ALBERT

Description: Benthic foraminiferal d18O and Mg/Ca of sediment cores off tropical NW Africa are used to study the properties of Atlantic central waters during the Last Glacial Maximum (LGM) and Heinrich Stadial 1 (HS1). We combined our core top data with published results to develop a new Mg/Ca-temperature calibration for Planulina ariminensis, which shows a Mg/Ca-temperature sensitivity of 0.19 mmol/mol per °C. Estimates of the LGM and HS1 thermocline temperatures are comparable to the present-day values between 200 and 400 m water depth, but were 1.2-1.5°C warmer at 550-570 m depth. The HS1 thermocline waters (200-570 m depth) did not show any warming relative to the LGM. This is in contrast to previous climate model studies, which concluded that tropical Atlantic thermocline waters warmed significantly when Atlantic meridional overturning circulation was reduced. However, our results suggest that thermocline temperatures of the northeastern tropical Atlantic show no pronounced sensitivity to changes in the thermohaline circulation during glacial periods. In contrast, we find a significant increase in thermocline-water salinity during the LGM (200-550 m depth) and HS1 (200-400 m depth) with respect to the present-day, which we relate to changes in the wind-driven circulation. We infer that the LGM thermocline (200-550 m depth) and the HS1 upper thermocline (200-400 m depth) in the northeastern tropical Atlantic was ventilated by surface waters from the North Atlantic rather than the southern-sourced waters. This suggests that the frontal zone between the modern South Atlantic and North Atlantic Central Waters was probably shifted southward during the LGM and HS1.

Description: Sedimentary specimens of the planktonic foraminifera Globorotalia inflata can provide much needed information on subsurface conditions of past oceans. However, interpretation of its geochemical signal is complicated by possible effects of cryptic diversity and encrustation. Here we address these issues using plankton tow and sediment samples from the western South Atlantic, where the two genotypes of G. inflata meet at the Brazil-Malvinas Confluence Zone. The δ18O and δ13C of encrusted specimens from both genotypes from a core within the confluence zone are indistinguishable. However, we do find a large influence of encrustation on δ18O and Mg/Ca. Whereas crust Mg/Ca ratios are at all locations lower than lamellar calcite, the crust effect on δ18O is less consistent in space. Plankton tows show that encrusted specimens occur at any depth and that even close to the surface crust Mg/Ca ratios are lower than in lamellar calcite. This is inconsistent with formation of the crust at lower temperature at greater depth. Instead we suggest that the difference between the crust and lamellar calcite Mg/Ca ratio is temperature-independent and due to the presence of high Mg/Ca bands only in the lamellar calcite. The variable crust effect on δ18O is more difficult to explain, but the higher incidence of crust free specimens in warmer waters and the observation that a crust effect is clearest in the confluence zone, hint at the possibility that the difference reflects advective mixing of specimens from warmer and colder areas, rather than vertical migration.

Description: Modern variability in upwelling off southern Indonesia is strongly controlled by the Australian-Indonesian monsoon and the El Niño-Southern Oscillation, but multi-decadal to centennial-scale variations are less clear. We present high-resolution records of upper water column temperature, thermal gradient and relative abundances of mixed layer- and thermocline-dwelling planktonic foraminiferal species off southern Indonesia for the past two millennia that we use as proxies for upwelling variability. We find that upwelling was generally strong during the Little Ice Age (LIA) and weak during the Medieval Warm Period (MWP) and the Roman Warm Period (RWP). Upwelling is significantly anti-correlated to East Asian summer monsoonal rainfall and the zonal equatorial Pacific temperature gradient. We suggest that changes in the background state of the tropical Pacific may have substantially contributed to the centennial-scale upwelling trends observed in our records. Our results implicate the prevalence of an El Niño-like mean state during the LIA and a La Niña-like mean state during the MWP and the RWP.

