Geochemical data for three sedimentary records of the Frasnian–Famennian extinction and Upper Kellwasser Event, from the H-32 core (Iowa, USA), the Kowala Quarry (Poland), and Sinsin (Poland), together with time vs depth information for the revised age model of the H-32 core. Nitrogen-isotope and nitrogen content data are included for all three sites, together with biomarker information for Kowala and the H-32 core, carbon isotope data for Kowala and Sinsin, total organic carbon for the H-32 core and Sinsin, Rock Eval data from the H-32 core, phosphorus and aluminium content data, and P/Al and TOC/P ratios, from the H-32 core, and time-depth data from the H-32 core.
This data collection contains new observations of sediment color, mineral‑geochemical properties, and benthic foraminiferal stable isotopes that characterise "green bands" in Quaternary marine sediments and evaluate their value as markers of deep‑ocean ventilation changes. Parameters reported include (i) benthic foraminifera δ¹⁸O and δ¹³C (species Cibicidoides wuellerstorfi and Uvigerina spp.), (ii) down‑core colour information extracted from high‑resolution digital core photographs, expressed as green‑pixel percentages for a set of hue–saturation–value (HSV) bins, (iii) X‑ray‑fluorescence (XRF) core‑scan elemental ratios (e.g. Fe/Ti, Si/Al), (iv) pyrite counts in >63 µm coarse fractions, and (v) ancillary site‑characterisation variables such as sedimentation rate, bulk CaCO₃, organic‑carbon content, bottom‑water [O₂], and satellite‑derived export productivity.
Sediments span 0–1.2 Ma for Site U1474 and Site U1313, with depth/age tie points listed in Tables S2. Primary data derive from International Ocean Discovery Program (IODP) Sites U1474 (Southwest Indian Ocean; 31°13.00′S, 31°32.71′E, 3045 below sea level) and (North Atlantic; 41°0.068'N, 32°57.439'W, 3412 meters below sea level). The accompanying global survey (Table 3) covers observations from the first core of 2122 holes collected since the inception of the Deep Sea Drilling Program, covering all major ocean basins. Green colour banding has been linked to redox fronts associated with bottom‑water ventilation; quantifying its stratigraphic occurrence may provide an easily transferable proxy for past oxygenation. The dataset supports tests of that hypothesis and enables reuse in broader studies of sediment diagenesis, colour imaging, and benthic δ¹³C. Benthic foraminiferal δ¹⁸O and δ¹³C were measured on Cibicidoides wuellerstorfi and Uvigerina peregrina picked at 12 cm (≈ 3 kyr) intervals from the >150 µm fraction of 400 samples down to 48 m at IODP Site U1474. Analyses used a Thermo Finnigan MAT 253 IRMS coupled to a Kiel IV carbonate device at Cardiff University; long‑term external precision is ±0.05 ‰ for δ¹⁸O and ±0.021 ‰ for δ¹³C (1 σ) and results are reported relative to VPDB after calibration to internal standard BCT63 and application of standard vital‑effect offsets (+0.64 ‰ for Cibicidoides, 0 ‰ for Uvigerina). A Site U1474 δ¹⁸O stack was generated by averaging species‑specific values and tuning the composite record to the LR04 stack for age control. To assess the spatial occurrence of colour banding, we screened 2 121 split‑core photographs (first 1.5 m of piston cores) from all IODP expeditions via Texas A&M's LIMS archive, noting the depth (< 40 cm or > 40 cm) and character of oxidative colour transitions. Environmental parameters (bottom‑water O₂, export productivity, sedimentation rate, bulk CaCO₃, organic C) were extracted from published global grids and compared with banding frequency. For Sites U1474 and U1313, relic green bands were mapped with a computer‑vision workflow in Python. Split‑core images (native resolution ≈ 50 µm px⁻¹) were analysed with OpenCV's inRange filter: HSV colour space (H 0–360, S/V 0–100) was divided into 3 600 ten‑unit cubes, narrowed to 627 sediment‑relevant hues guided by Munsell chips, and the 20 most selective "green" cubes were retained. Summed depth‑series ("green‑pixel %") were smoothed and peaks delineated with SciPy's prominence‑based algorithm (prominence 9/7, distance 1/10 points for U1474/U1313), then manually quality‑controlled. Sand layers identified from core photographs and Si/Al XRF scans (Avaatech, 30 kV, 200 µA, 10 s, 2 mm beam) were masked before colour analysis. Finally, a 14‑site global benthic δ¹³C stack was compiled by standardising, 5 ka‑smoothing, and averaging published depth‑age–resolved records (Atlantic, Indian, Pacific basins) together with the new U1474 data.
