Analysis of far-infrared (FIR) high spectral resolution data for cloud studies

The Far-Infrared Outgoing Radiation Understanding and Monitoring (FORUM) mission has recently been selected by the European Satellite Agency (ESA) as 9 th Earth Explorer mission. FORUM mission aims at studying the water vapor and clouds by filling the long-standing gap in Far-Infrared (FIR) spectral observations from space. In the framework of the FORUM mission, this thesis analyses FIR measurements to characterize the spectral signatures of radiance in presence of ice clouds. At this purpose, a cloud identification and classification code (named CIC) is implemented. CIC is an innovative machine learning algorithm, based on principal component analysis, able to perform cloud detection and scene multi-class classification. CIC is easily adaptable to different datasets and type of spectral sensors. It is firstly tested against a synthetic dataset comprising simulated measurements of the FORUM mission. Subsequently, CIC is applied to airborne interferometric data and finally it is used for the analysis of measured downwelling radiances collected in very dry conditions on the Antarctic Plateau. Provided the excellent performances of the algorithm, especially in the identification of thin cirrus clouds, CIC is adopted as the classificator in the official ESA FORUM End-to-End simulator (FE2ES). The FE2ES is a complex chain of codes used to simulate the entire FORUM mission from satellite orbit and geometry to level 2 product analysis. An extensive use of CIC is performed on ground-based radiances collected in Antarctica. The dataset is exploited to test and to optimize the CIC algorithm and for the developing of punctual statistic of cloud occurrence in the Antarctic Plateau. Meteorological conditions from this region are also analysed and correlated with the presence of clouds. The studies presented in this work showed the potentiality and the importance of the exploitation of spectral radiance measurements in the FIR for cloud identification and classification.