Description: The Southern Westerly Winds (SWW) exert a crucial influence over the world ocean and climate. Nevertheless, a comprehensive understanding of the Holocene temporal and spatial evolution of the SWW remains a significant challenge due to the sparsity of high-resolution marine archives and appropriate SWW proxies. Here, we present a north-south transect of high-resolution planktonic foraminiferal oxygen isotope records from the western South Atlantic. Our proxy records reveal Holocene migrations of the Brazil- Malvinas Confluence (BMC), a highly sensitive feature for changes in the position and strength of the northern portion of the SWW. Through the tight coupling of the BMC position to the large-scale wind field, the records allow a quantitative reconstruction of Holocene latitudinal displacements of the SWW across the South Atlantic. Our data reveal a gradual poleward movement of the SWW by about 1-1.5° from the early to the mid-Holocene. Afterwards variability in the SWW is dominated by millennial-scale displacements in the order of 1° in latitude with no recognizable longer-term trend. These findings are confronted with results from a state-of-the-art transient Holocene climate simulation using a comprehensive coupled atmosphere-ocean general circulation model. Proxy-inferred and modeled SWW shifts compare qualitatively, but the model underestimates both orbitally forced multi-millennial and internal millennial SWW variability by almost an order of magnitude. The underestimated natural variability implies a substantial uncertainty in model projections of future SWW shifts.

Description: We present a new coccolithophore productivity reconstruction spanning the last 300 ka in core GeoB12613-1 retrieved from the western tropical Indian Ocean (IO), an area that mainly derives its warm and oligotrophic surface waters from the eastern IO. Application of a calibrated assemblage-based productivity index indicates a reduction in estimated primary productivity (EPP) from 300 ka to the present, with reconstructed EPP values ranging from 91 to 246 g C/m2/yr. Coccolithophore assemblages and coccolith fraction Sr/Ca indicate three main phases of productivity change, with major changes at 160 and 46 ka. The productivity and water-column stratification records show both dominant precession and obliquity periodicities, which appear to control the paleoproductivity in the study area over the last two glacial-interglacial cycles. Shallowing of the thermocline due to strengthening of the trade winds in response to insolation maxima resulted to peaks in EPP. Comparison with the eastern IO productivity and stratification coccolithophore data reveals good correspondence with our records, indicating a strong tropical Pacific influence in our study area. Both of these records show high productivity from 300 ka to 160 ka, interpreted to be due to stronger Walker Circulation while the declining productivity from 160 ka to the present day is a consequence of its weakening intensity.

Description: The core descriptions (chapter 7) summarize the most important results of the analysis of each sediment core following procedures applied during ODP/IODP expeditions. All cores were opened, described, and color-scanned. In the core descriptions the first column displays the lithological data that are based on visual analysis of the core and are supplemented by information from binocular and smear slide analyses. The sediment classification largely follows ODP/IODP convention. Lithological names consist of a principal name based on composition, degree of lithification, and/or texture as determined from visual description and microscopic observations. In the structure column the intensity of bioturbation together with individual or special features (turbidites, volcanic ash layers, plant debris, shell fragments, etc.) is shown. The hue and chroma attributes of color were determined by comparison with the Munsell soil color charts and are given in the color column in the Munsell notation. A GretagMacbethTM Spectrolino spectrophotometer was used to measure percent reflectance values of sediment color at 36 wavelength channels over the visible light range (380-730 nm) on all of the cores. The digital reflectance data of the spectrophotometer readings were routinely obtained from the surface (measured in 1 cm steps) of the split cores (archive half). The Spectrolino is equipped with a measuring aperture with folding mechanism allowing an exact positioning on the split core and is connected to a portable computer. The data are directly displayed within the software package Excel and can be controlled simultaneously. From all the color measurements, for each core the red/blue ratio (700 nm/450 nm) and the lightness are shown together with the visual core description. The reflectance of individual wavelengths is often significantly affected by the presence of minor amounts of oxyhydroxides or sulphides. To eliminate these effects, we used the red/blue ratio and lightness.

