Towards quantitative sea ice reconstructions in the northern North Atlantic: A combined biomarker and numerical modelling approach

Organic geochemical analyses of marine surface sediments from the continental margins of East Greenland and West Spitsbergen provide for a biomarker-based estimate of recent sea ice conditions in the northern North Atlantic. By means of the sea ice proxy IP25 and phytoplankton derived biomarkers (e.g. brassicasterol and dinosterol) we reconstruct sea ice and sea surface conditions, respectively. The combination of IP25 with a phytoplankton marker (in terms of a phytoplankton marker-IP25 index; PIP25) proves highly valuable to properly interpret the sea ice proxy signal as an under- or overestimation of sea ice coverage can be circumvented. A comparison of this biomarker-based assessment of the sea ice distribution in the study area with (1) modern remote sensing data and (2) numerical modelling results reveal a good agreement between organic geochemical, satellite and modelling observations. The reasonable simulation of modern sea ice conditions by means of a regional ocean–sea ice model demonstrates the feasibility to effectively integrate the complex atmospheric and oceanic circulation features as they prevail in the study area. The good correlation between modelled sea ice parameters and the biomarker-based estimate of sea ice coverage substantiates that linking proxy and model data occurs to be a promising concept in terms of a cross-evaluation. This combinatory approach may provide a first step towards quantitative sea ice reconstructions by means of IP25. Future IP25 studies on marine surface sediments from the Arctic realm, however, are recommended to extend and validate this new attempt of using IP25 in combination with a phytoplankton marker as a quantitative measure for sea ice reconstructions.

Research Highlights
► We present a new combinatory biomarker approach for Arctic sea ice reconstructions.
► A high-resolution regional ocean–sea ice model shows realistic sea ice conditions.
► We find a considerable correlation between biomarker, model, and satellite data.
► Cross-evaluation of proxy and model data backs quantitative sea ice reconstruction.