Modeling of Submillimeter Passive Remote Sensing of Cirrus Clouds

Abstract

The scattering properties of cirrus clouds at submillimeter-wave frequencies are analyzed and characterized in this paper. This study lays a theoretical foundation for using radiometric measurements to investigate and monitor cirrus properties from high-flying aircraft or satellite. The significance of this capability is that it would provide data on the global distribution of cloud ice mass that is currently required to validate climate models. At present, these needs remain unmet by existing and planned observational systems. In this study the brightness temperature depression (?Tb) of upwelling radiation due to cirrus clouds is simulated at 150, 220, 340, 500, 630, and 880 GHz. The effects of a range of size distributions, eight ice particle shapes, and different atmospheric profiles are modeled. The atmospheric transmission is high enough in the submillimeter windows to allow upper-tropospheric sensing from space, but absorption by water vapor reduces the sensitivity to lower cirrus clouds in a simply predictable manner. It is shown that frequencies above 500 GHz have adequate sensitivity to measure cirrus cloud properties. For these higher frequencies, the ?Tb is closely proportional to ice water path (IWP) for median mass equivalent sphere diameters (Dme) above 125 ?m. The differing sensitivity with frequency allows two channels to determine particle size. A two-channel Bayesian algorithm is developed to assess retrieval accuracy with a Monte Carlo error analysis procedure. Particle shape, size distribution width, and receiver noise are considered as error sources. The rms errors for a nadir view with 630/880 GHz are less than 40% for IWP > 5 g m?2 and Dme > 100 ?m, while using an oblique viewing angle of 73° results in the same accuracy down to an IWP of 1 g m?2 (visible optical depth less than 0.1). The two-channel algorithm and error analysis methods are used to show how submillimeter radiometer and millimeter radar measurements may be combined.