We try to answer one question fast and honestly: is this park open, partially closed, or closed, and why? Here's exactly how we get there — and where the method has limits.
Where the data comes from
Every status is built from the National Park Service's own Data API — the same source that powers the alerts on NPS.gov. We pull two things: each park's active alerts, and its published operating hours. We don't scrape park pages or use unofficial sources.
How often it updates
The NPS refreshes its alert data on roughly a two-hour cycle, so we refresh on the same cadence. That means our "last updated" time is the honest ceiling — we can't be more current than the source. For a fast-moving emergency, the official park page or a phone call is always more up-to-date than any status board, including ours.
How we decide open, partially closed, or closed
- We read the text of every alert, not just its type, because closures are frequently mis-filed under other categories.
- A whole-park closure, or a park outside its scheduled hours, reads closed.
- A closed trail, road, or area, or an active hazard, reads partially closed — the most common state by far.
- Everything else reads open. Cautions and general notices are surfaced on the park's detail panel but don't change the headline.
Where our reading can fall short
Our classifier is a set of rules applied to human-written alerts, so an unusual phrasing can occasionally be miscategorized. We tune it against real data regularly, but you should treat our status as a fast first read, not the final word. The official park page — which we link on every park — is always the authority.