Global, Seasonal Cloud Variations from Satellite Radiance Measurements. Part I: Sensitivity of Analysis

Abstract

Global, daily, visible and infrared radiance measurements from the NOAA-5 Scanning Radiometer (SR) are analyzed for the months of January, April, July and October 1977 to infer cloud and surface radiative properties. In this first paper in a three part series, the data and analysis method are described. A unique feature of the method is that it utilizes radiative transfer models that simulate the SR measurements using explicit parameters representing the properties of the surface, atmosphere, and clouds. The simulations also account for variations that depend on viewing geometry. The analysis combines several datasets so that the cloud contributions to the SR measurements can be isolated. The accuracy of all the results depends primarily on the proper separation of the total radiance distribution into those parts representing clear and cloudy scenes. Comparison of the surface properties retrieved from the clear scene radiances [see also Rossow et al. (1989)], sensitivity tests of the cloud detection algorithm, and comparisons of the resulting cloud amounts (see also Part II) provide an assessment of the accuracy of the method.