a Climate Index Derived from Satellite Measured Spectral Infrared Radiation.

This thesis introduces a climate index based on radiative transfer theory and derived from the spectral radiances typically used to retrieve temperature profiles. It is assumed that clouds and climate are closely related and a change in one will result in a change in the other. Because the index developed in this paper is a function of the cloud, temperature, and moisture distributions, it may be used as a climate index. The advantage is that the index is more accurately retrieved from satellite data than cloudiness per se. This index, hereafter referred to as the VIRES index (for Vertical Infrared Radiative Emitting Structure), is based upon the shape and relative magnitude of the broadband weighting function of the infrared radiative transfer equation. The broadband weighting curves are retrieved from simulated satellite infrared sounder data (spectral radiances). This thesis describes the retrieval procedure and investigates error sensitivities of this method. It also proposes index measuring options and possible applications of the VIRES index. Results indicate that the VIRES approach is a very effective use of satellite radiometer measurements. Retrieval advantages include; day and night capability, no need to know cloud radiative properties, retrieval ability when cloud fraction or cloud emittance is less than 1.0, minimal geometric assumptions, retrieved information below sensor resolution, and minimum influence on the index from low tropospheric retrieval errors. These advantages along with the approach of compositing scenes for an average VIRES curve greatly reduce the retrieval sensitivity to the kind of errors found under assumed normal operating conditions. A detailed error analysis indicated that the most important error sources are instrument system noise, and ill specified temperature and humidity profiles. Accurate VIRES retrievals are illustrated under a number of different error and atmospheric conditions. Furthermore, a statistical technique used to successfully discriminate between VIRES curves derived for different atmospheric conditions is discussed. Operational VIRES index retrieval strategies and a number of specific applications are proposed. It is suggested that the index be derived from geostationary satellite data and averaged to provide weekly regional values. These index values would be used in a regional climate monitoring mode. They would also be useful for verification of climate model generated infrared radiation to space values.