28 Mar 2023
 | 28 Mar 2023
Status: a revised version of this preprint is currently under review for the journal TC.

Improving Arctic sea ice thickness retrieved from CryoSat-2: A comprehensive optimization of a retracking algorithm, radar penetration rate, and snow depth

Yi Zhou, Yu Zhang, Changsheng Chen, Lele Li, Danya Xu, Robert C. Beardsley, and Weizeng Shao

Abstract. The Arctic sea ice thicknesses retrieved from the CryoSat-2 satellite are significantly influenced by sea ice surface roughness, snow backscatter, and snow depth on the sea ice. This study is the first to improve the retrieval of sea ice thickness from CryoSat-2 data derived by the Alfred Wegener Institute (AWI CS2) through applying a comprehensive optimization of an improved retracking algorithm, corrected radar penetration rate, and new snow depth. The radar freeboard data was obtained from the improved retracking algorithm of the Lognormal Altimeter Retracker Model (LARM). The radar penetration rates were corrected to 0.77 for first-year ice (FYI), 0.96 for multi-year ice (MYI), and 0.91 for all ice types. The new snow depth data was derived from the Chinese satellite Feng Yun-3B with the MicroWave Radiometer Imager (FY3B/MWRI). Three individual and one combined optimization cases were created by focusing on the retracking algorithm, radar penetration rate, and snow depth, which were validated with in situ observations collected from the National Aeronautics and Space Administration Operation IceBridge (OIB), Ice Mass Balance buoys (IMB), CryoSat Validation Experiment (CryoVEX), and Alfred Wegener Institute IceBird Program (AWI IceBird). The assessment results showed that all the optimization cases had the ability to effectively improve the sea ice thickness, with similar correlation coefficients. In the validation with the four kinds of observed data, the optimization cases reduced the RMSE of AWI CS2 up to 0.23 m (25.0 %), 0.27 m (29.7 %), 0.26 m (25.5 %), and 0.22 m (23.9 %). The improved sea ice thickness retrieval retained the major distribution patterns generated by AWI CS2, but generally showed thinner sea ice thickness. In particular, in the MYI region, the difference in thickness became increasingly evident from fall to spring. The differences in the variation trend between the optimization cases and AWI CS2 were significant in some coastal regions and the central Arctic. The experiments revealed that the radar penetration rate calculation was more sensitive to sea ice density than to snow density. The sensitivity experiments suggested that the snow depth of TOPAZ4, in addition to that of FY3B/MWRI, was also applicable in improving the retrieval of sea ice thickness. The updated scheme of sea ice densities (FYI = 925 kg m−3 and MYI = 902 kg m−3) can be combined with the use of comprehensive optimization to improve the retrieval of sea ice thickness. This successful optimization provided new insights into improving the sea ice thickness retrieved from CryoSat-2 and helped to further understand and quantify the spatiotemporal variations of sea ice thickness.

Yi Zhou et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Yi Zhou et al.

Yi Zhou et al.


Total article views: 666 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
455 186 25 666 13 16
  • HTML: 455
  • PDF: 186
  • XML: 25
  • Total: 666
  • BibTeX: 13
  • EndNote: 16
Views and downloads (calculated since 28 Mar 2023)
Cumulative views and downloads (calculated since 28 Mar 2023)

Viewed (geographical distribution)

Total article views: 648 (including HTML, PDF, and XML) Thereof 648 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 02 Dec 2023
Short summary
This study used an improved retracking algorithm, considered the corrected radar penetration rates, and included the new snow depth from the Feng Yun-3B satellite to enhance the accuracy of Arctic sea ice thickness derived from the CryoSat-2 satellite. This comprehensive optimization was the first to improve the sea ice thickness retrieval. Compared with the sea ice product derived by the Alfred Wegener Institute, the optimization cases could successfully reduce the errors above 20 %.