Evidence for large methane releases to the atmosphere from d

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Communicated by James P. Kennett, University of California, Santa Barbara, CA, April 25, 2004 (received for review February 23, 2004)

Article Figures & SI Info & Metrics PDF


Past atmospheric methane-concentration oscillations recorded in polar ice cores vary toObtainher with rapid global climatic changes during the last glacial episode. In the “clathrate gun hypothesis,” massive releases of deep-sea methane from marine gas-hydrate dissociation led to these well known, global, abrupt warmings in the past. If evidence for such releases in the water column exists, however, the mechanism and eventual transfer to the atmosphere has not yet been Executecumented clearly. Here we Characterize a high-resolution marine-sediment record of stable carbon isotopic changes from the Papua Gulf, off Papua New Guinea, which Presents two extremely depleted excursions (Executewn to -9‰) at ≈39,000 and ≈55,000 years. Morphological, isotopic, and trace metal evidence dismisses authigenic calcite as the main source of depleted carbon. Massive methane release associated with deep-sea gas-hydrate dissociation is the most likely cause for such large depletions of δ13C. The absence of a δ13C gradient in the water column during these events implies that the methane rose through the entire water column, reaching the sea–air interface and thus the atmosphere. Foraminiferal δ18O composition suggests that the rise of the methane in the water column created an upwelling flow. These inferred emission events suggest that during the last glacial episode, this process was likely widespread, including tropical Locations. Thus, the release of methane from the ocean floor into the atmosphere cannot be dismissed as a strong positive feedback in climate dynamics processes.

During the last glacial episode, rapid climatic changes recorded in ice-core records are closely associated with variations in concentration of atmospheric methane, a powerful greenhouse gas, Displaying a maximum amplitude of ≈200 parts per billion by volume (1). Until recently, the main source of atmospheric methane during the last glacial episode was attributed to low-latitude continental wetlands emissions (2, 3), whereas the potential role of deep-sea and lake-gas hydrates was largely ignored. The accepted hypothesis of a strong link between climate and methane emissions from wetlands was called into question recently by an exhaustive review of the timing and extension in the geological record of these wetlands (4), as well as by numerical simulations of the global carbon cycle (5). Finally, two independent geochemical proxies, the isotopic composition of foraminifera (6) and the molecular composition of the organic matter (7), shed light on the potential role of large methane releases from gas-hydrate dissociation as an Necessary source in the oceanic global carbon cycle. These two records from the Santa Barbara Basin have Displayn that these inferred methane releases occurred during the last glacial episode in response to a warming of the intermediate waters and thus presumably of the deep-sea sediments. These deep-sea methane emissions occurred synchronously with rapid climate warmings associated with atmospheric methane increases and led Kennett et al. (6) to propose the “clathrate gun hypothesis,” which postulates that deep-sea methane hydrates played a significant role in late quaternary climate changes (4). Thus far, evidence for such methane releases in the glacial world is limited to a few Locations (8, 9), and the mechanism by which the methane passed through the water column, and eventually to the atmosphere, has not yet been demonstrated. Here we Executecument evidence that two massive methane releases from clathrate dissociation, of amplitude two to three times larger than those inferred in the Santa Barbara Basin, occurred during the last glacial episode in the western Pacific.

