Sea ice thickness estimates in the Weddell sea using laser altimetry and microwave radiometry

Abstract

This thesis deals with the estimation and changes of sea-ice thickness in the area of Weddell Sea, Antarctica. Sea-ice thickness estimates were derived from satellite sensors, which posses the unique ability to obtain data from high altitudes above the Earth. The introduction and background to this work are presented in Chapter 1 and the target area in Chapter 2. The satellite sensors used for the estimation of sea-ice thickness employ laser altimetry and microwave radiometry. Some theoretical basics of remote sensing in polar areas are described in Chapter 3. In order to obtain sea-ice thickness, snow thickness from the Advanced Microwave Scanning Radiometer - EOS (AMSR-E) is subtracted from freeboard height. The freeboard height is the total height of the snow cover plus the height of sea ice above the sea surface which can be inferred from data gathered from the Geoscience Laser Altimeter System (GLAS). The AMSR-E and GLAS satellite sensors are described in Chapter 4.This thesis presents some methods previously utilised for the estimation of sea-ice thickness. A well established method for the estimation of sea ice thickness has been modified and applied to data for the Weddell Sea. This thesis is concerned with the process of sea-ice thickness estimation and focuses above all on the choice of geoid, application of sea-ice concentration, averaging of snow data, inclusion of slush layers, and uncertainties in magnitudes regarding the calculation of sea-ice thickness. These sensitivity studies are included in Chapter 5.The resulting data are discussed in Chapter 6: Freeboard minus snow reveals that up to 50% of the Weddell Sea ice may consist of negative freeboard. A slush model is applied to the calculation of sea-ice thickness in the areas with negative freeboard heights. The sea-ice thicknesses in the Weddell Sea are estimated for 14 austral seasons between 2003 and 2008. The map resolution is 50 km x 50 km, and mean sea-ice thicknesses range from 1.64 to 2.28 m in autumn 2003 - 2008, from a thickness of 1.39 to 1.78 m in winter 2004 - 2006, and from a thickness of 2.49 to 3.11 m in spring 2003 - 2007. Repeating seasonal cycles are reflected in the maps with a thinning of the sea ice from spring to autumn. Gratifyingly, the histograms of sea-ice thickness distributions produced by these analyses are strikingly similar to those in the most recent publication of Yi et al. (2011).

This thesis deals with the estimation and changes of sea-ice thickness in the area of Weddell Sea, Antarctica. Sea-ice thickness estimates were derived from satellite sensors, which posses the unique ability to obtain data from high altitudes above the Earth. The introduction and background to this work are presented in Chapter 1 and the target area in Chapter 2. The satellite sensors used for the estimation of sea-ice thickness employ laser altimetry and microwave radiometry. Some theoretical basics of remote sensing in polar areas are described in Chapter 3. In order to obtain sea-ice thickness, snow thickness from the Advanced Microwave Scanning Radiometer - EOS (AMSR-E) is subtracted from freeboard height. The freeboard height is the total height of the snow cover plus the height of sea ice above the sea surface which can be inferred from data gathered from the Geoscience Laser Altimeter System (GLAS). The AMSR-E and GLAS satellite sensors are described in Chapter 4.This thesis presents some methods previously utilised for the estimation of sea-ice thickness. A well established method for the estimation of sea ice thickness has been modified and applied to data for the Weddell Sea. This thesis is concerned with the process of sea-ice thickness estimation and focuses above all on the choice of geoid, application of sea-ice concentration, averaging of snow data, inclusion of slush layers, and uncertainties in magnitudes regarding the calculation of sea-ice thickness. These sensitivity studies are included in Chapter 5.The resulting data are discussed in Chapter 6: Freeboard minus snow reveals that up to 50% of the Weddell Sea ice may consist of negative freeboard. A slush model is applied to the calculation of sea-ice thickness in the areas with negative freeboard heights. The sea-ice thicknesses in the Weddell Sea are estimated for 14 austral seasons between 2003 and 2008. The map resolution is 50 km x 50 km, and mean sea-ice thicknesses range from 1.64 to 2.28 m in autumn 2003 - 2008, from a thickness of 1.39 to 1.78 m in winter 2004 - 2006, and from a thickness of 2.49 to 3.11 m in spring 2003 - 2007. Repeating seasonal cycles are reflected in the maps with a thinning of the sea ice from spring to autumn. Gratifyingly, the histograms of sea-ice thickness distributions produced by these analyses are strikingly similar to those in the most recent publication of Yi et al. (2011).