Retrieval of Cloud Microphysical Properties from Thermal Infrared Observations by a Fast Iterative Radiance Fitting Method

Abstract

An algorithm is presented for inferring the IR optical depth, effective radius, and liquid water path of clouds from multispectral observations of emitted thermal radiation, which takes advantage of the larger number of spectral channels of future satellite IR sounders. The proposed technique consists of fitting the measured radiances with theoretical results obtained from full radiative transfer simulations, to allow the correct description of vertically inhomogeneous cloud layers. The retrieval is made efficiently treatable by using a fast multiple scattering infrared radiative transfer model and a fast iterative fitting procedure. The theoretical basis of both parts of the algorithm is given in detail, together with the underlying equations. The sensitivity of the method to measurement and modeling errors is investigated systematically from synthetic spectra covering the range of variability of observations of moderately thick, low-water clouds. Instrument absolute calibration errors, as well as uncertainties in the cloud and surface temperatures and in the vertical humidity profile, are identified as major sources of retrieval errors. Taken together, they restrict the application of the algorithm to clouds with an IR optical depth (at 11 ?m) ranging between 1 and 4. While the algorithm has, with this, little applicability for water clouds, it should be useful for the retrieval of cirrus cloud and aerosol layer microphysical properties.