Integrated assessment of aerosol optical properties over West Africa: multivariate classification, seasonality, and long-term variability

Despite substantial advances in aerosol remote sensing, objective characterization of mixed aerosol regimes and their seasonal evolution over West Africa remains poorly constrained, limiting accurate assessments of aerosol–climate interactions and regional air quality. This study addresses this gap by developing a robust, data-driven framework that integrates Principal Component Analysis (PCA) and Gaussian Mixture Models (GMM) to objectively classify aerosol regimes using four complementary aerosol optical properties (AOD, SSA, AAOD, and AE) derived from integrated AERONET, MODIS, MISR and MERRA-2 observations during 2005–2022. Unlike conventional threshold-based approaches, the proposed framework simultaneously incorporates aerosol loading, scattering, absorption, and particle-size characteristics within a probabilistic classification scheme, enabling improved discrimination of mixed and transitional aerosol populations. Distinct aerosol regimes were identified across the six AERONET stations, revealing pronounced spatial heterogeneity in aerosol optical characteristics. The six-component GMM solution achieved the lowest Bayesian Information Criterion (BIC) among the models evaluated in PC1–PC2 space. Long-term trend analyses showed that statistically significant changes occurred primarily in aerosol composition (SSA: τ = −0.52 to 0.31, p < 0.05; AAOD, p < 0.01), whereas AOD remained largely stable, indicating that changes in aerosol radiative properties outpaced changes in atmospheric aerosol loading. Seasonal variability ( p < 0.001) emerged as the dominant control on aerosol behaviour, reflecting the coupled influences of Saharan dust transport, biomass-burning emissions, West African monsoon dynamics, and wet deposition. These findings provide new insights into the processes governing aerosol optical variability over West Africa and demonstrate that the integrated PCA–GMM framework offers a robust, reproducible, and transferable approach for aerosol climatology, satellite comparison, air-quality assessment, and regional climate-impact studies.