Exploiting far-infrared signals of water vapour and ice clouds: A new retrieval capability developed in support of the ESA FORUM mission

This thesis presents the extension of the Rutherford Appleton Laboratory Infrared Microwave Sounding (IMS) optimal estimation retrieval scheme into the far-infrared in preparation for the European Space Agency’s Far-infrared Outgoing Radiation Understanding and Monitoring (FORUM) mission. Studies have indicated that upwelling far-infrared radiances (100-667cm−1) hold significant information about ice cloud microphysics, and the distribution of water vapour. This work has extended IMS to exploit these far-infrared signals, and applied it to observations of upwelling far-infrared radiances that have been adapted to mimic the FORUM Sounding Instrument’s (FSI) spectral characteristics. The work in this thesis is presented in three stages. First, the retrieval scheme has been optimised for clear-sky retrievals from simulated FSI radiances, and compared against the existing IMS configuration for the Infrared Atmospheric Sounding Interferometer (IASI). Comparable retrieval biases were observed for temperature and water vapour, however there was an increase of∼1 degree of freedom for water vapour and temperature for the FSI configuration. Second, IMS has been applied to eight clear-sky observations of upwelling far- and mid-infrared radiances taken during the PIKNMIX-F campaign. Retrievals from these radiances using the modified code showed a strong agreement with contemporaneous in-situ measurements of the atmospheric state, reducing the RMSE by 18% for water vapour from the a-priori, giving confidence in its performance. Finally, IMS has been used to perform the first retrieval of ice cloud properties (cloud optical thickness, cloud effective radius, cloud top height, ice crystal habit) from an airborne observation of upwelling far- and mid-infrared radiances taken above a cirrus cloud layer. Despite the known limitations of IMS, the retrieved values agree with in-situ measurements of the cloud. The inclusion of the far-infrared enables a distinction between two habits that was not possible using only the mid-infrared (667-2000 cm−1), and halves the uncertainties in the retrieved outputs.