Antarctic Ice Cloud Identification and Properties Using Downwelling Spectral Radiance From 100 to 1, 400 cm−1. Issue 8 (29th April 2019)

• Record Type: Journal Article
• Authors: Maestri, Tiziano; Arosio, Carlo; Rizzi, Rolando; Palchetti, Luca; Bianchini, Giovanni; Del Guasta, Massimo
• Is Part Of: Journal of geophysical research. Volume 124:Issue 8(2019)
• Page Start: 4761
• Page End: 4781
• Publication Date: 2019-04-29
• Subjects: far infrared -- cirrus clouds -- cloud properties retrievals; Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1029/2018JD029205
• Languages: English
• ISSNs: 2169-897X

Abstract

Abstract: One year (2013) of high spectral resolution measurements of downwelling radiance in the 100–1, 400 cm −1 range, taken by the Fourier Transform Spectrometer REFIR‐PAD at the research station Concordia (Antarctic Plateau), is analyzed. Optically thin ice clouds are identified by means of a new identification/classification tool based on a Support Vector Machine algorithm. The use of transparent microwindow channels in the Far InfraRed (FIR) spectral region (100–667 cm −1 ) is shown to be of great importance in the identification and classification of cloud type. In particular, the channels between 380 and 575 cm −1 are key channels for the clear/cloud and phase identification due to their sensitivity to cloud properties; in addition, FIR channels down to 238 cm −1 are exploited for the selection of precipitating or nonprecipitating cases because of their sensitivity also to water vapor content. A subset of 26 cases of nonprecipitating ice clouds is selected based on the presence of colocated LIDAR and radiosonde data. REFIR‐PAD channel in the 800–1, 000 cm −1 are used to derived optical and microphysical properties for four different assumptions concerning the crystal habits. Results, showing a correlation between cloud base temperature, optical depth, and particle size distribution effective dimensions, are compared with what found in literature. Based on the retrievals, forward simulations are also run over the whole sensor spectral interval, and results are compared to data. The simulation‐data residuals in the FIR are evaluated for selected "window" channels and analyzed in relation to crystal's habit assumption, cloud retrieved features, and atmospheric water vapor content. Key Points: Spectral radiance measurements in the far infrared are used for cloud identification/classification Cloud properties are retrieved