stilt.Trajectories.endpoints#

Trajectories.endpoints()[source]#

Per-particle trajectory endpoints (the far end of each particle’s path).

For each particle the row at the largest |time| from release is kept: where the air came from for a backward run (is_backward, the usual receptor case) or went for a forward run. This is the point to sample a boundary/background field at (e.g. lair.noaa.CarbonTracker.background) for a backward run. Direction is handled implicitly – max |time| is the most-negative offset for a backward run and the most-positive for a forward run.

Every particle contributes one endpoint, whether it ran the full duration or left the domain early. An early exit is a real endpoint: the point where that air entered (backward) or left (forward) the domain, which is exactly where the background should be sampled.

Returns:: One row per particle with columns indx, time (endpoint absolute time, UTC, ready for field sampling), lati, long, zagl, endpoint_age_min (signed minutes from release), and run_time (receptor release time). Ready to pass to CarbonTracker.sample()/.background().
Return type:: pandas.DataFrame