HR: 0830h
AN: GC31B-0187    [PDF]
TI: Testing the Dynamical Properties of Climate Models with Observations of Spectrally Resolved Thermal Radiance
AU: * Clark, D
AF: Harvard University, 12 Oxford Street, Cambridge, MA 02138 United States
AU: Dykema, J A
EM: dykema@fas.harvard.edu
AF: Harvard University, 12 Oxford Street, Cambridge, MA 02138 United States
AU: Anderson, J G
AF: Harvard University, 12 Oxford Street, Cambridge, MA 02138 United States
AB: Spectrally resolved, absolutely calibrated infrared radiance measurements made from space are being increasingly recognized as a key observational element of a program to detect and forecast future climate change. Space-based measurements have a number of important characteristics for climate research. They can provide complete global coverage, completing our knowledge of isolated ocean regions where in situ measurements are sparse. The homogeneity of these measurements reduces the complexity and uncertainty inherent in their analysis. Being provided by a single platform means that a careful absolute calibration for one instrument provides reproducible and demonstrable accuracy for an entire mission-length global dataset. One of the main disadvantages of space deployment is that instruments can only directly measure radiance, not the climatic state variables to which we are most accustomed. Although the radiances may be inverted to produce atmospheric profiles and records of other state variables, the complexity and ill-conditioned nature of this problem (Keith and Anderson 1998) make analysis of the radiances themselves a very attractive option. The analysis of the statistical properties of previous spectrally infrared missions has illustrated the detailed information that can be extracted from the means (Harries et al. 2001) and the higher statistical moments (Haskins and Goody 1998). The higher order moments are particularly important in understanding the response of the climate to an external forcing (Leith 1975). In this paper we examine the information content of the dynamical characteristics of a simulated infrared radiance dataset. We show that the dynamical properties can be separated from the statistical properties, yielding different information about the physical mechanisms operating in the model atmosphere. Changes in parameterizations of physical processes in a simple one-dimensional model are shown to independently affect the covariances of the radiation field determined from the empirical orthogonal functions (EOF) and dynamics determined from linear inverse modeling (LIM).
DE: 0360 Transmission and scattering of radiation
DE: 0394 Instruments and techniques
SC: Global Climate Change [GC]
MN: 2003 Fall Meeting