Spatio-temporal variability and decadal trends of snowmelt processes on Antarctic sea ice observed by satellite scatterometers

• Published in: The Cryosphere
• Main Authors: Arndt, Stefanie, Haas, Christian
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Copernicus 2019
• Subjects: Antarc*; Antarctic; Antarctica; Sea ice; The Cryosphere; Weddell Sea
• Online Access: https://epic.awi.de/id/eprint/52664/; https://epic.awi.de/id/eprint/52664/1/tc-13-1943-2019.pdf; https://hdl.handle.net/10013/epic.754dbe7a-9e6a-408e-ba9a-b5f98d31b8fd

The timing and intensity of snowmelt processes on sea ice are key drivers determining the seasonal sea-ice energy and mass budgets. In the Arctic, satellite passive microwave and radar observations have revealed a trend towards an earlier snowmelt onset during the last decades, which is an important aspect of Arctic amplification and sea ice decline. Around Antarctica, snowmelt on perennial ice is weak and very different than in the Arctic, with most snow surviving the summer. Here we compile time series of snowmelt-onset dates on seasonal and perennial Antarctic sea ice from 1992 to 2014/15 using active microwave observations from European Remote Sensing Satellite (ERS-1/2), Quick Scatterometer (QSCAT) and Advanced Scatterometer (ASCAT) radar scatterometers. We define two snowmelt transition stages: A weak backscatter rise indicating the initial warming and destructive metamorphism of the snowpack (pre-melt), followed by a rapid backscatter rise indicating the onset of thaw-freeze cycles (snowmelt). Results show large interannual variability with an average pre-melt onset date of 29 November and melt onset of 10 December, respectively, on perennial ice, without any significant trends over the study period, consistent with the small trends of Antarctic sea ice extent. There was a latitudinal gradient from early snowmelt onsets in mid-November in the northern Weddell Sea to late (end-December) or even absent snowmelt conditions in the southern Weddell Sea. We show that QSCAT Ku-band (13.4 GHz signal frequency) derived pre-melt and snowmelt onset dates are earlier by 20 and 18 days, respectively, than ERS and ASCAT C-band (5.6 GHz) derived dates. This offset has been considered when constructing the time series. Snowmelt onset dates from passive microwave observations (37 GHz) are later by 14 and 6 days than those from the scatterometers, respectively. Based on these characteristic differences between melt onset dates observed by different microwave wavelengths, we developed a conceptual model which illustrates ...