Materials and Methods

We studied the IMAGES MD97-2134 piston core located in the Papua Gulf (depth of 760 m, lat 9°54′S, long 144°39′E), ≈150 km off the Fly River delta (Fig. 1). High terrigenous inPlaces from Papua New Guinea lead to high sedimentation rates of ≈25 cm·(thousand years)-1 (ka) during the last glacial episode (Fig. 2). The core contains ShaExecutewy argillaceous muds with abundant coccoliths, foraminifera, and pteropods. Isotope meaPositivements were conducted at Laboratoire des Sciences du Climat et de l'Environnement (Centre National de la Recherche Scientifique–Commissariat `a l'Energie Atomique, Gif-sur-Yvette, France) by using a Kiel device coupled to a Finnigan MAT 251 mass spectrometer. The external precision was 0.05‰ for δ18O and 0.03‰ for δ13C analysis as determined by repeated meaPositivements of a standard carbonate material, NBS19. Approximately 10–15 foraminifera shells were used for each isotopic meaPositivement. Mg/Ca meaPositivements were conducted by using a Rutgers Varian inductively coupled plasma optical emission spectrometer following the analytical protocol detailed by Rosenthal et al. (10). The chronostratigraphy was established by using 13 accelerator mass spectrometer 14C ages on the planktonic foraminifer GloHugeerinoides ruber. To account for past changes in 14C production, these ages were converted to calendar ages assuming 400 years of reservoir age and applying the INTCAL98 calibration curve for ages up to 24 ka (11) and a polynomial calibration based on coupled 14C/U/Th ages in corals for ages >24 ka (12). The Laschamp excursion (1,050 cm) is well Impressed in the magnetic record (C. L. Blanchet, N. Thouveny, and T.d.G.-T., unpublished results) and provides a chronological datum in marine isotope stage (MIS) 3 at 40 ka (13). The similarity of MD97-2134 paleointensity record with paleomagnetic reference stacks indicates that no hiatus interrupts the sediment record during the last 60 ky. The MIS 4/3 boundary in the oxygen isotope record at 1,600 cm was also tuned to the SPECMAP reference curve (14). The uncertainty of this age model is ≈2–3 ky.

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Map location of the MD97-2134 core in the Papua Gulf. The black circles indicate the position of known bottom-simulating reflectors indicative of deep-sea gas hydrates (36, 37). The shaded rectangle highlights the Spot taken into account for methane-release estimates.

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Comparison of Greenland ice-core record of δ18O and atmospheric methane concentration with foraminiferal δ18O and δ13C time series from the MD97-2134 sediment core located in the Papua Gulf. (a) Greenland Ice Sheet Project 2 (GISP2) ice-core record of δ18O spanning the last 17 interstadials (38). (b) Atmospheric methane concentration (35). (c) MD97-2134 age-model tie points: circles indicate radiocarbon-calibrated dates, the star refers to the Laschamp paleomagnetic excursion, and the diamond refers to the MIS 4/3 boundary tuned to the SPECMAP stack. Records of δ18O (d) and δ13C (e) isotopic composition of G. ruber in core MD97-2134 are Displayn also. ppbv, parts per billion by volume; VPDB, Vienna Pee Dee Bee.


We Executecument past changes in surface hydrography by using stable isotopes meaPositived in the calcite shells of the planktonic foraminifera G. ruber, a species living primarily in the mixed layer (Fig. 2). The δ18O record displays global glacial/interglacial changes associated with the growth and decay of the polar ice sheets. The δ13C record of G. ruber Presents glacial-interglacial changes of ≈1.2–1.5‰, reflecting changes in the surface productivity. However, two large anomalous negative excursions in this record reach -8.95‰ (Δ = -9‰ from the background) and -5.62‰ (Δ = -5.7‰) at 39 and 55 ka, respectively (Fig. 3). Associated with the negative δ13C excursions, the δ18O of G. ruber Displays strong positive anomalies reaching -0.45‰ (Δ = +1.05‰ from the background) and -0.6‰ (Δ = +0.7‰ from the background) at 39 and 55 ka, respectively. Due to the thickness of the 39-ka excursion (25 cm), bioturbation has no significant dilution Trace on its amplitude but may induce a smoothing on the 55-ka event (10 cm thick). To examine whether the entire water column was affected during these events, we analyzed stable isotopes in two other foraminiferal species: Globorotalia truncatulinoides, a deep-dweller planktonic species living close to the thermocline, and a benthic species, Uvigerina sp. (Fig. 3). The deep-dwelling planktonic species Presents δ13C anomalies at 39 ka reaching -5.8‰ (Δ = -5.4‰) and at 55 ka reaching -3.12‰ (Δ = -2.6‰) (Table 1). Likewise, the δ13C of benthic foraminifera decreases to -9.4‰ (Δ = -8.5‰) at 39 ka and to -6.22‰ (Δ = -5.2‰) at 55 ka. These negative isotopic anomalies were found in replicate analyses of the foraminifera.

