One of the primary goals of the Atmospheric Infrared Sounder (AIRS) is to retrieve atmospheric profiles. Beyond the clear sky sounding capability, the high spectral resolution infrared radiance contains useful information about clouds, particularly for high-level, optically thin cirrus. In the present research, we present methodology for the retrieval of cirrus optical thickness and effective particle size from AIRS high spectral resolution infrared radiances.

A fast cloudy radiative transfer model is developed to simulate AIRS radiances when clouds are present. Calculations show that the brightness temperature difference between this fast cloudy radiative transfer model developed specifically to simulate clouds and a discrete ordinates (DISORT) radiative transfer model is within 0.5K for most cloud cases. However, the computing speed of the fast model is 1000 times faster than DISORT. The fast model for cloudy radiance simulations was merged subsequently with the clear-sky fast model for AIRS sounding purpose developed at the University of Maryland-Baltimore County.

Various cirrus cloud studies have been conducted, which include: (1). the sensitivity of AIRS brightness temperatures to optical thickness and effective particle size, and (2) the sensitivity of simulated brightness temperatures to the microphysical properties of ice clouds for a range of wavenumbers in the IR region. Based on the sensitivity studies, methodology is suggested to retrieve the optical thickness and the effective particle size of cirrus clouds from AIRS measurements. For the implemention of the retrieval algorithm, ECMWF atmospheric profiles and collocated MODIS cloud altitude data are used for forward modeling calculations. The applicability of the retrieval algorithm is demonstrated by a case study.