ERBE Scene Identification as seen by ISCCP: Results from ScaRaB Data

Broadband-based ERBE scene identification algorithm and narrowband-based ISCCP cloud detection and classification algorithm are applied to ScaRaB (Scanner for Radiation Budget) simultaneous measured broadband (SW and LW) and narrowband (VIS and IR) radiances to study the relation between ERBE scene ID (four cloud cover intervals) and ISCCP scene type (clear or 15 cloud types). On average, ISCCP method detects more clear pixels than ERBE method. For the ISCCP clear pixels, ERBE says clear only on about half of them. For the pixels claimed by ISCCP to be cumulus and altocumulus, ERBE most likely classify them to be partly cloudy. For the pixels claimed by ISCCP to be stratocumulus, stratus and altostratus, ERBE most likely classify them to be mostly cloudy. For the pixels claimed by ISCCP to be cirrus, ERBE most likely classify them to be either mostly cloudy or overcast. For the pixels claimed by ISCCP to be altostratus, nimbostratus, cirrostratus, and deep convective cloud, ERBE most likely classify them to be overcast. The study directly demonstrates the cloud optical properties (i.e., cloud height and optical thickness) are implicitly included into the four ERBE cloud cover intervals.