About the Atlas

The single navigable destination for "I want to understand Earth observation data." Structured data on every active and recent satellite, every key dataset across NASA's 12 DAACs — plus free layers from NOAA, ESA and others — a library of grounded science questions with working code, and cross-dataset fusion recipes.

This is an AI-curated guide, built solo with Claude Code: the catalog's plain-English names and much of the curation are AI-generated, and the flagship questions are verified end-to-end against live data before publishing. It is a personal project — not affiliated with, endorsed by, or representing NASA or UAH (the way Hannah Ritchie's Our World in Data is independent of any one institution). Every dataset belongs to its provider; sources are cited per page. Content is CC-BY-4.0; code is MIT.

The thinking behind it is in the essay From Photons to Proof ↗ — why an Earth-data answer should have to prove itself, from raw light to a verified result. (More essays like it will live here on the site over time.)

What this is not

• Not a chat app. Chat apps go stale; literacy stays useful.
• Not an MCP server. MCP servers are commodity by mid-2026.
• Not a science paper. Papers are gated; atlases are open.

The content model

Satellite	instruments → measurements → datasets; launch, status, orbit
Dataset	short name, theme, resolution, cadence, swath, formats, DAAC, snippet
Question	plain-language goal → datasets + code template + expected output + caveats
Recipe	cross-dataset fusion pattern with auth / format / temporal-alignment notes
Glossary	visual concept explainer
DAAC	one of NASA's 12 archive centers — what they curate, format quirks

How it's built

• Astro — static-first generator with content collections + MDX.
• MDX content in src/content/; each page cites primary sources with a last_verified date.
• Live islands — the home globe, Explore map, Live dashboard, Connections graph, and Tracks all run client-side against NASA GIBS / CMR / Celestrak.