Global-fit approach to the analysis of limb-scanning atmospheric measurements.

A method for the retrieval of concentration profiles of atmospheric constituents from spectra, recorded by balloon-borne spectrometers with the limb-scanning technique, is presented. The method uses a nonlinear least-squares fit procedure to fit simultaneously the whole concentration profile on a limb-scanning sequence of spectra. A use in interferometric measurements of the stratospheric emission is shown and a comparison is discussed with the results obtained from the analysis of the same data set, by using the onion-peeling method in which the error propagation, over concentrations, is taken into account. With the global-fit, error bars smaller than with the onion-peeling analysis are obtained. Computational details are also discussed.