Here we compiled stable carbon isotope data from 23 ODP sites in the Atlantic and Pacific, binned by marine isotope stage.Stable carbon isotope data were binned and averaged according to the marine isotope age assignments in the LR04 and the reported ages of the samples from each site. Only Cibicidoides wuellerstorfi values were used, although sometimes measured species was not reported for each sample by the original publication. Please note that some sites have data originating from multiple publications.
Calculated values for whole basin δ13C of dissolved inorganic carbon, calculated Δδ13C, and associated uncertainties from the values compiled in MPWP_benthic_isotope_data_compilation. Benthic δ13C data from all sites below 2000 meters water depth were binned by marine isotope stage as defined by the age bounds provided with the LR04 stack. Because of our focus on specific MIS events, sedimentary records with low-resolution age models were excluded. All data are derived from Cibicidoides wuellerstorfi. In datasets composed of isotope measurements of two or more benthic foraminifera species (e.g., ODP 883), only C. wuellerstorfi δ13C data were used. The mean of the binned data across the entire isotope stage from each site was then used to calculate a mean deep Pacific and mean deep Atlantic δ13CDIC estimate for all 11 marine isotope stages in the MPWP, assuming a 1:1 relationship between C. wuellerstorfi δ13C and δ13CDIC. All sites were weighed equally.
Late Pliocene (~2.6-3.4 million years ago) multiproxy paleoceanographic datasets from Ocean Drilling Program subpolar North Pacific Leg 145 sites 883 (51.12°N, 167.46°E, 2384 m water depth) and 887 (54.22°N, 148.27°W, 3631.2 m water depth). Composite splice for site 883 was created by visual correlation of the shipboard gamma ray attenuation and porosity evaluator (GRAPE) wet bulk density (WBD) measurements between 883B and 883C starting at core 9H and ending at 18H. An astronomically tuned age model for site 883 was developed by correlation of benthic foraminiferal δ18O (3.0–3.385 Ma) to the δ18O probabilistic stack and the 883 GRAPE WBD composite (2.841–3.0 Ma & 3.385–3.465 Ma) to the nearby astronomically-tuned ODP 882 carbonate weight percent record. Benthic foraminiferal stable oxygen and carbon isotope ratios were quantified with a Finnigan Mat 252 with Carbonate Kiel III autosampler. Alkenone concentrations and the UK'37 sea surface temperature index were determined by an Agilent Technologies 6890 gas chromatograph flame ionization detector (GC-FID), with an Agilent Technologies DB-1 column (60 m, 0.32 mm diameter, 0.10 mm film thickness). Calcium carbonate content was determined by acidification of bulk sediment samples with 2M hydrogen sulfide and quantification of the resulting gaseous carbon with a UIC Coulometrics CO2 Coulometer. A composite section was developed for ODP 887 by correlation of shipboard gamma ray attenuation and porosity evalvuator (GRAPE) wet bulk density between 887A and 887C. An improved age model was developed by correlation of the GRAPE composite to that of ODP 883. Stable carbon isotope data were compiled for each marine isotope stage of the mid-Piacenzian warm period from 23 IODP/ODP/DSDP core sites. These data were then binned by marine isotope stage, averaged, and grouped by ocean basin.