Snow Property Controls on Modeled Ku-Band Altimeter Estimates of First-Year Sea Ice Thickness: Case Studies From the Canadian and Norwegian Arctic

Vishnu Nandan*, Randall K. Scharien, Torsten Geldsetzer, Ronald Kwok, John J. Yackel, Mallik S. Mahmud, Anja Rosel, Rasmus Tonboe, Mats Granskog, Rosemary Willatt, Julienne Stroeve, Daiki Nomura, Markus Frey

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


Uncertainty in snow properties impacts the accuracy of Arctic sea ice thickness estimates from radar altimetry. On first-year sea ice (FYI), spatiotemporal variations in snow properties can cause the Ku-band main radar scattering horizon to appear above the snow/sea ice interface. This can increase the estimated sea ice freeboard by several centimeters, leading to FYI thickness overestimations. This article examines the expected changes in Ku-band main scattering horizon and its impact on FYI thickness estimates, with variations in snow temperature, salinity, and density derived from ten naturally occurring Arctic FYI Cases encompassing saline/nonsaline, warm/cold, simple/complexly layered snow (4-45 cm) overlying FYI (48-170 cm). Using a semi-empirical modeling approach, snow properties from these Cases are used to derive layer-wise brine volume and dielectric constant estimates, to simulate the Ku-band main scattering horizon and delays in radar propagation speed. Differences between modeled and observed FYI thickness are calculated to assess sources of error. Under both cold and warm conditions, saline snow covers are shown to shift the main scattering horizon above from the snow/sea ice interface, causing thickness retrieval errors. Overestimates in FYI thicknesses of up to 65% are found for warm, saline snow overlaying thin sea ice. Our simulations exhibited a distinct shift in the main scattering horizon when the snow layer densities became greater than 440 kg/m 3 , especially under warmer snow conditions. Our simulations suggest a mean Ku-band propagation delay for snow of 39%, which is higher than 25%, suggested in previous studies.
Original languageEnglish
Pages (from-to)1082-1096
Number of pages15
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Publication statusPublished - 17 Feb 2020
Externally publishedYes

Cite this