Progress Report for Week 6 (July 7 – July 13)
1. What did I get done this week?
- Explored and analyzed the NASA POWER Zarr timeseries datacube using AWS S3 and Xarray.
- Loaded and inspected the datacube structure, including dimensions, variables, and metadata.
- Developed and validated functions to convert the Zarr datacube into GeoCroissant JSON-LD, with type inference and metadata cleaning.
- Converted both the full datacube and a targeted 2020 T2M subset into GeoCroissant format, adding detailed monthly descriptions.
- Validated the GeoCroissant metadata using
mlcroissant
tools to ensure compliance. - Processed the 2020 T2M timeseries into monthly subsets and created visualizations of monthly and annual patterns.
2. Plan for Next Week ( July 14 - July 20):
- UMM to GeoCroissant conversion support.
3. Am I blocked on anything?
- No, I am not currently blocked on anything.
Links to Work Done:
- Public Repository: ZOO-AI-DATASET-MAAS/Datacube to GeoCroissant at main · HarshShinde0/ZOO-AI-DATASET-MAAS · GitHub
- Updated Project Wiki Page: GSoC 2025 AI‐ready Dataset Metadata as a Service using ZOO‐Project · HarshShinde0/ZOO-AI-DATASET-MAAS Wiki · GitHub
Best Regards,
Harsh Shinde