Arctic sea ice parameters from AMSR‐E data using two techniques and comparisons with sea ice from SSM/I

Abstract

We use two algorithms to process Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR‐E) data in order to determine algorithm dependence, if any, on the estimates of sea ice concentration, ice extent and area, and trends and to evaluate how AMSR‐E data compare with historical Special Sensor Microwave Imager (SSM/I) data. The monthly ice concentrations derived from the two algorithms from AMSR‐E data (the AMSR‐E bootstrap algorithm, or ABA, and the enhanced NASA team algorithm, or NT2), differ on average by about 1–3%, with data from the consolidated ice region being comparable for ABA and NT2 retrievals while data in the marginal ice zones and thin ice regions show slightly higher values when the NT2 algorithm is used. The iceextents and areas derived separately from AMSR‐E using these two algorithms are, however, in good agreement, with the differences (ABA – NT2) being about 0.07 × 106 km2 on average for ice extents and −0.07 × 106 km2 for ice area, which are small compared with mean seasonal values of 10.5 × 106 and 9.8 × 106 for ice extent and area, respectively. Likewise, extents and areas derived from the same algorithm but from AMSR‐E and SSM/I data are consistent but differ by about 0.24 × 106 km2 and 0.14 × 106 km2, respectively. The discrepancies are larger with the estimates of extents than area mainly because of differences in channel selection and sensor resolutions. Trends in extent during the AMSR‐E era were also estimated, and results from all three data sets are shown to be in good agreement (within errors).