Sea ice remote sensing

Due to the remoteness and hostile nature of polar regions, satellite observations are particularly useful for observing the Arctic and the state of its sea ice cover. Sea ice remote sensing can contribute to both climate research and practical applications, such as shipping. We use passive microwave measurements to estimate two important sea ice parameters from space: ice thickness (of thin ice) and snow thickness (on thick ice). 

Thin sea ice with a thickness of less than 50 cm plays a fundamental role for the energy exchange between the atmosphere and the ocean in the polar regions, as well as for the dynamics of sea ice. Thus, a continuous large-scale observation of thin ice areas in the polar regions would provide an important contribution to monitoring our climate system. The European Space Agency's (ESA) Soil Moisture and Ocean Salinity (SMOS) satellite mission was launched in 2009, and, for the first time, allows for globally measuring the Earth's radiation at a low microwave frequency of 1.4 GHz (about 21 cm wave length). Although SMOS was originally designed to provide global measurements of soil moisture and ocean salinity (as its name says), it has also been used to retrieve information on sea ice. In contrast to the higher frequencies at which previous passive microwave sensors have been operated, at 1.4 GHz, radiation measured above sea ice originates not only from the surface of the ice but also from deeper parts of the ice and, if the ice is not too thick, from the water underneath the ice. As a result, SMOS measurements provide information on the thickness of ice, up to thicknesses of about 50 cm to 1 m. In our group, an innovative retrieval algorithm has been developed and is used to operationally produce daily SMOS-derived thin sea ice thickness maps of the Arctic. These maps are available via icdc.zmaw.de.

Monthly mean sea ice thickness derived from SMOS brightness temperatures in the Arctic for November 2012. Note that the maximum retrievable ice thickness with SMOS is about 1 m. Thus, here these values indicate ice thicknesses of at least 1 m.

Snow is an important component of the Arctic climate system. Due to its high albedo and its low thermal conductivity, snow on sea ice largely determines the surface energy balance in polar regions. Snow also modifies the signal observed from satellites, and information on the snow cover is required for the estimation of ice thickness from altimetry measurements (i.e. measurements of the satellite's height over the Earth's surface), for example. Until now, the only continuous large-scale measurements of snow thickness have been provided by passive microwave satellites operating at 19 and 37 GHz frequencies. However, these satellite measurements are not suitable for the retrieval of snow thickness on multi-year ice in the Arctic because at these frequencies the measured signals from multi-year ice and from water are not distinguishable. Multi-year ice is ice that has survived at least one summer melt season in the Arctic. It is usually less saline and more permeable than first-year ice, i.e. sea ice newly formed during the current ice growth season.

Measurements from the European Space Agency's (ESA) Soil Moisture and Ocean Salinity (SMOS) satellite mission have been used to infer ice thickness up to about 50 cm to 1 m. Compared to previously operated microwave satellites, SMOS measures at a very low frequency of 1.4 GHz. We found that over thick sea ice (where we cannot infer the ice thickness with SMOS) there is a relationship between the measured 1.4 GHz radiation and the thickness of the snow layer on top of the ice. A first comparison of SMOS-derived snow thicknesses and snow thicknesses measured during a flight campaign in the Arctic in spring 2012 shows a reasonable agreement (see Figure below).

Our aims are to further validate the new SMOS-based approach for an Arctic-wide retrieval of snow thickness, to investigate the retrieval's uncertainties regarding the ice conditions, and to possibly improve the retrieval model. Finally, we plan to provide Arctic-wide snow thickness maps (including uncertainties) for a longer time sequence, which can be used as input to sea ice models, e.g. for short- and medium term forecasting.

Snow thickness as retrieved from SMOS satellite data (background) and as measured during NASA's "Operation IceBridge" flight campaign (circles) in March/April 2012