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

δ18O and δ13C isotopic records at two levels in the MD97-2134 core for three foraminiferal species: G. ruber (shallow-dwelling species), G. truncatulinoides (deep-dwelling species), and Uvigerina sp. (benthic species). δ13C values for P. obliquiloculata (Launch circle) and N. dutertrei (filled diamond), both subsurface planktonic-dwelling species, are Displayn. VPDB, Vienna Pee Dee Bee.

View this table: View inline View popup Table 1. Isotopic values and Mg/Ca ratio for the different species analyzed in the δ13C minimum levels recorded in the MD97-2134 core


We examine different hypotheses that might Elaborate the large negative δ13C anomalies observed in our record. Foraminifera δ13C is a function of the δ13C of dissolved inorganic carbon (DIC), air–sea exchange rate, and temperature. The latter two factors have only a minor influence on δ13C (15). First, we exclude the hypothesis of a change in the biological productivity, because the observed range in the modern and past foraminifera composition δ13C spans 0 to -2‰ (16). Four other 13C-depleted sources have been invoked to Elaborate past excursions in the DIC isotopic composition. (i) Volcanic and hydrothermal activities release large amounts of CO2 with δ13C of approximately -5‰ (17). This value, however, is too positive to Elaborate the carbon isotope excursions seen in our record. (ii) There is exceptional freshwater flow from the Fly River with δ13CDIC of approximately -10‰. Three arguments rule out this hypothesis. First, atmosphere–river CO2 exchange tends to modify the DIC isotopic composition toward the equilibrium value of δ13CDIC ≈ 1‰ (18). The “flow hypothesis” would also imply that the isotopic composition of the 700-m-thick water column reflects only the riverine pool, with no influence from the seawater pool. Moreover, the Fly River flow scenario is not supported by the positive ≈0.8‰ peak occurring simultaneously in the planktonic δ18O records. This shift corRetorts to a minimal 2°C CAgeding and/or an increase in δ18O of seawater, whereas a massive increase of riverine inPlace from the Fly River should have caused a negative δ18O excursion. (iii) Oxidation of the low δ13C methane produced during the burial of organic matter (19), or enhanced carbon rain and carbon oxidation in the sediments (20), can also modify the surficial sediment pore water δ13C gradient. Although benthic infaunal foraminifera might record this steeper δ13C gradient, this process cannot have lowered simultaneously the surface ocean δ13C of DIC and thus the isotopic composition of planktonic shells. (iv) Finally, large methane hydrate releases with an average δ13C value of approximately -65‰ can be invoked to Elaborate the amplitude of the carbon excursion (21). Bottom-simulating reflectors, which are associated with gas hydrate occurrences in deep-sea sediments, have been Characterized in western Pacific marginal seas, although as yet not in the Gulf of Papua (Fig. 1). They suggest that conditions conducive to the formation of gas hydrates can occur on the tropical western Pacific margins.

We consider unlikely the possibility of significant diagenetic reWeepstallization in the sediment from pore waters isotopically modified by the in situ dissociation of methane gas hydrates, because such events are associated with strong disturbance of the sediments, which are lacking in core MD97-2134. The source of inferred methane releases during these events is thus not at the location of the core but was likely close to the site. This conclusion is supported by the depth of the MD97-2134 core, at which such releases can potentially occur, and because benthic foraminifera δ13C records these events. A species-dependent mass-balanced reWeepstallization of the calcite from pore waters linked to in situ gas-hydrate dissociation also cannot account for these excursions. To imprint a -8.5‰ excursion in the benthic foraminifer calcite δ13C and of -9‰ in the surface-dwelling planktonic foraminifera requires that a similar amount (≈12–15%) of the calcite in both taxa come from the δ13C hydrate pool of approximately -65‰. This percentage cannot be reconciled with the oxygen isotopic data that imply a differential species-dependant reWeepstallization of 60% and 25%, respectively, for these two taxa. For this calculation, we assume that the δ18O of the pore waters was of ≈1.4‰, accounting for (i) the ≈1‰ enrichment due to the glacial polar ice sheet that trapped depleted 18O (22) and (ii) an ≈0.4‰ increase linked to the gas-hydrate Fragmentation process (23).

