
How was the list of HLS tiles over land excluding Antarctica derived.
-
The land mask was derived using the shorelines dataset from NOAA https://www.ngdc.noaa.gov/mgg/shorelines/data/gshhg/latest/
- Shapefiles are not included in the github repo and should be downloaded from the link above.
- The full resolution GSHHG level 1 shapefile is used (GSHHS_shp/f/GSHHS_f_L1.shp) a) Level 1 shapefile includes the land/ocean boundary - inland water bodies are not masked. b) Lower resolution shapefiles are available. If users want to reduce processing time with a lower resolution shapefile, update the file path in the params.json file c) A 0.01 degree buffer is added to the land boundary prior to finding the S2 grid intersection.
-
The KML of the MGRS grid is provided via Copernicus and can be converted as stored as a geojson using create_S2_geojson.py:
Input KML: https://sentinel.esa.int/documents/247904/1955685/S2A_OPER_GIP_TILPAR_MPC__20151209T095117_V20150622T000000_21000101T000000_B00.kml/ec05e22c-a2bc-4a13-9e84-02d5257b09a8 Output: s2_grid.json
-
The file list is generated by executing HLS_land_tiles.py a) params.json is required for execution. This file requires the following inputs i) "path_to_gshhs_sh" ii) "S2_kml_url" b) Output: HLS.land.tiles.txt c) Runtime: ~5 hours to complete the HLS land tile grid
There are 18952 tiles in HLS.land.tiles.txt. The coverage of the tiles is given in HLS_global_coverage.jpg.