The abstract is read at the publisher; the summary is JClub's.
An upwind-transition-receptor framework for 1–4-day PM10 forecasts at three northern Taiwan receptors found the upwind-inclusive CatBoost model lowest aggregate RMSE at 11.04 μg m−3.
- Why it matters: Transboundary PM10 forecasting needs to separate what upwind measurements provide from what air-mass trajectory timing adds, to improve short-range prediction.
- What they did: The study built nested models that added local conditions, regional background, nine continental particulate records, and a transport-timing composite from historical trajectories, then evaluated 1–4-day forecasts at three receptors.
- The result: Nested comparisons attributed most added skill to distributed upwind observations, while trajectory timing gave receptor-dependent refinement; lower-tropospheric trajectories acted as timing evidence with decreasing vertical representativeness at longer lookbacks.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.