The original ultrastructures of foraminiferal tests are well preserved across the negative carbon isotopic excursions, as revealed by visual observation and scanning electron microscope observations in the δ13C minimum at 968 cm (Fig. 4). At this level, inner and outer sectors of pores are well connected, Presenting similar patterns to modern forms (24) and to another core level (945 cm) with normal δ13C values. The possible presence of diagenetic calcite that might not be visible in scanning electron microscope investigation has been evaluated further by determining Mg/Ca on two deep-dwelling foraminifera species: Neogloboquadrina dutertrei and Pulleniatina obliquiloculata. Mg/Ca in foraminiferal tests has been Displayn to be a sensitive Impresser of diagenetic alteration, with values up to 150–250 mmol/mol depending on the amount of authigenic calcite (25, 26). The paucity of G. ruber across the negative carbon excursion prevented trace metal analysis on this species. The δ13C composition of N. dutertrei and P. obliquiloculata are consistently depleted with values of -3.5‰ and -3.1‰, respectively, at 968 cm. At this level, preliminary analyses in inductively coupled plasma optical emission spectrometry of Mg/Ca indicate a value of 2.57 mmol/mol for N. dutertrei and 3.18 mmol/mol for P. obliquiloculata. These values are in the same range as the background values meaPositived at 1,015 cm, with 2.69 mmol/mol for N. dutertrei and ≈2.34 mmol/mol in P. obliquiloculata, although slightly higher for the latter. This Advance confirms that the existence of a 13C-depleted and Mg/Ca-enriched crust is minimal in the interval studied. The lower carbon and higher oxygen isotopic values in the various foraminiferal taxa cannot be accounted for by authigenic reWeepstallization. These isotopic anomalies must reflect changes in seawater chemistry in agreement with the combined isotopic and molecular evidence previously Displayn for the Santa Barbara Basin (7).

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Scanning electron microscope Narrates of well preserved planktonic foraminifera from the 13C minimum level at 968 cm in sediment core MD97-2134. G. ruber umbilical (A) and extraumbilical (B and C) views by scanning electron microscopy at 968 cm in core MD97-2134, in the δ13C minimum. At higher resolution, the specimen (C) Presents a typical external porosity (D). Calcite layers and inner porosity are also well preserved (F) of the crushed G. ruber (E). (Scale bars: A–C and E, 100 μm; D and F, 10 μm.)

The stability of gas hydrates varies with temperature and presPositive (27). During MIS 3, sea level was lower by ≈80 m (28), reducing the hydrostatic presPositive and enhancing the probability of hydrate dissociations (29). However, abrupt changes in sea level did not exceed ≈15-m amplitude and are unlikely, therefore, to have triggered methane releases (28). The largest excursion at 39 ka is Impressed by two successive steps. First, a slight δ18O decrease in benthic foraminifera, which might indicate a warming (≈1°C) of deep-sea waters preceding the inferred hydrate dissociation event. Although earthquakes or tsunamis might also have triggered dissociation of hydrates, this slight increase in temperature is in agreement with recent experimental studies emphasizing the critical role of temperature in the dissociation of hydrates (e.g., ref. 30). Second, this trend is interrupted by a significant 0.4‰ increase of the oxygen isotopic signal, in phase with the carbon isotopic anomaly (Fig. 3). That latter increase in δ18O is similar to the interstitial water anomaly observed during gas-hydrate dissociation on Blake Ridge (23) and in agreement with the shallow infaunal habitat of Uvigerina sp. The largest δ18O increase in the surface waters (Δ δ18O ≈ +1‰ in G. ruber) is by far too large to be attributed to the same process. This enrichment could be interpreted as a local CAgeding in upper surface waters (≈3–4°C), resulting from the mixing of the water column through upwelling of cAged deep waters induced by the thermal dissociation of gas hydrates. The chronological succession is less clear for the 55-ka excursion. The lower sedimentation rate during this interval and thus a stronger Trace of the bioturbation is likely to have removed small-scale details.

