As more information about global aerosol properties has become available from remotely sensed retrievals and in situ measurements, it is prudent to evaluate this new information, both on its own and in the context of satellite retrieval algorithms. Using the climatology of almucantur retrievals from global Aerosol Robotic Network (AERONET) Sun photometer sites, we perform cluster analysis to determine aerosol type as a function of location and season. We find that three spherical-derived types (describing fine-sized dominated aerosol) and one spheroid-derived types (describing coarse-sized dominated aerosol, presumably dust) generally describe the range of AERONET observed global aerosol properties. The fine-dominated types are separated mainly by their single scattering albedo (ω ₀), ranging from nonabsorbing aerosol (ω ₀ ∼ 0.95) in developed urban/industrial regions, to moderately absorbing aerosol (ω ₀ ∼ 0.90) in forest fire burning and developing industrial regions, to absorbing aerosol (ω ₀ ∼ 0.85) in regions of savanna/grassland burning. We identify the dominant aerosol type at each site, and extrapolate to create seasonal 1° × 1° maps of expected aerosol types. Each aerosol type is bilognormal, with dynamic (function of optical depth) size parameters (radius, standard deviation, volume distribution) and complex refractive index. Not only are these parameters interesting in their own right, they can also be applied to aerosol retrieval algorithms, such as to aerosol retrieval over land from Moderate Resolution Imaging Spectroradiometer. Independent direct-Sun AERONET observations of spectral aerosol optical depth (τ) are consistent the spectral dependence of the models, indicating that our derived aerosol models are relevant.