Cloud optical property retrievals from layered-cloud radiances derived from AVHRR data

Abstract

This paper describes a new methodology forcomputing cloud optical properties from the newcloud remote sensing algorithms ofNOAAINESDIS. The recently enhanced cloudremote sensing algorithms of NOAAINESDISdescribed by Stowe et al (1998) and Davis et al(1998) compute cloud amounts and the radiancestatistics for each AVHRR channel for each layerof each cloud type. The ultimate aim of this studyto produce a multi-year cloud propertyclimatology based on the AVHRR data recordsimilar to the AVHRR Pathfinder reprocessingeffort (Stowe and Jacobowitz, 1998) whosecloud products were limited to total cloudamount and to total cloud radiances.The goal of this study is to demonstratetechniques that can be used to estimate cloudoptical properties from the layered-cloudradiances including the cloud optical depth, theparticle effective radius and the cloud toppressure. Currently, CLAVR computes the firsttwo moments of the radiance distribution for theensemble of pixels of each layer and each cloudtype over a 1° equal-area grid-cell. Using theensemble radiance statistics to perform retrievalsis a factor of N times faster than the standardpixel by pixel retrieval approach where N is thenumber of pixels of any one cloud type in a grid-cell. This increase in computational speed iscrucial to this study since the ultimate outcomewill be cloud property climatology for theapproximately 20-year AVHRR data-record.To illustrate the type of information produced bythe current CLAVR processing system, Fig 1shows a sample AVHRR image of the channel 4