High-resolution maps of Arctic surface skin temperature and type retrieved from airborne thermal infrared imagery collected during the HALO–(AC)3 campaign

Abstract

Two retrieval methods for the determination of Arctic surface skin temperature and surface type based on radiance measurements from the thermal infrared (TIR) imager VELOX (Video airbornE Longwave Observations within siX channels) were developed. VELOX captured TIR radiances in terms of brightness temperatures for wavelengths from 7.7 to 12 µm in six spectral channels. The imager was deployed on the High Altitude and LOng Range research aircraft (HALO) during the HALO–(AC)3 aircraft field campaign conducted in the framework of the Arctic Amplification: Climate Relevant Atmospheric and SurfaCe Processes and Feedback Mechanisms (AC)3 research programme. The measurements were taken over the Fram Strait and the central Arctic in March and April 2022. Radiative transfer simulations assuming cloud-free atmospheric conditions were performed showing that the influence of water vapour on the measured brightness temperature can be neglected. Therefore it was possible to apply a single-channel retrieval technique to obtain the surface skin temperature from the VELOX data. The retrieval results were compared with data from the MODerate-resolution Imaging Spectroradiometer (MODIS) showing an agreement within 2.0 K. Secondly, a pixel-by-pixel surface classification retrieval was developed using a random forest algorithm. It classifies surfaces into types of open water, sea-ice–water mixture, thin sea ice, and snow-covered sea ice. The resulting sea-ice concentrations were compared with satellite data, yielding a mean absolute difference (MAD) of 5 %. In addition, the classified pixels were aggregated into segments of the same surface type, providing different segment size distributions for all surface types. When grouped by the distance to the sea-ice edge, the segment size distribution showed a shift to fewer but larger floes in the direction of the pack ice.