Multiangle remote sensing data can discriminate among aerosol air mass type
s, as represented by climatologically probable, external mixtures of compon
ent particles. Retrievals are performed over a comparison space of four-com
ponent mixtures, selected from six commonly observed components having assu
med, fixed microphysical properties but with mixing ratios free to vary fro
m 0% to 100%. We refer to this approach, which assumes climatologically pro
bable component particles and derives aerosol mixtures from the observation
s, as a "climatological retrieval." On the basis of simulated Multiangle Im
aging Spectroradiometer (MISR) observations over dark water, the retrieval
can distinguish mixtures containing large, spherical particles (sea salt),
nonspherical particles (accumulation and coarse mode dust), and small, dark
particles (black carbon) to within 20% or better of each component's true
mixing ratio. This is sufficient to distinguish maritime from continental a
erosol air masses. The retrievals, which use all nine MISR angles and the t
wo wavelengths least affected by ocean surface reflectance (672 and 867 nm)
, are not good at distinguishing medium, spherical, nonabsorbing (sulfate)
from medium, spherical, absorbing (carbonaceous') particles. However, the s
um is retrieved to within 20% of the true mixing ratio or better. This is s
ignificantly more detail about the properties of particle mixtures than has
previously been retrieved from satellite data, and in all cases, the deriv
ed total column aerosol optical depth remains well constrained, to at least
0.05 or 20%, whichever is larger. We expect the MISR data, with its freque
nt global coverage, to complement in situ and field data, which can provide
greater detail about aerosol size and composition locally. This combined e
ffort should advance our knowledge of aerosol behavior globally and our abi
lity to model the impact of aerosols on the climatically important solar ra
diation budget.