Satellite Remote Sensing of Multiple Cloud Layers

Abstract

The goals of the current study are threefold: 1) to present a multispectral, multiresolution (MSMR) methodology for analysis of scenes containing multiple cloud layers; 2) to apply the MSMR method to two multilevel cloud scenes recorded by the NOAA Advanced Very High Resolution Radiometer (AVHRR) and the High Resolution Infrared Radiometer Sounder (HIRS/2) instruments during the First International Satellite Cloud Climatology Program (ISCCP) Regional Experiment (FIRE) on 28 November 1991; and 3) to validate the cloud-top height results from the case study analyses through comparison with lidar, radar, aircraft and rawin-sonde data. The measurements available from FIRE Cirrus II enable detailed examination of two complex cloud scenes in which cirrus and stratus appear simultaneously. A ?fuzzy logic? classification system is developed to determine whether a 32?32 array of AVHRR data contains clear sky, low-level cloud, midlevel cloud, high-level cloud, or multiple cloud layers. With the addition of the fray logic cloud classification system, it is possible for the first time to find evidence of more than one cloud layer within each HMS field of view. Low cloud heights are determined through application of the spatial coherence method to the AVHRR data, while mid- to high-level cloud heights are calculated from the HIRS/2 15-µm CO2 band radiometric data that are collocated with the AVHRR data. Cirrus cloud heights retrieved from HIRS 15-µm CO2 band data are improved for optically thin cirrus through the use of the upper-tropospheric humidity profile. The MSMR-derived cloud heights are consistent with coincident lidar, radar, and aircraft data. Cirrus and stratus cloud-top heights and cirrus effective emittances are retrieved for data within an ISCCP 2.5° grid cell that encompasses the FIRE experimental region.