The Atlas Series · Data-Access · v0.2 · 2026-07-10

Where astronomical data lives — and how to get it

A practical field-guide to the public archives. For every major survey and mission: what kind of data it holds, which data center it lives in, and exactly how to pull it — from a click-through portal to a copy-paste query. Built for undergraduates finding their first dataset, and for researchers who just need the endpoint. Includes the Chinese archives most Western tools leave out.

"The data is already public. The hard part is knowing which door to knock on. This is the map of the doors."

Looking for how much of each archive has actually been studied? → The Coverage Atlas
Region
Data center
Access method

By region

By data type

By wavelength

How data actually reaches you — a 60-second VO primer

Almost every archive here can be reached several ways. Undergraduates usually start at the portal; researchers usually end up at TAP or astroquery. They are doors into the same room.

Portal (web)

Point-and-click search in a browser. Best for exploring, previewing, and grabbing a handful of files. No code.

astroquery (Python)

An Astropy package with one submodule per archive (astroquery.mast, astroquery.ipac.irsa…). The friendliest programmatic door.

TAP + ADQL

The Virtual Observatory table protocol. Send SQL-like ADQL to a /TAP endpoint (via pyvo) and get a table back. Same language across every center — including China's tap.china-vo.org.

Bulk / cutouts / VO

Bulk file mirrors for whole surveys; image cutout services; and VO protocols (cone search, SIA for images, SSA for spectra) for standardized position queries.

How this atlas is organized

Every archive is filed under the data center that actually serves its bits — MAST, HEASARC, IRSA, the ESA archives, CDS/VizieR, NOIRLab's Astro Data Lab, NRAO, ESO/ALMA, JAXA/DARTS, and China's NADC/China-VO, IHEP, PMO and lunar/planetary system. That is the honest unit of "distribution": a mission's data doesn't float in the abstract, it lives in a specific center with a specific front door. Each card standardizes four things a data user needs before committing: what data type, what wavelength, what sky coverage, and which access methods work. Toggle Researcher to swap every plain-language blurb for a runnable access snippet, and use the Region and Data center chips to isolate, say, every Chinese facility at once.

Note. Access snippets are illustrative starting points — table and collection names evolve, so check each archive's current docs for exact schema. Some facilities (many TeV arrays, some Chinese missions) release data only to their collaborations or on request; those are marked with a portal link rather than a query.

References & standards

The doors are standardized. These are the protocol documents and archive papers behind the cards above.

← Showcase