Application of Neural Algorithms for a Real-Time Estimation of Ozone Profiles From GOME Measurements.

The thermal structure of trace gases, their distribution in the atmosphere, and their circulation mechanisms result from a complex interplay between radiative, physical, and dynamical processes. Neural-network algorithms can be a useful tool to face such complexities in retrieval operations. In this paper, their potentialities have been exploited to design real-time procedures for the estimation of vertical profiles of ozone concentration from spectral radiances measured by GOME, the first instrument of the European Space Agency capable of monitoring global distribution of ozone and other trace gases. [ABSTRACT FROM AUTHOR]