Effect of summer-to-fall transition on ERS-1 SAR and SSM/I images of sea ice

Microwave images of sea ice are difficult to interpret during the seasonal transitions because of the rapidly changing ice and atmospheric conditions. In this paper the authors examine the summer-to-fall transition of sea ice from satellite active and passive microwave observations and investigate the use of a hybrid fusion algorithm for improving estimates of ice type concentration during this season. Based on previous in situ measurements the authors have observed that the backscattering coefficient of multiyear ice increases sharply at the onset of freeze-up in the Arctic. A 6-dB increase in the mean backscattering coefficient is also observed from ERS-1 synthetic aperture radar data. The change in the observed backscatter is caused by refreezing of moisture on or near the ice surface, which allows for increased penetration into the ice volume. This change in backscatter can be used to determine the onset of freeze-up in the Arctic from SAR. In addition, the authors used DMSP SSM/I data in this study to compute ice type concentration using the NASA Team algorithm. These estimates have been shown to be inaccurate during the freeze-up season. To improve estimates of ice type concentration after the onset of freeze-up we used a hybrid fusion method for incorporating information from the ERS-1 SAR into an existing passive microwave algorithm.