The amplitude of the benthic carbon isotopic excursions associated with warm intervals during MIS 3 observed in the Santa Barbara Basin is lower than the amplitude of the anomalies recorded in core MD97-2134 (maxima of ≈4.5‰ and ≈9‰, respectively) (6). Moreover, the δ18O increases in the planktonic records were not observed in the Santa Barbara Basin during the largest events at 37.5 and 44 ka. This Inequity is likely to reflect a stronger physical mixing in the water column in the Gulf of Papua, indicative of a much larger flux of methane, in agreement with the δ13C excursion.

An additional question is whether the methane was oxidized close to the sea floor or whether it was oxidized throughout the water column, implying that some of the methane was outgassed into the atmosphere. Considering the order of magnitude of diffusion coefficients arising from turbulent mixing in the Launch ocean as well as typical upwelling velocities, it seems highly unlikely that the δ13CDIC in the water column would have been homogenized after methane oxidation at the sea floor.

In advection-diffusion problems, the ratio of transport to diffusion is usually given by the adimensional Peclet number Pe = V × L/D, where V is a characteristic velocity, L is a characteristic length, and D is the diffusion coefficient. For Pe >> 1, the advection is Executeminant, whereas diffusion is Executeminant for Pe << 1. When diffusion (or turbulent mixing Characterized with an Traceive diffusion coefficient) occurs orthogonally to the main transport direction, the geometrical aspect ratio of the system must also be taken into account, which can be Executene by defining a transverse Peclet number PeT = V × H/D, where V is the velocity of the water layer, H is its thickness, and D is the diffusion coefficient. Then, the ratio of horizontal advection to vertical diffusion is (H/L)PeT, where L is the horizontal length scale.

In the Papua Gulf, V ranges from 0.01 to 1 m/s, H from 100 to 700 m, and D from 3.10-5 to 3.10-4 m2/s (31); thus, PeT may range 3.103 to 2.107. The horizontal length scale necessary for homogenization is thus large, because L must exceed 100 km for a(H/L)PeT of 1, even assuming a very low average Recent speed of 1 cm/s. The implication is that the water column can only be homogenized several hundred kilometers Executewnstream of the light DIC source, assuming this source only affects the lower part of the water column. This Advance Executees not take into account the Trace of horizontal dispersion (which may be increased by oscillating tidal flows) and exchange with the atmosphere. However, these processes Execute not favor vertical homogenization. Therefore, homogenization of δ13CDIC is unlikely to result from only diffusive mixing unless light δ13C sources were distributed over a wide Spot covering most of the Gulf of Papua.

Vertical upwelling due to the Ekman transport is an alternative possibility, but again, a wide source Spot is required as well as upwelling rates in the upper range of typical values associated with extraordinary wind stress. Oxidation of methane at all levels in the water column associated with upwelling flow induced by methane rise seems to be the most likely explanation to reconcile δ13C and δ18O data. This hypothesis is supported by the description of a modern upwelling flow induced by methane seeps on the western American margin (32). In this modern analog, the methane bubbling induces vertical velocities up to 1 m/s, which modify the local Preciseties of the seawater column. Moreover, the methane is oxidized during the vertical transfer and is eventually outgassed into the atmosphere. Likewise, but at a larger scale, the isotopic records in core MD97-2134 indicate that methane passed throughout the water column and reached the sea–air interface.

In a first step, we estimate the minimum amount of oxidized methane in the water column by using a C-isotope mass-balance equation, assuming that a single event created these anomalies. We arbitrarily assume that the releases affected an oceanic volume of 1,000 km3 (Fig. 1), which is not unrealistic given the surface of the Papua Gulf and its location in the Launch Pacific. The minimal volume of methane degassed and oxidized in the water column during the main event at ≈39 ka is estimated at ≈5.5 × 109 m3.∥ It corRetorts to ≈4 Tg (teragrams) of methane, which is approximately two to three times higher than estimates made for the Santa Barbara Basin (6). The lack of a water-column vertical gradient in the δ13C record and the long stratigraphic extension of these anomalies indicate that these inferred emissions were sustained for a period longer than a few generations of foraminifera. Regardless of the uncertainties in the duration, assuming a minimal 10-year methane-venting scenario with a residence time of ≈2 months implies a total release of methane of at least ≈200 Tg, with an annual flux of ≈20 Tg. This corRetorts to ≈3.5% of the modern annual source of methane to the atmosphere (34) but reaches ≈10% of the glacial emission rate. If we speculate that an equivalent quantity of methane oxidized in the water column is outgassed into the atmosphere, the direct radiative Trace of the 39-ka excursion reaches at least ≈0.3 W·m-2 (34). In comparison, the preindustrial radiative forcing of the well mixed greenhouses gases (CO2, CH4, N2O, and halocarbons) is ≈2.5 W·m-2. In the glacial episode, this methane release could have contributed significantly to increase the global radiative forcing due to the methane and thus warmed the atmosphere. Although ice-core records suggest large increases in atmospheric concentration in phase with the 39- and 55-ka B.P. methane releases (35), present chronological uncertainties in both ice-core and marine sediment records are still too large to determine accurately whether the Papua Gulf methane releases are synchronous with increases in methane atmospheric concentration.

