Science Service System

Summary of Proposal OCE0116

TitleThe potential of TerraSAR-X dual-polarization data for dedicated tasks in sea-ice remote sensing
Investigator Stammer, Detlef - University of Hamburg, Inst. fuer Meereskunde
Team MembersNo team members defined
Summary

Objectives:
Investigate the potential of dual-polarization TerraSAR-X measurements to

  • 1.) distinguish thin ice from other surfaces (thick ice, open water)
  • 2.) distinguish between different thin-ice surface types (bare, snow, slush, frost flowers)
  • 3.) utilize the X-Band measurements to map frost flowers.
  • 4.) determine the optimum frequency to estimate the melt-pond cover fraction on Arctic first-year and perennial ice

Method:
Utilize dual-polarization radar backscatter measurements acquired by TerraSAR-X in combination with radar backscatter data provided as simultaneous as possible at other frequencies (L- and C-Band) by, e.g., ALOS-PALSAR and Envisat-ASAR:

  • to improve the discrimination of thin ice and between different thin-ice surface properties (particularly frost flowers),
  • to develop a method to derive the thickness of thin ice using these measurements,
  • to identify melt ponds and to estimate their areal fraction,
  • by investigating the co-polarization ratio at the different considered frequencies bands. By assuming that the co-polarization ratio of sea ice at higher frequencies (X- and Ku-Band) is relatively uniform, i.e. close to zero decibel, except for very thin ice, while the one at lower frequencies (L- to C-Band) is changing with young ice thickness, the thin ice can be identified and its thickness derived. It will be investigated whether and by which amount the co-polarization ratio at L-/C-Band is influenced by frost flowers.
    During summer no young ice is present so that any change in the co-polarization ratio observed at either of the bands (L- to Ku) could be attributed to a change in the melt-pond fraction provided that the ice concentration is close to 100% and does not change. It will be investigated, which frequency would give the best result with regard to accuracy, the minimum observable melt-pond cover fraction (change), influence by changes in ice type and degree of deformation, and change in surface wind speed.

    One backbone for these goals are helicopter-borne multi-frequency, multi-polarization scatterometer measurements at L-, C-, X-, and Ku-Band with the HELISCAT instrument [Kern and Gade, 2004] carried out during field campaigns as simultaneous as possible to overpasses of relevant satellite sensors. Together with in-situ measurements of snow and ice properties, and air-borne VIS-reflectance / IR-temperature measurements, HELISCAT data will be used to develop methods to obtain the thin-ice thickness, identify (and map) frost flowers, and identify (and map) melt ponds and derive the sea-ice albedo. It will be investigated, to which extent X-Band data are a necessary component of these methods which in a next step will be modified to be applied to satellite data. Synchronous HELISCAT and TerraSAR-X data are needed for this modification and to evaluate the resulting methods.

    The second backbone are preparatory multi-sensor studies involving multi-frequency satellite active microwave data available at the date for two key regions. These studies shall provide information about most suitable incidence angles (probably above 40) and frequency combinations to accomplish above-mentioned goals. These studies are a preparation for the field campaigns and the modification of the methods from using airborne to satellite data.

    Data Requirements:
    Ideal would be dual-polarization ScanSAR!! This seems not to be possible. We request in total 90 SC and 90 SM, GEC, for Ross Sea, Beaufort Sea, and wider Fram Strait region, start: 03/2007, end: 10/2008.

    Funding:
    will be requested from DFG within SPP 1158

    Deliverables:
    Methods to identify and map frost flowers, derive the thin-ice thickness, and to identify melt-ponds and their coverage using a multi-frequency approach; final report

    Back to list of proposals

    DLR 2004-2016