Comparison of temporal cloud variability simulated by a GCM with observations from the Nimbus-7 Satellite

Abstract

This paper describes temporal cloudiness variability from Nimbus-7 satellite observations and a GCM simulation. The GCM clouds are parameterized using the cumulative normal distribution function to relate fractional cloud amount with large-scale relative humidity. The global distributions of cloudiness characteristics from both data sources are presented, including their temporal means and standard deviations. For the Nimbus-7 observations, the periods are for DJF and JJA, 1979-1985 while the simulated data are for perpetual January and July. Variances contributed by fluctuations with periods between 2.5 to 6 days, and >10 days are extracted. Both observed and simulated cloud variability fields show strong zonal dependence. High-frequency fluctuations are more predominant in the middle latitudes while low-frequency fluctuations predominate in the tropics and the polar regions. In general, unfiltered cloud variabilities are highest in regions associated with cyclolysis. The GCM performs creditably in not only simulating the seasonal means but also the variabilities in the temporal bands examined. Some reasons for the extracted cloud variability patterns at different locations and for different cloud heights are given.