Top/bottom multisensor remote sensing of Arctic sea ice

Abstract

The Arctic sea ice cover has been studied using near simultaneous observations by passive and active (synthetic aperture radar, SAR) microwave sensors, upward looking and sidescan sonars, a lidar profilometer, and an infrared sensor. Data from two aircraft and a submarine over an approximately 100 km track of central Arctic sea ice were registered and analyzed to evaluate the characteristics of the ice cover and the utility of each sensor in ice studies. The results of comparative and correlation analyses are as follows. The probability density functions of ice draft from sonar and elevation from lidar were found to be almost identical when isostasy is taken into account. This result suggests that the basic ice thickness distribution can be derived from the surface topography measurements alone. Reasonable correlation was found between SAR backscatter and ice draft (or elevation) especially when scales were adjusted such that 15–20 SAR pixels were averaged. However, surface roughness derived directly from standard deviations in the lidar elevation data was found to be poorly correlated to the SAR backscatter. These results indicate that the SAR values are affected more by scattering from the ice than from the snow‐covered surface. The active and passive microwave sensors are shown to generally complement each other as the two sensors are especially sensitive to different physical properties of the sea ice. Undeformed first‐year ice showed low backscatter values but high brightness temperatures while some multiyear ice showed high backscatter values and low brightness temperature. However, surfaces identified as multiyear ice by the passive system have a large spread in the unaveraged SAR backscatter, indicating limitations when using a one‐channel SAR for ice type identification at the highest resolution. Also, ridged ice identified by sonar and SAR data covers a large range of passive microwave emissivity, suggesting considerable variability in the age and salinity of this type of ice. Significant variations (about 0.11) in the minimum emissivity of consolidated multiyear ice are observed in different regions of the Arctic using the high‐resolution (30 m) passive microwave data. This suggests that regional variations in texture and scattering characteristics of multiyear ice in the Arctic are present, likely influenced by different histories of formation of the ice in different regions.