Dates feel reassuring. They suggest precision and commitment. But for natural phenomena, a single "best date" is often misleading - it implies a level of predictability that nature rarely delivers, and it shifts attention from planning to expectation.

Instead, this platform uses timing windows and confidence levels. Together, they give you a more honest and ultimately more useful picture of when to go.

Timing windows

Every event on this platform has a broad seasonal window — the outer bounds within which the phenomenon is plausible — and a peak window, a plain-language description of when conditions are typically best. The peak window might be a month range ("July–September"), a condition trigger ("1–3 weeks after autumn rain"), or a recurring pattern ("around each new moon from May to August").

These windows are conservative by design. We would rather you arrive slightly early and have a longer opportunity than arrive at a single "peak date" and find conditions have shifted.

Confidence levels

Each event is assigned one of three confidence levels, reflecting how predictable the event is in a typical year.

High: The event occurs reliably each year within a well-established window. Timing is relatively consistent across years, and the main variables (weather, access) are manageable with preparation. Examples: southern right whale season, Milky Way core visibility.

Medium: The event occurs most years, but timing, intensity, or location can shift meaningfully depending on seasonal conditions. A good year requires a reasonable confluence of factors. Examples: wildflower blooms following winter rainfall.

Low / Variable: The event is real and recurring over longer cycles, but not reliably annual or predictable within a season. Worth knowing about; not worth planning a trip around without current local intelligence. Examples: pelican breeding following major inland flooding.

What drives variability

For most natural phenomena, the primary drivers of variability are temperature, rainfall, and timing relative to those conditions in the weeks or months prior. We include variability notes on each event page to make these drivers explicit rather than leaving them as background noise.

Confidence levels are reviewed periodically. If a pattern appears to be shifting — seasonally, climatically, or otherwise — confidence ratings may be revised, and those changes are logged.