Skip to content

Choosing a Sensor

AerEO separates the core framework from search and I/O plugins. The table below shows a few common combinations, but it is not exhaustive: any constellation works as long as you pair a search function with a compatible reader and reprojector. The same search_earthaccess + read_satpy pair can read MODIS, VIIRS, Sentinel-3, and many other NASA datasets; the same search_stac can hit any STAC catalog. If your sensor is not listed, swap the plugin or write your own — see Build a Plugin.

flowchart LR
    Sensor["Your sensor / catalog"] --> Search["Search plugin"]
    Search --> Read["Read plugin"]
    Read --> Reproject["Reproject plugin"]
    Reproject --> Write["Write plugin"]

Sensor cheat sheet

Sensor Install Search Read Reproject Notes
Sentinel-2 MSI uv add aereo / pip install aereo search_stac read_odc_stac reproject_odc Easiest starting point; works with any STAC catalog.
MODIS, VIIRS, Sentinel-3, etc. uv add aereo aereo-read-satpy / pip install aereo aereo-read-satpy search_earthaccess read_satpy reproject_swath NASA Earthdata login required.
GOES ABI uv add aereo aereo-search-aws-goes aereo-read-satpy / pip install ... search_aws_goes read_satpy reproject_odc Public AWS S3, no auth.
GeoTessera uv add aereo aereo-search-tessera aereo-read-tessera / pip install ... search_tessera read_tessera — Public data, no auth required.

The install column is only a suggestion. If you already have a search function that returns AssetSchema, you can reuse the built-in reader/reprojector; if you have a custom reader, you can point any search function at it.

When to use each search provider

  • search_stac — any STAC catalog (Sentinel-2, Landsat, custom catalogs, etc.).
  • search_earthaccess — NASA Earthdata collections accessible through earthaccess (MODIS, VIIRS, Sentinel-3, etc.).
  • search_aws_goes — GOES ABI data on the public AWS S3 bucket.
  • search_tessera — GeoTessera tile catalogs.

When to use each reader

  • read_odc_stac — for STAC-backed, analysis-ready data cubes (Sentinel-2 L2A, Landsat, etc.). Uses odc-stac and returns an xr.Dataset aligned to the asset's native geobox.
  • read_satpy — for swath or Level-1b data that needs Satpy-style scene reading and resampling (MODIS, VIIRS, Sentinel-3, GOES, etc.).
  • read_tessera — for GeoTessera tile catalogs.

When to use each reprojector

  • reproject_odc — reproject a raster xr.Dataset to a target CRS and resolution using odc-geo.
  • reproject_swath — reproject 2-D lat/lon swaths to a UTM GeoBox using odc-geo and nearest-neighbor lookup. Use this for swath sensors such as VIIRS and Sentinel-3.

Not listed? Mix and match

AerEO plugins are plain functions. To add a new constellation you usually only need to answer two questions:

  1. How do I find the data? Use an existing search provider, or write one that returns a GeoDataFrame[AssetSchema].
  2. How do I open the data? Use an existing reader, or write one that turns an ExtractionTask into an xr.Dataset.

If a reader already understands the file format, you can point a different search provider at it. For example, read_satpy does not care whether the assets came from search_earthaccess, search_aws_goes, or your own custom search function — it only cares that the asset metadata describes a file Satpy can open.

Next steps

  • Follow the tutorial for your sensor in the Examples section.
  • Read the Search guide for collection and band naming tips.
  • Learn how to configure plugins in the Configuration section.