IOTA publishes prediction refinements continuously, with major bundles each month and event-specific tightening in the days before a pass. A 6-hour cache during peak observing seasons keeps the pages within one refinement cycle of upstream; a 24-hour cache works well for dormant periods. The cache refresh is per-row, so an event that hasn't tightened costs nothing while the volatile ones update on demand.

Yes. Add a status column with values like upcoming, in-progress, or historical. A selector mapping renders a status badge on the page, but the URL stays live. Sitemap inclusion stays on so historical traffic continues to find the page. Search engines see a stable URL with relevant body content, just flagged as past.

Encode the event date and asteroid number into the slug, like 2024-03-15-432-pythia. Date plus asteroid is a unique pair within IOTA's catalog, and the format sorts naturally. Even if a future pass overlaps in name, the date keeps the URL space distinct, and the per-row mappings still target the correct base page.

Yes. Each row carries the asteroid number. Either run a parallel /asteroids/{slug}/ page group that lists every event for that body, or use a category column inside the occultation page group to render an inline link to the asteroid's encyclopedia URL on every event page.

IOTA provides 1-sigma uncertainty bands per prediction. Include them in the path coordinates payload and render two map overlays: the nominal centre line and the uncertainty band. A meta mapping points at the combined image stored by slug, refreshed when the export changes. Observers see the latest uncertainty without code changes.

Most occultation targets are referenced by catalog identifiers like TYC, HIP, or UCAC4. Use the canonical identifier in the target_star column, and consider a friendly_name column for the small set with proper names. Selector mappings handle both: the canonical identifier in the title, the friendly name as a parenthetical.

Yes. Add an editorial_notes column to the source. A selector mapping injects the cell into the page below the stat block, so the structured data stays canonical and the human commentary lives alongside it. The base page renders both halves cleanly without competing for space.

Yes. The same schema works for occultations of stars by TNOs, KBOs, or planetary satellites. Add a body_type column to differentiate, run a sub-group with /occultations/tno/{slug}/ as the URL pattern, and reuse the base page. The matrix grows naturally as new occultation campaigns publish predictions.