The Arctic Ocean is under profound transformation. Observations and model predictions show dramatic decline in sea ice extent and volume. Despite its importance, our understanding of the pacing of Arctic sea ice retreat is incomplete largely due to a paucity of observations. The launch of the Soil Moisture and Ocean Salinity (SMOS) mission, in 2009, marked the dawn of a new type of space-based microwave observations. Although the mission was originally conceived for hydrological and oceanographic studies [1,2], SMOS is also making inroads in the cryospheric sciences. SMOS carries an L-band (1.4 GHz, or 21-cm wavelength), passive interferometric radiometer (the so-called MIRAS) that measures the electromagnetic radiation emitted by the Earth’s surface, at about 50 km spatial resolution, full polarization, continuous multi-angle viewing, large wide swath (1200-km), and with a 3-day revisit time at the equator, but more frequently at the poles. A significant difference of the L-band microwave radiometers with respect to higher frequency radiometers, such as SSMI/AMSR-E/AMSR-2, is that the former can also “see through the ice.” That is because ice is more transparent (i.e., optically thinner) at 1.4 GHz than at higher frequencies (19-89 GHz). In radiometric terms, the brightness temperature measured by an L-band antenna radiometer does not correspond to the emissivity of the topmost surface layer but of a larger range of deeper layers within the ice (about 60 cm, depending on ice conditions). Thanks to that increased penetration in the medium, L-band radiometers can provide, for the first time, thin ice thickness from space [3, 4]. A novel radiometric method to determine sea ice concentration (SIC) is presented. The method exploits the distinctive radiative properties of sea ice and seawater when observed at low microwave frequencies and from a range of incidence angles, to discern both media. The Bayesian-based Maximum Likelihood Estimation (MLE) approach is used to retrieve SIC. The advantage of this approach with respect to the classical linear inversion is that the former takes into account the uncertainty of the tie-point measured data in addition to the mean value, while the latter only uses a mean value of the tie-point data. When thin ice is present, the SMOS algorithm underestimates SIC due to the low opacity of the ice at this frequency. However, using a synergistic approach with data from other satellite sensors, it is possible to obtain accurate thin ice thickness estimations with the Bayesian-based method. Despite its lower spatial resolution relative to SSMI or AMSR-E, SMOS-derived SIC products are little affected by the atmosphere and the snow (almost transparent at L-band). This new dataset can contribute to complement ongoing monitoring efforts in the Arctic Cryosphere

2016 European Space Agency (ESA) Living Planet Symposium, 9-13 May 2016, Prague, Czech Republic

Peer Reviewed