Open pbranson opened 2 years ago
Hi @pbranson, glad you found this and very glad you're interested in using it!
This is a very early test, and is really just proof of concept work. What datasets are you working with, as we can prioritise those as the next test set of Collections and Items. I hope we'll have a development STAC API endpoint up in the next couple of weeks that you can point and and evaluate.
Cool!
Some of the more traditional remote sensing datasets we are using are referenced here: https://github.com/IOMRC/intake-aodn/blob/main/intake_aodn/catalogs/aodn.yaml These datasets have had their binary chunks mapped and aggregated using the kerchunk library (https://github.com/fsspec/kerchunk) It would be interesting to consider where/how these types of lower level indexes fit with STAC. Kerchunk essentially creates a virtual, zarr-like key,value store. virtual in the sense that that values refer to the url and byte-ranges of the contiguous binary chunks of the actual source data - so not a copy.
Other less traditional datasets are the various ocean wave datasets, which are cataloged here: https://github.com/rom-py/rompy/blob/main/rompy/catalogs/aodn.yaml (many of which seem to be currently broken due to an issue with the AODN WFS - see https://github.com/rom-py/rompy/issues/20)
Hey @pbranson our STAC Catalog got a lot bigger! It's pretty messy, but has lots of stuff in there. Let me know if you have any thoughts?
Catalog: https://stac.dev.aodn.org.au/
View of the catalog: https://radiantearth.github.io/stac-browser/#/external/stac.dev.aodn.org.au/
Hey @alexgleith,
This is cool. We have developed some intake drivers to wrap the older functionality provided by the WFS and nclist services, but a STAC catalog would mean much less boilerplate code.
Let us know when/if you would like anything tested out.
FYI @hot007, @blasee, @benjaminleighton