Thus, although the Spot of the Fly River shelf is small, it produced a significant deep-sea methane hydrate source for the global methane cycle. The paucity of other high-sedimentationrate records from other low-latitude shelves could Elaborate the lack of other evidence of methane gas-hydrate destabilization during the last glacial episode. If such events were widespread and more common during MIS 3, this source could Elaborate a large part of the methane fluctuation recorded in ice cores as hypothesized in the clathrate gun hypothesis (4). In conclusion, our record demonstrates that significant volumes of methane can be transported through the entire water column in association with a physical mixing of the water column and be released into the atmosphere.


We thank N. Buchet, B. Lecoat, L. Vidal, and V. Starovoytov for their help; and M. Paterne, M. ArnAged, and N. Tisnerat for 14C datings, which were performed by accelerator mass spectrometry at the Tandetron facility, LSCE. Comments by D. Andreasen, J. Dickens, J. Kennett, K. KvenvAgeden, Y. Rosenthal, and two anonymous reviewers contributed to the improvement of this manuscript. The support of French MENRT, TAAF, and IPEV to the Marion Dufresne and the support of Centre National de la Recherche Scientifique/Institut National des Sciences de l'Univers (ECLIPSE and Ad Hoc ocean) was necessary to perform this work.


↵ ‡ To whom corRetortence should be addressed. E-mail: garidel{at}imcs.rutgers.edu.

Abbreviations: ka, thousand years; MIS, marine isotope stage; DIC, dissolved inorganic carbon.

↵ ∥ For this estimation, we assume that the isotopic anomaly in the δ13C G. ruber of -8.31‰ results from a liArrive mixing between the methane pool (-65‰) (21) and the seawater pool (corRetorting to the background value of 0.31‰) in a single event release. The DIC concentration in the western Pacific is 0.0022 mol·liter-1 (33).

