An analysis of global aerosol type as retrieved by MISR

• Published in: Journal of geophysical research. Atmospheres 2015-05, Vol.120 (9), p.4248-4281
• Main Authors: Kahn, Ralph A., Gaitley, Barbara J.
• Format: Article
• Language: English
• Subjects: aerosol remote sensing; aerosol type; Aerosols; Albedo; Algorithms; Argon oxygen decarburizing; Atmospheric aerosols; Classification; Climatology; desert dust; Discrimination; Geophysics; Ground truth; Meteorology; MISR; multiangle imaging; Optical analysis; particle pollution; Retrieval
• ISSN: 2169-897X; 2169-8996
• DOI: 10.1002/2015JD023322

In addition to aerosol optical depth (AOD), aerosol type is required globally for climate forcing calculations, constraining aerosol transport models and other applications. However, validating satellite aerosol‐type retrievals is more challenging than testing AOD results, because aerosol type is a more complex quantity, and ground truth data are far less numerous and generally not as robust. We evaluate the Multiangle Imaging Spectroradiometer (MISR) Version 22 aerosol‐type retrievals by assessing product self‐consistency on a regional basis and by making comparisons with general expectation and with the Aerosol Robotic Network aerosol‐type climatology, as available. The results confirm and add detail to the observation that aerosol‐type discrimination improves dramatically where midvisible AOD exceeds about 0.15 or 0.2. When the aerosol‐type information content of the observations is relatively low, increased scattering‐angle range improves particle‐type sensitivity. The MISR standard, operational product discriminates among small, medium, and large particles and exhibits qualitative sensitivity to single‐scattering albedo (SSA) under good aerosol‐type retrieval conditions, providing a categorical aerosol‐type classification. MISR Ångström exponent deviates systematically from ground truth where particle types missing from the algorithm climatology are present, or where cloud contamination is likely to occur, and SSA tends to be overestimated where absorbing particles are found. We determined that the number of mixtures passing the algorithm acceptance criteria (#SuccMix) represents aerosol‐type retrieval quality effectively, providing a useful aerosol‐type quality flag. Key Points MISR multiangle retrievals provide global aerosol‐type classification mapping Aerosol‐type discrimination improves dramatically for aerosol optical depth > 0.2 The number of successful mixtures represents aerosol‐type retrieval quality