Description: The advection of relatively fresh Java Sea water through the Sunda Strait is presently responsible for the low-salinity "tongue" in the eastern tropical Indian Ocean with salinities as low as 32 per mil. The evolution of the hydrologic conditions in the eastern tropical Indian Ocean since the last glacial period, when the Sunda shelf was exposed and any advection via the Sunda Strait was cutoff, and the degree to which these conditions were affected by the Sunda Strait opening are not known. Here we have analyzed two sediment cores (GeoB 10042-1 and GeoB 10043-3) collected from the eastern tropical Indian Ocean off the Sunda Strait that cover the past ~40,000 years. We investigate the magnitude of terrigenous supply, sea surface temperature (SST), and seawater d18O (d18Osw) changes related to the sea level-driven opening of the Sunda Strait. Our new spliced records off the Sunda Strait show that during the last glacial, average SST was cooler and d18Osw was higher than elsewhere in the eastern tropical Indian Ocean. Seawater d18O decreased ~0.5 per mil after the opening of the Sunda Strait at ~10 kyr B.P. accompanied by an SST increase of 1.7°C. We suggest that fresher sea surface conditions have persisted ever since due to a continuous transport of low-salinity Java Sea water into the eastern tropical Indian Ocean via the Sunda Strait that additionally increased marine productivity through the concomitant increase in terrigenous supply.

Description: Measurement were conducted using a NewWave UP193 solid-state laser coupled to a Thermo‐Finnigan Element2 HR-ICP-MS at the Department of Geosciences, University of Bremen. Isotope abundances of 25Mg, 27Al, 43Ca, 55Mn, 57Fe and 87Sr in foraminifera shells were determined at an irradiance of 1.3 GW cm-2, a laser pulse rate of 5 Hz and a spot size of 50–100 μm. Plasma power was 1200 W, and Helium (0.7 L min-1) and Argon (0.9 L min-1) were used as sample and make-up gases, respectively. All isotopes were analysed at low resolution with five samples in a 20% mass window and a total dwell time of 25 ms per isotope. Blanks were measured for 30 s prior to ablation and every analysis was followed by a wash out time of at least 60 s to avoid cross-contamination between samples. A glass reference material (NIST610) was measured before every five ablation spots as an external calibration standard using the values of Jochum et al. (2011). To assess data quality, a pressed pellet of MACS-3 carbonate standard powder (n=4) and the USGS reference material BHVO‐2G (n=18) were analysed as control standards along with the samples. External precision is better than 3.9%, and accuracy as determined by comparison of our reference material data with the GeoReM database (as of July 2020) is better than 3.4% for all elements except for Mg (7.7%). The 43Ca isotope was used as an internal standard. The files contain raw and despiked element counts, calibration factors and calibrated element/calcium ratios.

Description: Research expedition M140 took place from August 11th, 2017 (Mindelo, Cabo Verde) to September 5th, 2017 (Las Palmas, Spain). CTD data for 16 stations with a total of 37 individual casts along the cruise track were recorded using a Sea & Sun Technology CTD90M (SN 979) down to depths of 700m. The CTD was equipped with the following sensors: Temperature sensor Pt100 model 1509 (Thermal Developments International), Conductivity sensor 7-pole platinum coated electrode cell in quartz glass (Sea & Sun Technology), Seapoint Chlorophyll Fluorometer (Seapoint) and dissolved oxygen sensor (DO522M18, Clark type, OxyGuard). Additionally data for two stations were also recorded with Seabird Electronics (SBE) 9 plus CTD (SN 979) mounted in a SBE water sampler rosette. The data files contain the data for temperature, salinity, density, chlorophyll a concentration and dissolved oxygen concentration; as raw data and processed and flagged according to the recommendations for real-time data processing of EuroGOOS and GTSPP, as well as the outlier detection method CoTeDe (https://github.com/castelao/CoTeDe). Biogeographic regions are determined according to Spalding et al. (2012). TEOS-10 unit conversions have been performed with the GSW Oceanographic Toolbox (McDougall and Barker, 2011).