Using radiative properties to detect mineral dust aerosols over the ocean: A comparison of MODIS and CALIPSO techniques

Detection of mineral dust aerosols from satellites is challenging. Mineral dust aerosols are often spatially localized, but can be transported over large areas. They typically originate from known source regions and can have strong interactions with surface, cloud, and meteorological conditions. Mineral dust aerosol spectral optical properties differ from those of clouds and some other aerosols. Satellite dust detection algorithms try to capitalize on these differences. Detection methods have been developed with varying degrees of success. We will present an evaluation of several dust detection techniques using the Moderate Resolution Imaging Spectroradiometer (MODIS) level 1B data measured reflectances and brightness temperatures. Specifically, we will compare: (1) Ackerman et al.'s (1989) brightness temperature difference procedure, (2) Roskovensky and Liou's (2005) dust differentiation algorithm and (3) Zhao et al.'s (2009) multi-channel imager algorithm. These are all based on using visible and IR spectral regions to discriminate dust areas from cloud and clear-sky condition. The results of these MODIS algorithms will be compared with results from CALIPSO.