Synergy of radar altimetry and L-Band passive microwave data to improve Arctic sea ice thickness information

Abstract:

Satellite remote sensing of sea ice thickness can be achieved by different sensors that are based on independent physical retrieval methods. Consequently, these methods differ in sensitivity for certain sea ice types and thickness ranges as well as in the spatial and temporal resolution of the geophysical product. However, estimating Arctic sea ice volume requires retrieval methods that are able to obtain thickness information across the entire sea ice thickness range. Currently, this requirement is only met by laser altimeter and radar data, such as from ICESat or CryoSat-2. Other retrieval strategies that are realized by the evaluation of surface emissivity in microwave frequencies are restricted to the thin first-year sea ice thickness range, where the upper thickness threshold is defined by the wavelength among other factors. It is however the thin thickness ranges, where altimeter-based sea ice thickness estimates show the highest relative uncertainty, since the method requires a detectable elevation of ice floes above the local sea surface.

We show first result of a data fusion of sea ice thickness fields obtained from CryoSat-2 freeboard and L-Band surface emission measurements of the SMOS satellite. We implemented the data fusion with an optimal interpolation between the two datasets on a weekly basis, where the parameters of the optimal interpolation are based on the sensor characteristics and the uncertainties of the sea ice thickness data fields from both sources. The differences of both data products are used to discuss the potential preferential sampling in both datasets. In addition, we use dedicated airborne and shipborne validation data in the marginal ice zone in the Barents Sea to assess the characteristics the individual and merged sea ice thickness information.