STAC for Oceanographic Data Discovery

Finding ocean datasets without drowning in metadata.

The Problem

Oceanographic data is scattered: NOAA, Copernicus, NASA, institutional repositories. Each has different APIs, metadata formats, and access patterns. Finding the right dataset takes longer than analyzing it.

STAC: SpatioTemporal Asset Catalog

STAC is a specification for describing geospatial data. Think of it as a standardized card catalog for Earth observation data.

Core concepts:

• Item — A single spatiotemporal asset (one satellite image, one day of buoy data)
• Collection — A group of related items (all Sentinel-2 images, all Argo floats)
• Catalog — Collections organized hierarchically
• API — Search across catalogs with standard queries

Searching Ocean Data

from pystac_client import Client

# Connect to a STAC API
client = Client.open("https://api.example.com/stac")

# Search by bounding box and time
results = client.search(
    collections=["sea-surface-temperature"],
    bbox=[-180, -60, 180, 60],
    datetime="2024-01-01/2024-12-31"
)

for item in results.items():
    print(item.id, item.properties["sst_mean"])

Why STAC for Ocean Science

1. Federated search — Query multiple repositories with one API
2. Cloud-native — Assets can be COGs, Zarr, or cloud buckets
3. Extensions — Scientific metadata (units, quality flags, processing level)
4. Reproducibility — Exact asset versions, not just “SST data from NOAA”

Getting Started

Major ocean data providers with STAC:

• Microsoft Planetary Computer
• Element 84 Earth Search
• C4IROcean Ocean Data Platform

---

One API to find them all.

Source: C4IROcean-OceanDataPlatform