Cloud Properties inferred from 812-µm Data

A trispectral combination of observations at 8-, 11-, and 12-µm bands is suggested for detecting cloud and cloud properties in the infrared. Atmospheric ice and water vapor absorption peak in opposite halves of the window region so that positive 8-minus-11-µm brightness temperature differences indicate cloud, while near-zero or negative differences indicate clear regions. The absorption coefficient for water increases more between 11 and 12 µm than between 8 and 11 µm, while for ice, the reverse is true. Cloud phase is determined by a scatter diagram of 8-minus-11-µm versus 11-minus-12-µm brightness temperature differences; ice cloud shows a slope greater than 1 and water cloud leer than 1. The trispectral brightness temperature method was tested upon high-resolution interferometer data resulting in clear-cloud and cloud-phase delineation. Simulations using differing 8-µm bandwidths revealed no significant degradation of cloud property detection. Thus, the 8-µm bandwidth for future satellites can be selected based on the requirements of other applications such as surface characterization studies. Application of the technique to current polar-orbiting High-Resolution Infrared Sounder (HIRS)-Advanced VM High Resolution Radiometer (AVHRR) datasets is constrained by the nonuniformity of the cloud scenes sensed within the large HIRS field of view. Analysis of MAS (MODIS Airborne Simulator) high-spatial resolution (500 m) data with all three 8-, 11-, and 12-µm bands revealed sharp delineation of differing cloud and background scenes from which a simple automated threshold technique was developed. Cloud phase, clear-sky, and qualitative differences in cloud emissivity and cloud height were identified on a case study segment from 24 November 1991, consistent with the scene. More rigorous techniques would allow further cloud parameter clarification. The opportunities for global cloud delineation with the Moderate-Resolution Imaging Spectrometer (MODIS) appear excellent. The spectral selection, the spatial selection, and the global coverage are all well suited for significant advances.