Generalization of the complete data fusion to multi-target retrieval of atmospheric parameters and application to FORUM and IASI-NG simulated measurements
Cecilia Tirelli, Simone Ceccherini, Nicola Zoppetti, Samuele Del Bianco, Ugo Cortesi
Abstract
In the context of a growing need for innovatory techniques to take advantage of the largest amount of information from the great number of available remote sensing data, the Complete Data Fusion (CDF) algorithm was presented as a new method to combine independent measurements of the same vertical profile of an atmospheric parameter into a single estimate for a concise and complete characterization of the atmospheric state. The majority of the atmospheric composition measurements determine the altitude distribution of a great number of quantities: multi-target retrievals (MTRs) are increasingly applied to remote sensing observations to determine simultaneously atmospheric constituents with the purpose to reduce the systematic error caused by interfering species. In this work, we optimised the CDF for the application to MTR products. We applied the method to simulated retrievals in the thermal infrared and in the far infrared spectral ranges, considering the instrumental specifications and performances of IASI-NG (Infrared Atmospheric Sounding Interferometer New Generation) and FORUM (Far-Infrared Outgoing Radiation Understanding and Monitoring) instruments, respectively. The obtained results show that the CDF algorithm can cope with state vectors from MTRs, that must share at least one retrieved variable. In particular, the results show that the fused profile has the greatest number of degrees of freedom and the smallest error for all considered cases. The comparison between the CDF products and the synergistic retrieval ones shows the equivalence of the two methods when the linear approximation is adopted to simplify the treatment of the retrieval problem.