Cloud variability as revealed in outgoing infrared spectra: Comparing model to observation with spectral EOF analysis

Abstract

Spectrally resolved outgoing radiance is a potentially powerful tool for testing climate models. To show how it can be used to evaluate the simulation of cloud variability, which is the principal uncertainty in current climate models, we apply spectral empirical orthogonal function (EOF) analysis to satellite radiance spectra and synthetic spectra derived from a general circulation model (GCM). We show that proper averaging over a correct timescale is necessary before applying spectral EOF analysis. This study focuses on the Central Pacific and the western Pacific Warm Pool. For both observation and GCM output, cloud variability is the dominant contributor to the first principal component that accounts for more than 95% of the total variance. However, the amplitude of the first principal component derived from the observations (2 ∼ 3.4 W m-2) is 2 ∼ 6 times greater than that of the GCM simulation. This suggests that cloud variability in the GCM is significantly smaller than that in the real atmosphere.