Description: Arctic sea ice is a critical component of the climate system, known to influence ocean circulation, earth's albedo, and ocean-atmosphere heat and gas exchange. Current developments in the use of IP25 (a sea ice proxy with 25 carbon atoms only synthesized by Arctic sea ice diatoms) have proven it to be a suitable proxy for palaeo-sea ice reconstructions over hundreds of thousands to even million years timescales. In the NE-Baffin Bay, off NW-Greenland, Melville Bugt is a climate-sensitive region characterized by strong seasonal sea ice variability and strong melt-water discharge from the Greenland Ice Sheet (GIS). Here, we present a centennial-scale resolution Holocene sea ice record, based on IP25 and open-water phytoplankton biomarkers (brassicasterol, dinosterol and HBI III) using core GeoB19927-3 (73°35.26' N, 58°05.66' W). Seasonal to ice-edge conditions near the core site is documented for most of the Holocene period with some significant variability. A cold interval, in the lower-most part, characterized by extensive sea ice cover and very low local productivity is succeeded by an interval (~9.4-8.5 ka BP)) with reduced sea ice cover, enhanced GIS spring melting, and strong influence of the West Greenland Current (WGC) in the earliest part of the record. From ~8.5 untill ~7.8 ka BP, a cooling event is recorded by ice algae and phytoplankton biomarkers. They indicate an enhanced sea ice cover, possibly related to the opening of Nares Strait, which may have led to an increased influx of Polar Water into NE-Baffin Bay. The interval between ~7.8 and ~3.0 ka BP is characterized by generally reduced sea ice cover with millennial-scale variability of the (late winter/early spring) ice-edge limit, increased open-water conditions (polynya-type), and a dominant WGC carrying warm waters at least as far as the Melville Bugt area. During the last ~3.0 ka BP, our biomarker records do not reflect the late Holocene 'Neoglacial cooling' observed elsewhere in the Northern Hemisphere, possibly due to the persistent influence of the WGC and interactions with the adjacent fjords. Peaks in HBI III at about ~2.1 and ~1.3 ka BP, interpreted as persistent ice-edge situations might, correlate with the Roman Warm Period (RWP) and Medieval Climate Anomaly (MCA), respectively, in-phase with the North Atlantic Oscillation (NAO) mode. When integrated with marine and terrestrial records from other circum-Baffin Bay areas (Disko Bay, the Canadian Arctic, the Labrador Sea), the Melville Bugt biomarker records point to close ties with high Arctic and Northern Hemispheric climate conditions, driven by solar and oceanic circulation forcings.

Description: This dataset comprises previously published and new data on the composition of sediment cores from the Arctic Ocean. New data are supplied for cores PS51/038-4, PS72/396-5 and PS72/396-3: For PS51/038-4 and PS72/396-5, we provide data on 230Th excess, 238U, 234U/238U together with bulk Ca, Mn and S concentrations. For PS72/396-3, we provide data on planktonic foraminifera, their oxygen and carbon isotopic composition, allochtonous forminifera fractions and the abundance of Cibicides lobatulus subsp. grossa. Thorium and uranium isotopes were measured on a sector-field ICP-MS (Element2) via isotope dilution; elemental concentrations on an ICP-OES iCAP; foraminfera abundances after freeze-drying and sieving of a mean sample size of 87 g; and stable isotopes on a Finnigan MAT 251 isotope ratio gas mass spectrometer directly coupled to a Kiel I automatic carbonate preparation device; all at the Alfred Wegener Institute. 230Thex was calculated by subtracting the activity of 234U from 230Th. Existing data comprise 230Thex for PS2185 and PS2200 (recalculated from Strobl et al. 1998), 10Be for the same cores (Strobl 1998 and Spielhagen et al. 2004), and microfossil data for the same cores (Spielhagen et al. 2004). In addition, we provide 230Thex data For HLY-0503-11 and HLY0503-12 from Not and Hillaire Marcel (2010) and PS87/030-1 from Hillaire Marcel et al. (2017). We also include previously published original and now recalculated U-series data (as described in the publication) together with previously published data for 10Be, magnetic susceptibility, sand content, stable isotopes, planktonic foraminifera abundance, CaCO3 content, TOC content, and age information for PS1533-3; 230Thexcess data manually digitized from a published paper version for PS1235-2 (Eisenhauer et al. 1990); a copy of previously published U-series data for GIK23065-3 (Paetsch et al. 1991).