Olivier Marchal

Apparent long-term cooling of the sea surface in the northeast Atlantic and Mediterranean during the Holocene.

Reconstructions of upper ocean temperature (T) during the Holocene (10–0 ka B.P.) were established using the alkenone method from seven, high accumulation sediment cores raised from the northeast Atlantic and the Mediterranean Sea (361N–751N). All these paleo-T records document an apparent long-term cooling during the last 10 kyr. In records with indication of a constant trend, the apparent cooling ranges from � 0.27 to � 0.151C kyr � 1 . Records with indication of time-variable trend show peak-to-peak amplitudes in apparent temperatures of 1.2–2.91C. A principal component analysis shows that there is one factor which accounts for a very large fraction (67%) of the total variance in the biomarker paleo-T records and which dominates these records over other potential secondary influences. Two possible contributions are (1) a widespread surface cooling, which may be associated with the transition fromthe Hypsithermal interval ( B9–5.7 ka B.P.) to the Neoglaciation (B5.7–0 ka B.P.); and (2) a change in the seasonal timing and/or duration of the growth period of alkenone producers (prymnesiophyte algae). The first contribution is consistent with many climate proxy records from the northeast Atlantic area and with climate model simulations including Milankovitch forcing. The second contribution is consistent with the divergence between biomarker and summer faunal paleo-T fromearly to late Holocene observed in two cores. Further work is necessary, and in particular the apparent discordance between biomarker and faunal T records for the relative stable Holocene period must be understood, to better constrain the climatic and ecological contributions to the apparent cooling observed in the former records. r 2002 Elsevier Science Ltd. All rights reserved.

Feedback mechanisms and sensitivities of ocean carbon uptake under global warming

Global warming simulations are performed with a coupled climate model of reduced complexity to investigate global warming—marine carbon cycle feedbacks. The model is forced by emissions of CO2 and other greenhouse agents from scenarios recently developed by the Intergovernmental Panel on Climate Change and by CO2 stabilization profiles. The uptake of atmospheric CO2 by the ocean is reduced between 7 to 10% by year 2100 compared to simulations without global warming. The reduction is of similar size in the Southern Ocean and in low-latitude regions (32.5°S-32.5°N) until 2100, whereas low-latitude regions dominate on longer time scales. In the North Atlantic the CO2 uptake is enhanced, unless the Atlantic thermohaline circulation completely collapses. At high latitudes, biologically mediated changes enhance ocean CO2 uptake, whereas in low-latitude regions the situation is reversed. Different implementations of the marine biosphere yield a range of 5 to 16% for the total reduction in oceanic CO2 uptake until year 2100. Modeled oceanic O2 inventories are significantly reduced in global warming simulations. This suggests that the terrestrial carbon sink deduced from atmospheric O2/N2 observations is potentially overestimated if the oceanic loss of O2 to the atmosphere is not considered.

Ocean thermohaline circulation and sedimentary 231Pa/230Th ratio

Holocene sediments from the Atlantic are characterized by 231Pa/230Th ratios below the production ratio of the two radionuclides in the water column (0.093), whereas Holocene sediments from the Southern Ocean have 231Pa/230Th > 0.093. This pattern of 231Pa deficit and excess was ascribed to southward 231Pa export from the Atlantic by the Atlantic thermohaline circulation (THC) as Pa is scavenged less efficiently by marine particles and more effectively transported by the THC than Th. The same pattern is observed in deposits of the Last Glacial Maximum (LGM), which led to a previous contention that the THC strength did not vary markedly through the last glacial termination. Here we embed a description of trace metal scavenging into a zonally averaged, circulation-biogeochemistry ocean model to explore the sensitivity of 231Pa/230Th in Atlantic and Southern Ocean sediments to THC changes. Our results show that the production of biogenic opal (which, unlike other marine particles, poorly fractionates Th and Pa) in the Southern Ocean water column determines the spatial pattern of the sensitivity. Also, 231Pa/230Th increases in the North Atlantic but changes little in the South Atlantic and decreases in the Southern Ocean as THC is reduced. The mean 231Pa/230Th of the whole Atlantic is therefore less sensitive to THC changes than the mean 231Pa/230Th of the North Atlantic. The current uncertainties in Atlantic mean 231Pa/230Th are too large to rule out a twofold reduction of the THC at the LGM. However, the increase in North Atlantic mean 231Pa/230Th simulated in response to a twofold THC reduction is larger than the observed change in the North Atlantic mean 231Pa/230Th from the LGM to Holocene. Comparing this change with the modeled sensitivity of North Atlantic 231Pa/230Th to THC variations indicates that the THC at the LGM could not have been reduced by >30% of its present strength. Experiments of transient THC changes indicate that high-resolution 231Pa/230Th records from North Atlantic sediments could also document thermohaline oscillations on century-to-millennial timescales.

