Grids¶

AerEO uses the Major TOM grid (paper) to turn any AOI into a set of equal-area, globally consistent cells. Outputs are indexed by cell ID, so scenes from different sensors line up without manual reprojection.

The Major TOM equal-area grid. Image from the ESA Phi Lab repository.

Major TOM grid basics¶

The grid divides the world into fixed cells. When you set grid_dist, AerEO selects the cells that intersect your AOI and uses them as the extraction framework.

from aereo.pipeline import ExtractionJob

job = ExtractionJob(
    name="demo",
    grid_dist=10_000,  # 10 km cells
    output_uri="/tmp/demo",
    target_aoi=aoi,
    read=read_odc_stac,
    write=write_geotiff,
)

Grid cell size¶

Smaller cells mean more tasks and smaller files; larger cells mean fewer tasks and larger files.

How cells become tasks¶

flowchart LR
    AOI["AOI polygon"] --> Grid["Major TOM grid"]
    Grid --> Cells["Intersecting cells"]
    Cells --> Group["Group by time + native CRS"]
    Group --> Tasks["ExtractionTask list"]

build_grouped_tasks intersects the AOI with the grid, then groups the resulting cells by acquisition time and native CRS. Each group becomes one or more ExtractionTask objects, depending on cells_per_task.

Reprojecting to a cell's local UTM geobox¶

When reproject_mode is "grid", AerEO reprojects each cell to its local UTM zone before writing. This keeps pixels square and avoids warping a whole scene to a single CRS when it spans multiple UTM zones.

See Reprojection for details.

Grid helpers¶

from aereo.grid import build_grid_cells, intersect_cells

cells = build_grid_cells(aoi, grid_dist=10_000)
selected = intersect_cells(cells, aoi)

These helpers are useful when you want to inspect the grid before running an extraction. The Configuration section explains how grid_dist and target_aoi are declared in YAML.