**Note** from Bead: Alertwest Record · [canonical source](https://redfish.acequia.io/guerin/.agents/24882a83-509b-4a9b-8608-303726c0c624/2026-06-15/notes/00-the-test.md) · session 2026-06-15 · discussion: Talk: Alertwest Record
This bead is a **realtime.earth end-to-end test**, framed by Stephen: > archive the last 24 hours of the alertwest.live cameras that are within a 100 mile > radius of Los Alamos, New Mexico. This is a test of our realtime.earth system. > ... Store the imagery in directory in your bead as artifacts. and then when done, > make a STAC.
## Why this is the right test The team already built the AlertWest backup pipeline and a STAC over the 2025 Palisades / 2026 Sandy fires (bead `cbf8b027`, the `simtable.acequia.io/docs/alertcalifornia/` docs, and the `derive-alertwildfire-sidecars` skill in bead `65783732`) — but always pointed at a **past California incident** with a pre-curated camera list. This test inverts that: a **live** pull over **our home region** (northern New Mexico), letting the geometry pick the cameras (100 mi of Los Alamos), with no incident to anchor on. It exercises the full chain — roster → proximity filter → 24 h timelapse → image archive → STAC — exactly as realtime.earth would for a fresh event, and proves the pipeline works off a cold region rather than a warmed-up incident folder.
## What "realtime.earth" gets out of it - A **time-queryable spatiotemporal catalog** (STAC) of fixed-camera imagery, the same shape the incident-viewer / viewer-3d consume — so the archive can drop straight into those viewers. - A reusable, parameterized **`fetch-alertwest-archive.mjs`** (center + radius → 24 h archive) that is no longer incident-coupled. - A baseline of what northern-NM AlertWest coverage looks like on an ordinary day (11 cameras, Santa Fe / Tesuque / Picacho / Glorieta / Albuquerque-Double-Eagle), useful the next time a fire actually starts here. See `01-method-and-results.md` for the run.