Aerosol absorption coefficients across global sites: Impacts of chemical composition, emissions, and meteorology

Light-absorbing aerosols warm the atmosphere by absorbing solar radiation, and this warming is an important and poorly constrained uncertainty in anthropogenic radiative forcing. Absorption is governed by aerosol emission, chemical composition, and meteorology, yet these competing associations have rarely been examined together across diverse environments. This study characterizes aerosol light absorption across 25 SPARTAN sites spanning six continents, using filter-based ultraviolet–visible spectrophotometry on PM 2.5 samples collected during 2020–2024, (i) deriving the absorption coefficient (b abs,550 ), (ii) mass absorption coefficient (MAC 550 ), and (iii) Ångström absorption exponents (AAE 400-650 , AAE 650-900 ) over 400–900 nm wavelength range. Site-mean mid-visible absorption coefficients (b abs,550 ) range from 4.9 to 87.3 Mm -1 . Partial Spearman rank correlations across 20 predictors identify ambient black carbon concentration (ρ = +0.64) and minimum temperature (ρ = +0.41) as the dominant positive correlates of absorption, while secondary inorganic aerosols most strongly suppress mass absorption coefficient (MAC 550 ρ = −0.44; all p < 0.01). K-means clustering sorts individual filter samples into five optical regimes. Three regimes emerge as dominant site types — high b abs,550 , moderate b abs,550 , low b abs,550 — while the high-AAE and high-MAC regimes appear as within-site minority fractions. Absolute Principal Component Analysis with a two-point inverse-distance-weighting estimator reproduces withheld-site b abs,550 with reasonable accuracy where training analogues exist. These findings indicate that aerosol absorption is associated with a tractable set of environmental variables and hints towards approximation from routinely available data in the future through larger datasets.