Retrieval of Arctic Sea Ice Parameters by Satellite Passive Microwave Sensors: A Comparison of Eleven Sea Ice Concentration Algorithms

Abstract

Eleven algorithms for the passive-microwave measurement of sea-ice were implemented and inter-compared. Daily, monthly and annual Arctic sea-ice concentration, area and extent were calculated by the algorithms using daily microwave brightness temperatures from SMMR, SSM/I, and SSMIS for the period 1979–2012. The differences between the 11 sea-ice concentration estimates—structural uncertainties—were quantified and analyzed spatially and seasonally. The algorithms differ in annual sea-ice area by 0.0–1.3 million $\hbox{km}^{2}$ and in extent by 0.0–0.6 million $\hbox{km}^{2}$ . Linear trends for 34- and 21-year periods were calculated and compared for sea-ice concentration, area and extent. Low-frequency algorithms obtained annual Arctic sea-ice area decrease of 0.534–0.573 million $\hbox{km}^{2}$ per decade (0.439–0.491 million $\hbox{km}^{2}$ per decade for the extent) over the period 1979 to 2012, and a decrease of 0.866–0.975 million $\hbox{km}^{2}$ per decade (0.767–0.812 million $\hbox{km}^{2}$ per decade for the extent) for the 1992–2012 period. High-frequency algorithms obtained a decrease of 0.766–0.978 million $\hbox{km}^{2}$ per decade in the area and 0.758–0.814 million $\hbox{km}^{2}$ per decade in the extent over the period 1992–2012. Results for all the algorithms have close agreement on the strength of the negative trend in Arctic sea-ice area and extent, but are individually biased from the mean. The algorithms' ensemble mean and standard deviation in sea-ice concentration, describing part of the uncertainty, are presented to provide users with more insight into the uncertainties and potential biases of sea-ice concentration data.