[A-CG06_29AM2] Satellite Earth Environment Observation

[ACG06-20] Sea-ice production in Antarctic coastal polynyas estimated using AMSR-E data

*Sohey NIHASHI¹, Kay I. OHSHIMA² (1.Department of Mechanical Engineering, Tomakomai National College of Technolog, 2.Institute of Low Temperature Science, Hokkaido University)

Keywords:AMSR-E, Antarctic Ocean, Coastal polynyas, Sea-ice production, Antarctic Bottom Water

Coastal polynyas are newly-forming sea-ice areas formed by divergent ice drift due to prevailing winds and/or ocean currents. In coastal polynyas, huge amounts of heat flux from the ocean to the atmosphere occur because the heat insulation effect of sea-ice is greatly reduced in the case of thin ice, and accordingly sea ice is formed actively. Dense water formed in Antarctic coastal polynyas with the intense sea-ice production is a major source of Antarctic Bottom Water, which is a key player in the global climate system. In this study, an algorithm for estimating daily thin ice thickness is developed based on a relationship between polarization ratios (PR) of AMSR-E brightness temperatures (TBs) and thermal ice thickness. The TBs at 89 GHz and 36.5 GHz are used. The thermal ice thickness is based on heat flux calculation using ice surface temperatures derived from satellite thermal infrared images. We used cloud-free MODIS images. In the Antarctic Ocean, landfast sea-ice (fast ice), which is stationary sea ice attached to coastal features such as grounded icebergs, is formed along the coast. Antarctic coastal polynyas tend to be formed adjacent to fast ice. The AMSR-E ice thickness algorithm possibly mis-classifies fast ice as thin ice, because the PR values of thin ice and fast ice are similar. Thus, also the fast ice detection algorithm is developed. Monthly fast ice extent is detected based on microwave characteristics that the horizontally- and vertically-polarized TBs of fast ice tend to be lower than those of thin ice and are similar to those of ice sheet close to the coast. The spatial resolution of AMSR-E is about 6.25 km, and the pixel density is four times higher than that of SSM/I which has been used in previous studies. This advantage is critical for the coincident detection and monitoring of coastal polynyas and fast ice because their areal extent is fairly small (tens to a hundred kms at most). The accuracy of the created AMSR-E dataset is validated from comparisons with backscatter images acquired by ASAR on Envisat. Sea-ice production in Antarctic coastal polynyas is estimated based on heat flux calculation using the AMSR-E dataset. For the estimation, it is assumed that heat from the ocean below is negligible and that all of the heat loss to the atmosphere goes towards freezing. The sea-ice production estimated using the AMSR-E data has been improved from the SSM/I ice production because of the finer spatial resolution. First, the AMSR-E data can better resolve the high production area close to the coast. Second, false sea-ice production in the fast ice pixels mis-included by SSM/I is corrected because AMSR-E can detect fast ice that cannot be resolved by SSM/I. In fact, the total sea-ice production in each polynya by AMSR-E does not change much from the SSM/I ice production for many polynyas because these two effects of opposite direction compensate for each other. The AMSR-E dataset presented in this study would give the boundary/validation data of sea-ice production and fast ice for modeling studies.