PRODEX - Detection and quantification of thin sea ice in the Arctic

The overall objective of project ThinIce is to develop a pre-operational algorithm for the retrieval of thin sea ice thickness from thermal satellite data.

The sub-objectives of the project are:

  1. To determine whether thin sea ice thickness can be retrieved from thermal satellite data. If so:
  2. Develop a complete model and algorithm for thin ice thickness retrieval
  3. Develop a prototype processing chain that can handle large satellite data volume
  4. Evaluate the performance of the algorithm and the approach in general in a pre-operational setting
  5. Refine the approach and processing chain to make it fully suitable for operational use
  6. Transfer the prototype processing chain into an operational system within an operational environment.

There is a general lack of in situ data sets on sea ice thickness that can be used for validation of thin ice thickness data from thermal satellite data.
Data covering several years of in situ sea ice observations are available from two fjords in Svalbard: Storfjorden on the east side and Kongsfjorden on the west side of Spitsbergen. The two fjords are fundamentally different, with different precipitation scenarios, and with Kongsfjorden being a narrower fjord, influenced by Atlantic water from the West Spitsbergen Current, and Storfjorden being a wide fjord, with various hydrographic and ice-dynamic related processes. At those sites, ice thickness has been measured in the past years by drilling and electromagnetic (EM) induction using a Geonics EM31 instrument. In Storfjorden, larger scale thickness measurements have also been conducted using a helicopter-borne EM instrument (EM-bird). The in situ measurements also include snow thickness observations and temperature and salinity profiles of the sea ice. In situ and airborne data are also available from ship-based field campaigns conducted in the pack ice north of Svalbard in 2010 and 2011.

New data will be collected for this project by extending NPI’s activities in Storfjorden and Kongsfjorden to obtain best suitable dataset for validation of the ice thickness retrieval algorithm. The landfast sea ice monitoring in Inglefieldbukta, Storfjorden, will be complemented by extended Geonics EM31 surveys and repeated EM-bird flights over the entire Storfjorden sea ice cover at the beginning, the middle, and the end of the sea ice season. An ice mass balance buoy will be deployed for continuous monitoring of the sea ice thickness in Inglefieldbukta as baseline for the EM31 and EM-bird measurements. The regular monitoring in Kongsfjorden will be extended by more extensive Geonics EM31 transects and an additional EM-bird flight.

The satellite-derived ice thicknesses will be compared with the values obtained in the field both spatially and by summary statistics to assess the accuracy of the retrieval algorithm.