VARIATIONS OF OXYGEN-MINIMUM AND PRIMARY PRODUCTIVITY RECORDED IN SEDIMENTS OF THE ARABIAN SEA

Abstract Two deep-sea sediment cores from the northeastern and the southeastern Arabian Sea were studied in order to reconstruct the palaeoenvironments of the past glacial cycles. Core 136KL was recovered from the high-productivity area off Pakistan within the modern oxygen-minimum zone (OMZ). By contrast, modern primary productivity at the site of MD900963 close to Maldives is moderate and bottom waters are today well oxygenated. For both cores, we reconstructed the changes in palaeoproductivity using a set of biomarkers (alkenones, dinosterol and brassicasterol); the main result is that primary productivity is enhanced during glacial stages and lowered during interstadials. The proxies associated with productivity show a 23 kyr cyclicity corresponding to the precession-related insolation cycle. Palaeoredox conditions were studied in both cores using a new organic geochemical parameter (C35/C31-n-alkane ratio) developed by analysing surface sediments from a transect across the OMZ off Pakistan. The value of this ratio in core 136KL shows many variations during the last 65 kyr, indicating that the OMZ was not stable during this time: it disappeared completely during Heinrich- and the Younger Dryas events, pointing to a connection between global oceanic circulation and the stability of the OMZ. The C35/C31 ratio determined in sediments of core MD900963 shows that bottom waters remained rather well oxygenated over the last 330 kyr, which is confirmed by comparison with authigenic metal concentrations in the same sediments. A zonally averaged, circulation–biogeochemical ocean model was used to explore how the intermediate Indian Ocean responds to a freshwater flux anomaly at the surface of the North Atlantic. As suggested by the geochemical time series, both the abundance of Southern Ocean Water and the oxygen concentration are significantly increased in response to this freshwater perturbation.

How strong is the Harvardton‐Bear Constraint?

We compare the sensitivity of the partial pressure of CO 2 in the warm surface ocean and atmosphere to the influence of the oceans cold water outcrops in a wide spectrum of models. While in simple box models the cold ocean dominates, in three-dimensional ocean general circulation models, this influence is considerably smaller, suggesting that exchange processes between the warm and cold regime in the real ocean are extremely important in determining the distribution of chemical properties.

Impact of oceanic reorganizations on the ocean carbon cycle and atmospheric carbon dioxide content

A zonally averaged, circulation-biogeochemical ocean model is used to explore how the distribution of PO4 and δ13C in the major basins and the atmospheric pCO2 respond to rapid changes in the thermohaline circulation (THC). Different evolutions of the Atlantic THC are simulated by applying surface freshwater pulses typical, for example, of Heinrich events and the last deglaciation. In the model, when the THC completely collapses, PO4 increases (>0.5 mmol m−3) and δ13C decreases ( 1‰) because of a longer residence time of waters at the surface. These surface anomalies are propagated to the bottom North Atlantic when the THC resumes. When the thermohaline overturning is only partly reduced and at shallower depths, changes in bottom waters are accompanied by a PO4 decrease and δ13C increase at intermediate levels in the mid-latitude Atlantic. This results in enhanced vertical gradients of these properties consistent with chemical and isotopic reconstructions for the last glacial maximum. Finally, the atmospheric pCO2 increases during the cold period in the North Atlantic when the THC is reduced with an amplitude (7–30 µatm) and timescale (∼10² to 1–2 × 10³ yr) depending on the intensity of the THC change. This is qualitatively consistent with recent data from an Antarctic ice core documenting a pCO2 increase during the Younger Dryas and after Heinrich events 4 and 5.

