CLOUD COVER FROM AVHRR：ALGORITHMS AND THOUGHTS
Luo Gang（UCAR Visiting Scientist Program，NOAA Science Center，Washington，DC 2 0233，USA）
Owing to the important role clouds play in atmospheric radiation，many algorithms have been developed to retrieve cloud cover from Advanced Very High Resolution Radiometer（AVHRR）data. The threshold technique，used in International Satellite Cloud Climatology Project （ ISCCP），is a straightforward method and has provided the scientific community valuable results，Spatial Coherence Method，aiming at allowing for partly cloud filled satellite pixels，successfully identifies cloud layers and improves the quality of cloud overcast plxels．CLAVR（Clouds from AVHRR）and APOLLO（AVHRR Processing over land cloud and ocean），by making use of the MCSST（Multiple Channel Sea Srrface Temperature）retrieval，enables cloud type classification and improves the quality of cloud-free pixels. Bispectral Method，with the best- fit between the satellite data and theoretical calculations，obtains cloud cover，cloud particle size and cloud emissivitv simultaneously. To establish apromising cloud cover retrieval scheme，it seems nscessary to combine the valuable scientific thoughts behind each of the developed algorithms.