Maxar ARD sample data

[drwelby]

Hi from Maxar!

We have some COG gridded imagery in a public bucket that could be fun to play with:

s3://maxar-ard-samples/v3

Directory browser

In short, these are COGs plus STAC on a quadkey UTM grid, and are quadkey-addressable to blocks and even to pixels. For more info there's docs here.

Our team's previous project used Dask and we've thought about building Dask arrays with these images, so I'm interested to see how this project works with our STAC data.

We don't yet have Raster Extension metadata up to speed, so you'll need to set up the stack something like:

cat = Catalog.from_file('https://ard.maxar.com/samples/v3/rio_deforestation/order_collections/5763013822349948994_root_collection.json')
stack = stackstac.stack(cat, chunksize=512, epsg=32610, assets=['visual'], resolution=0.30517578125)

[drwelby]

With the above example we seem to be missing some bands, the "visual" imagery is RGB but loads to:

<xarray.DataArray 'stackstac-eaac4c5d9fdd6c3718ac94fee49663f7' (time: 19, band: 1, y: 17409, x: 17488)>

I haven't had time to investigate, this could just be due to our metadata structure, or I'm missing a keyword.

[kylebarron]

Seems like you're hitting this from the README:

Single-band raster data only! Each band has to be a separate STAC asset—a separate red, green, and blue asset on each Item is great, but a single RGB asset containing a 3-band GeoTIFF is not supported yet.

[drwelby]

ah there it is, thanks!