Abstract

Remote sensing instruments to be flown on planned NASA missions will generate unprecedented quantities of data. It is essential that these data be processed rapidly and the results conveyed quickly to the scientific community. Such considerations mandate substantial advances in current techniques for analyzing data. In particular major improvements in both computation technology and numerical analysis methods are necessary. We examine the extent to which sequential parameter estimation can be used to automate and expedite the retrieval of atmospheric parameters from remote sensing data.

Our algorithm utilizes a square root information filter to retrieve the vertical profiles of trace gases from spectroscopic data. The algorithm permits simultaneous retrieval of many atmospheric parameters such as species abundance, atmospheric temperature, atmospheric pressure, and water vapor profiles. Implementation of this algorithm in a parallel processing environment should result in dramatic increases in the rate at which data can be analyzed.

© 1990 Optical Society of America