Full(-ish) MGRS 1km grid. Data is downloaded data from https://earth-info.nga.mil/index.php?dir=coordsys&action=utm-1km-polyline-dloads (but 5 files were missing from their list). Goal is to make it easier to access - details in the readme.
This project crawled the map download links at https://earth-info.nga.mil/index.php?dir=coordsys&action=utm-1km-polyline-dloads
and put them into a GeoParquet files that follow the Best Practices for Distributing GeoParquet. The full directory contains the complete dataset as a single GeoParquet file. For now it just contains mgrs_clean.parquet which is all the
download links that worked, but removing the ShapeLen, ShapeArea and area columns, as they weren't all that consistent.
Size: 5.87 GB
Compression: ZSTD
Rows: 305,799,745
Row Groups: 2,483
GeoParquet Version: 1.0.0
CRS: EPSG:4326
Bounding Box: [-180.000000, -80.000000, 180.000000, 84.000193]
The partitioned/ directory currently includes gzd_partition/, containing ~700 partition files (one per GZD).
These partitions:
mgrs_clean.parquetSize: 9.02 MB
Compression: ZSTD
Rows: 187,646
Row Groups: 1
GeoParquet Version: 1.1.0
CRS: EPSG:4326
Bounding Box: [-180.000000, -80.000000, 180.000000, 84.000193]
The best way to use these is to treat the partition as a full parquet dataset and query just what you need out of it. With DuckDB you can do:
And then you'll have the table locally. You can go straight to a parquet file with