Day 4

Detailed paper information

Back to list

Paper title Detecting Arctic sea ice openings with Cryosat-2 InSAR observations as a basis for improved ocean surface circulation monitoring
Authors
  1. Tadea Veng DTU Space, Technical University of Denmark Speaker
  2. Felix L. Müller Technical University of Munich (DGFI-TUM)
  3. Denise Dettmering Technical University of Munich
  4. Florian Seitz Deutsches Geodätisches Forschungsinstitut (DGFI-TUM), Technical University of Munich
  5. Ole Baltazar Andersen DTU Space
Form of presentation Poster
Topics
  • A8. Ocean
    • A8.07 Oceanographic Change of the Arctic Ocean From Space
Abstract text Sea level observations from satellite altimetry in the Arctic Ocean are severely limited due to the presence of sea ice. To determine sea surface heights and enable studies of the ocean surface circulation, it is necessary to first detect openings in the sea ice cover (leads and polynyas) where the ocean surface is exposed. This is of particular interest in the coastal areas of the Arctic, where glaciers calve into the Arctic Ocean. The increasing freshwater influx in the last years leads to changes in the sea level and the thermohaline circulation.

The ESA Explorer mission Cryosat-2 was launched in 2010, aiming at the monitoring of the cryosphere. The satellite works in three different acquisition modes. One of these modes is the interferometric SAR (InSAR) mode. The radar returns (called waveforms) of this mode are characterized by a higher temporal resolution, which allows a more reliable detection of leads and polynyas in coastal areas. An unsupervised classification approach based on Machine Learning is implemented for Cryosat-2 InSAR waveforms. The classification approach utilizes differences in scattering properties from sea ice, open ocean, and calm enclosed ocean. By defining quantitative parameters from the waveform shape, the waveforms are grouped by comparing the similarity of the parameters without the necessity of pre-classified data. The classification performance is validated against optical images of spatiotemporally overlapping aircraft overflights. An algorithm is implemented to automatically detect leads from the optical images while minimizing the time difference between altimetry and optical observations.

The implementation of an unsupervised detection of open water in the Arctic Ocean environment is part of the recently launched AROCCIE project (ARctic Ocean surface circulation in a Changing Climate and its possible Impacts on Europe). The aim of the project is to combine satellite altimetry with numerical ocean modelling to determine changes in Arctic Ocean surface circulation from 1995 to present. AROCCIE will use the classification of InSAR data to create a more comprehensive dataset of sea surface heights for further analysis of ocean circulation changes in the vicinity of the Arctic's rugged coastlines.