Coccolithophore assemblage data of sediment cores 162-984C, 162-980, GeoB11035-1, GeoB5546-2, M35003-4 and GeoB3104-1 from the North Atlantic Ocean spanning at least the past 24 ka. The assemblage data cover major parts of the latitudinal and thermal gradients in the North Atlantic Ocean and were used to assess changes in biodiversity and assemblage composition during the past 24 ka, i.e., from the Last Glacial Maximum to the current warm period covering the last deglaciation.Prior to coccolithophore counting, the sediment samples were prepared using a combined dilution/filtering technique as described by Andruleit (1996, doi:10.2307/1485964). Between 50 and 150 mg of dry bulk sediment was brought into suspension using demineralized water buffered with ammonia (pH ~ 8.5-9). The suspension was ultrasonicated for 15 to 30 s, split (split factor = 100) using a rotary splitter and filtered onto polycarbonate membrane filters (0.4 µm pore size) using a vacuum pump. The samples have been dried at 40°C for 24 h before ~1x1 cm filter aliquots were mounted on an aluminium stub that was prepared with electroconductive, self-adhesive, carbon-based discs (PLANO Leit tabs). The mounted samples were sputtered with gold/palladium in a Polaron SC7640 Sputter Coater.Coccolith counts (N) were made manually using a Zeiss DSM 940A scanning electron microscope at 3,000x or 5,000x magnification. In some instances, counts were made on SEM images at different magnifications. Where possible a minimum of 300 coccoliths were counted. In some samples fewer coccoliths were counted due to low abundances. Sample mass (m), magnification, split factor (k) and the area of the scanned transect (A) and the filter (F) are given for all samples to allow calculation of the concentration of coccoliths per gram of dry bulk sediment, which was calculated using: (FN)/(kA*m).Coccolithophore taxonomy follows Young et al. (2003, doi:10.58998/jnr2297). Species not reported in the data are assumed to be absent (that is, zero abundance). Counts of unidentified and reworked specimens are provided. Where appropriate, the lumping of species and the mapping of subspecies and variants are indicated.The chronology of the time series presented here are based on revised age models included in the PALMOD 130k marine palaeoclimate data synthesis V1.1 (Jonkers et al., 2020, doi:10.5194/essd-12-1053-2020) or revised using the same approach as in Jonkers et al. (2020, doi:10.5194/essd-12-1053-2020). All age models are based on radiocarbon ages and mainly benthic foraminifera δ18O isotope data calibrated to regional benthic foraminifera δ18O stacks (Lisiecki and Stern, 2016, doi:10.1002/2016pa003002).
