FAQ#
Is TideTracker accurate enough to actually plan dives around?#
For most Salish Sea sites, yes — especially sites at or near reference stations, or sites with a high-confidence applied delta from diver reports. For sites with low-confidence corrections or distant subordinate stations, treat the prediction as a strong starting point that benefits from your own site knowledge. The app always shows you the confidence tier and how recently the correction was updated, so you can weight predictions accordingly. More detail in How the Data Works.
Where does the tide and current data come from?#
Directly from NOAA's CO-OPS prediction APIs. No scraping, no third-party aggregators. Corrections are layered on top of that raw data, and every metric in the briefing is tagged with the NOAA station it came from.
What's the difference between a reference station and a subordinate station?#
A reference station has its own long-term tide or current gauge with decades of data. Subordinate stations don't — their predictions are derived by applying a published offset to a reference station. Most dive sites sit on subordinate stations, which is why corrections matter.
What's an "applied delta"?#
It's the correction layered on top of the NOAA offset at a subordinate station, derived from diver reports and filtered through an IQR outlier model. It's the engine's best estimate of the additional adjustment a site actually needs beyond what NOAA publishes.
What do the confidence tiers mean?#
A site needs at least 10 logged observations before any correction is applied at all. Below that, the briefing shows "No correction data yet" (zero observations) or "Correction pending: n/10 required" (one to nine), and no confidence badge appears.
Once a correction is active, the badge tracks how much evidence sits behind it: Low confidence covers 10 to 19 observations — the applied delta exists but is still settling. Moderate covers 20 to 49 — the correction is meaningful but still tightening. High is 50 or more — the applied delta is well-established and stable. The tier is visible on each site so you can weight the prediction accordingly.
What does it mean when a site's correction is "stale"?#
Sites whose corrections haven't received new observations in over 12 months get a stale flag. It's a soft warning, not a verdict — a stale but well-established correction will still usually beat raw NOAA predictions alone. The flag is there so you know to weight the prediction with appropriate skepticism, and so that if you do dive the site, your report can include that context.
Can I trust a new site with only a few reports?#
The prediction will lean more heavily on NOAA's raw data until more reports come in. The site's confidence tier (Low) reflects that. Treat low-confidence sites the way you'd treat any partial information: as useful, but not as complete as a well-established site. If you do dive a low-confidence site, your report counts for more there than it would at a High-confidence site.
Does TideTracker account for weather?#
Not yet. Weather integration is on the roadmap and will surface as a surface-conditions layer — wind, barometric pressure, river outflow — distinct from the underlying NOAA depth-current predictions. Until then, check your preferred weather source separately.
Is my data private?#
Account data (email, password) is used only for authentication and never shared. Dive reports are aggregated into site-level corrections and are not publicly attributed to individual users. Full privacy details will live in a dedicated policy page before the beta ends.
Is it free?#
Yes, and that's the long-term plan, not a beta promotion. TideTracker isn't profit-driven and isn't going to be — there are no plans to charge users for access, ever. The app exists to serve the local diving community, and putting it behind a paywall would defeat the point.
That said, real infrastructure costs real money. Hosting, NOAA data refreshes, and especially the remote sensor network on the long-term roadmap all add up. If engagement grows to the point where those costs outpace what I can comfortably absorb personally, I'll set up a community donation system to cover operations and development — voluntary, transparent, and tied to actual costs rather than profit. Beta users will hear about it directly when and if that happens.
If TideTracker ever changes hands, the community will hear about that directly too. There's no current plan to sell, but I won't pretend it's impossible — what I will commit to is that any change of ownership has to preserve the free, community-first nature of the app, or it doesn't happen.
Can I help build TideTracker?#
Yes — and I'd welcome it. Beyond reporting dives, there's room for help with code, maintenance, documentation, and (down the line) sensor deployment and recovery work. If you're interested, the Community page has more detail on what kinds of contributions would be useful. The short version: get in touch through the feedback form and tell me what you'd like to work on.
Who's building this?#
One developer, me, in between dives. There's no team, no investors, no product roadmap committee. Feedback reaches me directly through the in-app form.
Where do I log a dive?#
The "Log a dive" button in the sidebar (route: /dashboard/log). The slack delta you log there feeds site-specific corrections if you opt in to share. The "Send feedback" button at the bottom of the sidebar is a different surface — that one's for bug reports and general notes, not per-dive observations.
How do I report a bug, a missing site, or a bad prediction?#
The "Send feedback" button at the bottom of the sidebar opens a small form for bug reports, station-pairing concerns, and general notes. It goes straight to my inbox. For per-dive observations, use "Log a dive" instead — that's the surface that feeds the correction engine.
Can I use TideTracker for diving outside Puget Sound and the Strait?#
Not currently. The correction engine is tuned to the Salish Sea and the sites loaded in the database reflect that. Expansion to other regions is possible long-term but not planned for the near term.