High-Resolution Snow Freeboard Maps from Operation IceBridge (OIB) ATM Measurements and Sea Ice Type Products based on Sentinel-1 data

This record contains two datasets: 1 m resolution snow freeboard (Fs) maps and ice type products based on Sentinel-1 data, related to article "On the Statistical Relationship between Sea Ice Freeboard and C-Band Microwave Backscatter -- A Study with Sentinel-1 and Operation IceBridge". 1. The first dataset contains processed Fs measurements derived from NASA's Airborne Topographic Mapper (ATM) observations during four April 2019 Arctic campaigns of Operation IceBridge (OIB), with particular focus on the April 8 and 12 flights which were organized into racetracks. This dataset related to article "On the Statistical Relationship between Sea Ice Topography and C-Band Microwave Backscatter – A Study with Sentinel-1 and Operation IceBridge." Specifically, two steps are carried out, as follows. (1) Construction of the per-pass 1m-scale Fs map As the first step, for each OIB pass, we converted OIB ATM samples into the Fs map. Both wide scan and the narrow scan of the OIB ATM are utilized. We first project each ATM sample under the polar stereographic projection according to its geolocation (i.e., its latitude and longitude). Then, we interpolate the samples into a 1m-scale elevation map, using linear interpolation. Afterwards, we apply corrections to the elevation, including the mean sea-surface height (DTU15 MSS model) and the atmospheric and tidal effects. Finally, we treat the corrected elevation as elevation anomalies, and apply the lowest elevation method to retrieve the freeboard. Specifically, the lowest 1‰ of elevation samples within each 10km segment are extracted and interpolated to construct the local water level (also at 1m-scale) using the Inverse Distance Weighting (IDW) method. The final 1m-scale Fs map is further validated with the standard 40m-scale Fs product from IDCSI. (2) Collocation between OIB passes and the construction of the merged Fs field We further merge the three OIB passes to form the Fs map that covers over 1.4km across the flight path. Since the central pass and the left pass were separated by 1∼2 hours, and the central pass and the right pass by 3∼4 hours, the sea ice cover potentially had undergone drift and deformation. Therefore, we first search for corrections between each of the two pairs of OIB passes. For each 3km segment, we maximize the correlation of the overlapping part of the Fs maps of the central and the left (or the right) pass, by adjusting the relative location of the left (or the right) pass with respect to the central pass. After the maximum correlation is attained, we record the corrections in both the along-track and the cross-track directions, and further merge the left and the right pass to the central pass, in order to form a unified Fs map. For detailed data descriptions, please refer to the manual document. 2. The second dataset is about the sea ice classification based on S1 EW images. The sea ice classification algorithm used in this study is based on: Lohse et al. (2020, 2021); Guo et al. (2023). Lohse et al.(2020) developed a supervised algorithm that accounts for the class-dependent IA effects, known as the GIA classifier. While this classifier performs well in addressing IA sensitivity, some misclassifications and ambiguities remain. To address these issues, Lohse et al. (2021) and Guo et al. (2023) enhanced the algorithm by incorporating GLCM texture features, resulting in improved separation between classes. This study uses this classification approach to produce sea ice type maps on the selected S1 scenes. The cross-polarization channel (HV) of the selected S1 scenes are de-noised using an adapted version of the Kalman-filter-based noise removal algorithm developed by Iqbal et al. (2012). Lohse, J., Doulgeris, A. P., and Dierking, W.: Mapping sea-ice types from Sentinel-1 considering the surface-type dependent effect of incidence angle, Annals of Glaciology, 61, 260–270, https://doi.org/10.1017/aog.2020.45, 2020. Lohse, J., Doulgeris, A. P., and Dierking, W.: Incident Angle Dependence of Sentinel-1 Texture Features for Sea Ice Classification, RemoteSensing, 13, https://doi.org/10.3390/rs13040552, 2021. Guo, W., Itkin, P., Singha, S., Doulgeris, A. P., Johansson, M., and Spreen, G.: Sea ice classification of TerraSAR-X ScanSAR images for the MOSAiC expedition incorporating per-class incidence angle dependency of image texture, The Cryosphere, 17, 1279–1297, https://doi.org/10.5194/tc-17-1279-2023, 2023. Iqbal, M., Chen, J., Yang, W., Wang, P., and Sun, B.: Kalman filter for removal of scalloping and inter-scan banding in scansar images, Prog. Electromagn. Res., 132, pp. 443–461, 2012.