On the Abyssal circulation in the glacial Atlantic

An inverse method is used to evaluate the information contained in sediment data for the Atlantic basin during the Last Glacial Maximum (defined here as the time interval 18–21 kyr before present). The data being considered are an updated compilation of the isotopic ratios 18 O/ 16 O( � 18 O) and 13 C/ 12 C( � 13 C) of fossil shells of benthic foraminifera (bottom-dwelling organisms). First, an estimate of the abyssal circulation in the modern Atlantic is obtained, which is consistent with (i) climatologies of temperature and salinity of the World Ocean Circulation Experiment, (ii) observational estimates of volume transport at specific locations, and (iii) the statements of a finite-difference geostrophic model. Second, estimates of water properties (� 18 O of equilibrium calcite or � 18 Oc and � 13 C of dissolved inorganic carbon or � 13 CDIC) derived from sediment data are combined with this circulation estimate to test their consistency with the modern flow. It is found that more than approximately 80% of water property estimates (� 18 Oc or � 13 CDIC) are compatible with the modern flow given their uncertainties. The consistency of glacial � 13 CDIC estimates with the modern flow could be rejected after two assumptions are made: (i) the uncertainty in these estimates is �0.1‰ (this uncertainty includes errors in sediment core chronology and oceanic representativity of benthic � 13 C, which alone appears better than this value on average); and (ii) � 13 CDIC in the glacial deep Atlantic was dominated by a balance between water advection and organic C remineralization. Measurements of � 13 C on benthic foraminifera are clearly useful, but the current uncertainties in the distribution and budget of � 13 CDIC in the glacial Atlantic must be reduced to increase the power of the test.

Application of an inverse method to interpret 231Pa/230Th observations from marine sediments

constraints. Then sediment 231 Pa/ 230 Th data for each time interval are combined with an advection‐ scavenging model in order to determine their (in)consistency with the modern circulation estimate. We find that the majority of sediment 231 Pa/ 230 Th data for the Holocene, LGM, or H1 can be brought into consistency with the modern circulation if plausible assumptions are made about the large‐scale distribution of 231 Pa and about model uncertainties. Moreover, the adjustments in the data needed to reach compatibility with a hypothetical state of no flow (no advection) are positively biased for each time interval, suggesting that the 231 Pa/ 230 Th data (including that for H1) are more consistent with a persistence of some circulation than with no circulation. Our study does not imply that earlier claims of a circulation change during the LGM or H1 are inaccurate, but that these claims cannot be given a rigorous basis given the current uncertainties involved in the analysis of the 231 Pa/ 230 Th data.

Can Paleoceanographic Tracers Constrain Meridional Circulation Rates

Abstract The ability of paleoceanographic tracers to constrain rates of transport is examined using an inverse method to combine idealized observations with a geostrophic model. Considered are the spatial distribution, accuracy, and types of tracers required to constrain changes in meridional transport within an idealized single-hemisphere basin. Measurements of density and radioactive tracers each act to constrain rates of transport. Conservative tracers, while not of themselves able to inform regarding rates of transport, improve constraints when coupled with density or radioactive observations. It is found that the tracer data would require an accuracy one order of magnitude better than is presently available for paleo-observations to conclusively rule out factor-of-2 changes in meridional transport, even when assumed available over the entire model domain. When data are available only at the margins and bottom of the model, radiocarbon is unable to constrain transport while density remains effective o...

Atmospheric radiocarbon during the Younger Dryas: production, ventilation, or both?

A new reconstruction of past atmospheric v 14 C( v 14 Catm) based on Polish lake varved sediments has suggested that previous v 14 Catm values (e.g. from the Cariaco basin record) for the beginning of the Younger Dryas cold event (YD) are overestimates and that the v 14 Catm rise at the YD onset could only be due to changes in atmospheric 14 C production (PCˇ14). This result would have profound climatic implications, for the YD is a paradigm example of abrupt climate change which is usually thought to have been triggered by a reduction in the northward heat flux by the Atlantic thermohaline circulation. Here we examine results from a large number of simulations (300) based on a zonally averaged ocean circulation model, to constrain the effect on v 14 Catm of PCˇ14 changes during the YD as reconstructed from a Greenland ice core record of 10 Be flux. Our results suggest that the scatter in the lake data set is too large to exclude the probable change in deep ocean ventilation at the onset of the YD. By contrast, the model fit to the higher v 14 Catm levels throughout the YD detected in the marine record is generally better when a substantial decrease in deep ocean ventilation is simulated. The early v 14 Catm drawdown that initiated during the first half of the YD, however, could entirely be due to production changes. If this was the case, the drawdown would not reflect an increasing formation of North Atlantic Intermediate Water or Southern Ocean water, as previously suggested. The rapid v 14 Catm rise at the YD onset documented in the marine record, however, remains unexplained. fl 2001 Elsevier Science B.V. All rights reserved.