Quantifying Spatio-Temporal Dynamics of African Dust Detection Threshold for PM10 Concentrations in the Caribbean Area Using Multiscale Decomposition

Abstract

Due to African dust, the Caribbean area is known to have one of the highest incidences of asthma on the planet. Consequently, it is crucial to dissociate the impact of local sources from large scale sources in this region. The aim of this study was to estimate the PM10 detection threshold for dusty events using a statistical approach and a dynamic approach. To carry out this analysis, PM10 time series from Martinique (MAR), Guadeloupe (GPE) and Puerto-Rico (PR) were used between 2006 and 2016. The statistical analysis highlighted that the distance from the African coast is a key feature for PM10 concentrations distribution with the highest at MAR (26.52 μg/m3) and the lowest at PR (24.42 μg/m3). The probability density function analysis showed that MAR-GPE-PR distributions converge towards a same point between the first and the second maximum probability value at 28 μg/m3. The dynamical analysis with the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and the Improved CEEMDAN (ICEEMDAN) validated the 28 μg/m3 found with the statistical analysis. The analysis of HYSPLIT back trajectories confirmed this threshold. Thus, our results indicated that 28 μg/m3 is the PM10 detection threshold for African dust in the Caribbean basin. It will therefore be a good indicator allowing the competent authorities to take the appropriate decisions to protect vulnerable populations during dusty events.