– CliC sponsored and web-hosted this workshop. Summary by Stefan Kern
The report from the International Sea Ice Concentration and Thickness Evaluation and Inter-Comparison Workshop is now available here
Summary of the report:
The International Sea Ice Concentration and Thickness Evaluation and Inter-Comparison Workshop took place in Hamburg, Germany, on Sep. 18-19, 2014. It was initiated as part of work package WP4000 of the ESA-CCI Sea Ice ECV project (SICCI project, http://esa-cci.nersc.no). Main objectives of WP4000 are the evaluation of the sea ice concentration, computed from satellite microwave radiometry, and sea ice thickness, computed from satellite radar altimetry, prototype products generated by the SICCI project, the inter-comparison of these prototype products with independent data, and the user assessment and usage of the prototype products in, e.g., numerical models.
Over open water, accuracy of the sea ice concentration is better than about 1%, precision is better than 2% and hence matches the average retrieval error. Sea ice concentration and its retrieval error are stable over the periods (1992-2008 and 2002-2011). At high ice concentrations accuracy is still an issue and is of the magnitude 3-5% during winter. In the freezing season accuracy is degraded over thin ice and by snow conditions impacting the radiometric signature of the sea ice. During the melting season evolution in snow and ice morphology and melt ponds impact the radiometric signature of sea ice. Melt ponds are seen as open water and hence cause a reduction in sea ice concentration. The remaining (wet) snow and sea ice between the melt ponds could counterbalance this reduction. Contamination by land in the sensors’ field-of-view is an uncertainty source for sea ice area and extent calculations and poses difficulties when comparing observations with model results.
Mean values of airborne sea ice thickness estimates such as NASA Operation Ice Bridge agree within 0.1 m with SICCI project sea ice thickness but the correlation is < 0.2. Inter-comparison with in-situ sea ice thickness observations and sea ice draft yield a positive bias of SICCI project sea ice thickness of between 0.5 and 1.0 m. An estimation of the radar altimeter freeboard accuracy and precision and hence of the sea ice thickness is pending. More research work is needed to quantify the impact of snow cover and sea ice properties on radar waveform and sampling. The sensitivity of the freeboard-to-thickness conversion to input parameters snow depth, ice and snow density and their uncertainties is investigated. For radar altimetry accurate snow depth and ice density are similarly important. The lack of accurate present-day information about snow depth on sea ice hampers accurate sea ice thickness retrieval in both hemispheres. An advance in sea ice thickness evaluation would require more direct observations of sea ice freeboard, snow depth, and especially sea ice thickness. With the exception of direct drill hole measurements of the sea ice thickness all other methods also only provide a parameter, which – similar to the satellite radar altimeter freeboard – needs to be converted into sea ice thickness.