Color Maps & Plots | Cloud Climatology On-Line Downloadable Datasets
Mesoscale Climatology On-Line Downloadable Datasets
Surface Climatology On-Line Downloadable Datasets

ISCCP DEFINITION OF CLOUD TYPES

Cloud Parameter Definitions And Measurement Methods

This parameter represents the fractional area covered by clouds as observed from above by satellites. It is estimated by counting the number of satellite fields-of-view (called pixels, about 5 km across for ISCCP) that are determined to be cloudy and dividing by the total number of pixels in a region about 280 km across. Cloud amount for lower-level clouds is only that fraction of the area actually observed to be covered by clouds at that level. This way of determining cloud amount assumes that each cloudy pixel is covered completely by clouds.

This parameter represents the location of the "radiating" top of the clouds; if there is a very tenuous upper portion, this value may be below where the first cloud particles are found. Cloud top pressure is determined from cloud top temperature, which the satellite measures more directly, using a profile of atmospheric temperature with pressure. It can be considered as equivalent to cloud top height above mean sea level.

The satellite measures the infrared radiation emitted by the cloudy scene, which is assumed to be covered completely by clouds at one level. If the cloud is optically thick, it radiates like a blackbody so that the observed emission is equivalent to the actual temperature at the top (if the upper portion of the cloud is tenuous, then the radiating level will be within the cloud). If the cloud is optically thin, then the emission will appear to be larger than that for the cloud top temperature because additional radiation is transmitted from the warmer atmosphere and surface below. The ISCCP parameter has been corrected for this effect using the measured optical thickness to determine the transmission (daytime only). The actual emission from a larger area, about 280 km across, is produced by a mixture of clouds at different levels; the deviation of this emission from that by a single cloud layer with the average temperature increases as the mesoscale variations of temperature increase. The mesoscale variability of cloud top temperature is represented by the time-averaged, spatial standard deviation.

This parameter represents the optical thickness of clouds at visible wavelengths (approximately 0.6 microns). It is determined from the satellite-measured visible solar reflectivity of cloudy scenes, assuming that the pixel is uniformly covered by clouds. The retrieval depends on an assumed particle size and shape. The standard ISCCP products assume that clouds warmer than 260 K are liquid clouds composed of spherical droplets with an effective radius of 10 microns and that colder clouds are ice clouds composed of crystals with a fractal shape (aspect ratio unity) that have an effective radius of 30 microns. The albedo of a cloud covering a larger area, about 280 km across, is reduced by the mesoscale variations of optical thickness; however, the ISCCP optical thickness values are averaged so as to give the correct albedo. If the mesoscale variability is needed to correct for this variability, then we provide a parameter, epsilon, that can be used for this purpose.

This parameter combines the measured cloud optical thickness with the assumed particle size distribution to calculate the column mass density of the cloud. The values given assume that the whole cloud layer has the same water phase and particle size as at the top.

An advanced analysis, exploiting an additional wavelength, allows for the retrieval of cloud particle size and the consistent optical thickness. Since these data are available only from the polar orbiting weather satellites and since the measurement uses reflected sunlight, we obtain only one measurement per day from each polar orbiter. We provide the climatology of cloud particle sizes and revised optical thickness values here.

A value of cloud top pressure and optical thickness is obtained for each cloudy pixel during the daytime. This information can be used to classify different cloud types as shown in the figure. The cloud type names here represent only an approximate climatological relationship between the satellite-measured optical parameters and the classical morphological cloud types. However, a detailed comparison of the satellite and surface-based cloud observations supports this assignment of names.

ISCCP Data Products | ISCCP Web Site Highlights