Advances in Understanding Clouds from ISCCP

This progress report on the International Satellite Cloud Climatology Project (ISCCP) describes changes made to produce new cloud data products (D data), examines the evidence that these changes are improvements over the previous version (C data), summarizes some results, and discusses plans for the ISCCP through 2005. By late 1999 all datasets will be available for the period from July 1983 through December 1997. The most significant changes in the new D-series cloud datasets are 1) revised radiance calibrations to remove spurious changes in the long-term record, 2) increased cirrus detection sensitivity over land, 3) increased low-level cloud detection sensitivity in polar regions, 4) reduced biases in cirrus cloud properties using an ice crystal microphysics model in place of a liquid droplet microphysics model, and 5) increased detail about the variations of cloud properties. The ISCCP calibrations are now the most complete and self-consistent set of calibrations available for all the weather satellite imaging radiometers: total relative uncertainties in the radiance calibrations are estimated to be ≲ 5% for visible and ≲ 2% for infrared; absolute uncertainties are < 10% and < 3%, respectively. Biases in (detectable) cloud amounts have been reduced to ≲ 0.05, except in the summertime polar regions where the bias may still be ~ 0.10. Biases in cloud-top temperatures have been reduced to ≲ 2 K for lower-level clouds and ≲ 4 K for optically thin, upper-level clouds, except when they occur over lower-level clouds. Using liquid and ice microphysics models reduces the biases in cloud optical thicknesses to ≲ 10%, except in cases of mistaken phase identification; most of the remaining bias is caused by differences between actual and assumed cloud particle sizes and the small effects of cloud variations at scales < 5 km. Global mean cloud properties averaged over the period July 1983–June 1994 are the following: cloud amount = 0.675 ± 0.012; cloud-top temperature = 261.5 ± 2.8 K; and cloud optical thickness = 3.7 ± 0.3, where the plus–minus values are the rms deviations of global monthly mean values from their long-term average. Long-term, seasonal, synoptic, and diurnal cloud variations are illustrated. The ISCCP dataset quantifies the variations of cloud properties at mesoscale resolution (3 h, 30 km) covering the whole globe for more than a decade, making it possible to study cloud system evolution over whole life cycles, watching interactions with the atmospheric general circulation. Plans for the next decade of the World Climate Research Programme require continuing global observations of clouds and the most practical way to fulfill this requirement is to continue ISCCP until it can be replaced by a more capable system with similar time resolutions and global coverage.