Evaluation of late summer passive microwave Arctic sea ice retrievals

被引:35
作者
Markus, T
Dokken, ST
机构
[1] NASA, Goddard Space Flight Ctr, UMBC, Joint Ctr Earth Syst Technol, Greenbelt, MD 20771 USA
[2] Chalmers Univ Technol, Dept Radio & Space Sci, S-41296 Gothenburg, Sweden
来源
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING | 2002年 / 40卷 / 02期
关键词
Arctic; passive microwave remote sensing; sea ice;
D O I
10.1109/36.992795
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
The melt period of the Arctic sea ice cover is of particular interest in studies of climate change due to the albedo feedback mechanisms associated with meltponds and openings in the ice pack. The traditionally used satellite passive microwave sea ice concentration algorithms have deficiencies during the summer months due to the period's highly variable surface properties. A newly developed ice concentration algorithm overcomes some of these deficiencies. It corrects for low ice concentration biases caused by surface effects through the use of 85 GHz data in addition to the commonly used 19 and 37 GHz data and, thus, the definition of an additional ice type representing layering and inhomogeneities in the snow layer. This new algorithm will be the standard algorithm for Arctic sea ice concentration retrievals with the EOS Aqua advanced microwave scanning radiometer (AMSR-E) instrument. In this paper, we evaluate the performance of this algorithm for the summer period of 1996 using data from the special sensor microwave imager (SSM/I) which has frequencies similar to the AMSR instrument. The temporal evolution of summertime passive microwave sea ice signatures are investigated and sea ice concentration retrievals from the standard NASA team and the new algorithm are compared. The results show that the introduction of the additional sea ice type in the new algorithm leads to improved summertime sea ice concentrations. The SSM/I sea ice retrievals are validated using SAR-derived ice concentrations that have been convolved with the SSM/I antenna pattern to ensure an appropriate comparison. For the marginal ice zone, with ice concentrations ranging from 40% to 100%, the correlation coefficient of SAR and SSM/I retrievals is 0.66 with a bias of 5% toward higher SAR ice concentrations. For the central Arctic, where ice concentrations varied between 60% and 100%, the correlation coefficient is 0.87 with a negligible bias.
引用
收藏
页码:348 / 356
页数:9
相关论文
共 23 条
  • [21] Observation of perennial Arctic sea ice melt and freeze-up using passive microwave data
    Smith, DM
    [J]. JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS, 1998, 103 (C12) : 27753 - 27769
  • [22] PASSIVE MICROWAVE REMOTE-SENSING OF THIN SEA-ICE USING PRINCIPAL COMPONENT ANALYSIS
    WENSNAHAN, M
    MAYKUT, GA
    GRENFELL, TC
    WINEBRENNER, DP
    [J]. JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS, 1993, 98 (C7) : 12453 - 12468
  • [23] ZAKHAROV VF, 1997, P ARCT CLIM SYST STU, P298