Rainfall events

rainfall_thresholds.events

Rainfall events from an hourly series, after Melillo et al. (2018).

The series is split wherever no_rain_break_hours or more consecutive dry hours occur; an event whose cumulated rainfall is below min_event_mm is dropped. An event runs from its first to its last wet hour, both included, so a single wet hour is a one-hour event.

This is a simplified reconstruction: Melillo et al. also vary the dry gap with the season and separate sub-events within a rainfall period. Use their CTRL-T tool when the reconstruction itself is what you are studying.

RainfallEvent dataclass

One rainfall event: first and last wet hour, duration and total.

intensity_mm_h property

intensity_mm_h: float

Mean intensity over the event.

reconstruct_events

reconstruct_events(series: Iterable[tuple[datetime, float]], *, no_rain_break_hours: int, min_event_mm: float) -> list[RainfallEvent]

Split an hourly (timestamp, mm) series into rainfall events.

The series is sorted by timestamp first. Dry hours are counted as samples, so the series is expected to be hourly and without gaps.

latest_event

latest_event(events: Iterable[RainfallEvent]) -> RainfallEvent | None

The event that ended last, or None.