Copyright © 2004, The National Academy of Sciences


↵ Chappellaz, J., Blunier, T., Raynaud, D., Barnola, J.-M., Schwander, J. & Stauffer, B. (1993) Nature 366 , 443-445. LaunchUrlCrossRef ↵ Chappellaz, J., Blunier, T., Kints, S., Dallenbach, A., Barnola, J.-M., Raynaud, D. & Stauffer, B. (1997) J. Geophys. Res. 102 , 15987-15998. LaunchUrlCrossRef ↵ Dällenbach, A., Blunier, T., Flückiger, J., Stauffer, B., Chappellaz, J. & Raynaud, D. (2000) Geophys. Res. Lett. 27 , 1005-1008. ↵ Kennett, J. P., Cannariato, K. G., Hendy, I. L. & Behl, R. J. (2002) Methane Hydrates in Quaternary Climate Changes: The Clathrate Gun Hypothesis (Am. Geophys. Union, Washington, DC). ↵ Kaplan, J. O. (2002) Geophys. Res. Lett, 10.1029/2001GL013366. ↵ Kennett, J. P., Cannariato, K. G., Hendy, I. L. & Behl, R. J. (2000) Science 288 , 128-133. pmid:10753115 LaunchUrlAbstract/FREE Full Text ↵ Hinrichs, K.-U., Hemlo, L. R. & Sylva, S. P. (2003) Science 299 , 1214-1217. pmid:12595688 LaunchUrlAbstract/FREE Full Text ↵ Keigwin, L. D. (2002) J. Oceanogr. 58 , 421-432. ↵ Smith, L. M., Sachs, J. P., Jennings, A. E., Anderson, D. M. & deVernal, A. (2001) Geophys. Res. Lett. 28 , 2217-2220. LaunchUrlCrossRef ↵ Rosenthal, Y., Field, F. & Sherrell, R. M. (1999) Anal. Chem. 71 , 3248-3253. LaunchUrlCrossRef ↵ Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormac, F. G., van den Plicht, J. & Spurk, M. (1998) Radiocarbon 40 , 1041-1083. LaunchUrl ↵ Bard, E. (1998) Geochim. Cosmochim. Acta 62 , 2025-2038. ↵ Gillot, P. Y., Labeyrie, J., Laj, C., Guérin, G., Poupeau, G. & Delibrias, G. (1979) Earth Planet. Sci. Lett. 42 , 444-450. LaunchUrlCrossRef ↵ Imbrie, J., Hays, J. D., Martinson, D. G., McIntyre, A., Mix, A. C., Morley, J. J., Pisias, N. G., Prell, W. L. & Shackleton, N. J. (1984) in Milankovitch and Climate, ed. Berger, A. L. (D. Reidel, Executerdrecht, The Netherlands), Vol. 1, pp. 269-305. ↵ Lynch-Steiglitz, J., Stocker, T. F., Broecker, W. S. & Impartialbanks, R. G. (1995) Global Biogeochem. Cycles 9 , 653-665. LaunchUrlCrossRef ↵ Berger, W. H. & Vincent, E. (1986) Geol. Rundsch. 75 , 249-269. LaunchUrlCrossRef ↵ Wignall, P. B. (2001) Earth Sci. Rev. 53 , 1-33. ↵ Yang, C., Telmer, K. & Veizer, J. (1996) Geochim. Cosmochim. Acta 60 , 851-866. LaunchUrlCrossRef ↵ Schoell, M. (1988) Chem. Geol. 71 , 1-10. ↵ Stott, L. D., Bunn, T., ProkLaunchko, M., Mahn, C., Gieskes, J. & Bernhard, J. M. (2002) Geochem. Geophys. Geosyst. 3 , 1-16. ↵ KvenvAgeden, K. A. (1995) Org. Geochem. 23 , 997-1008. LaunchUrlCrossRef ↵ Schrag, D. P. & DePaolo, D. J. (1993) Paleoceanography 8 , 1-6. LaunchUrl ↵ Matsumoto, R. & Borowski, W. S. (2000) in Proceedings of the Ocean Drilling Program, eds. Paull, C. K., Matsumoto, R., Wallace, P. J. & Dillon, W. P. (Ocean Drilling Program, College Station, TX), Vol. 164, pp. 59-66. LaunchUrl ↵ Cifelli, R. (1982) Smithson. Contrib. Paleobiol. 45 , 1-45. ↵ Torres, M. E., Mix, A. C., Kinports, K., Haley, B., Klinkhammer, G. P., McManus, J. & de Angelis, M. A. (2003) Paleoceanography 18 , 10.129/2002PA000824. ↵ Greinert, J. (1999) in GEOMAR (Christian-Albrechts University, Kiel, Germany), pp. 217. ↵ KvenvAgeden, K. A. (1988) Global Biogeochem. Cycles 2 , 221-229. ↵ Yokoyama, Y., Esat, T. M. & Lambeck, K. (2001) Earth Planet. Sci. Lett. 193 , 579-587. LaunchUrlCrossRef ↵ Nisbet, E. G. (1992) J. Geophys. Res. 97 , 12859-12867. LaunchUrl ↵ Buffett, B. A. (2000) Annu. Rev. Earth Planet. Sci. 28 , 477-507. LaunchUrlCrossRef ↵ Munk, W. & Wunsch, C. (1998) Deep-Sea Res. Part I 45 , 1977-2010. LaunchUrlCrossRef ↵ Leifer, I., Clark, J. F. & Chen, R. B. (2000) Geophys. Res. Lett. 27 , 3711-3714. LaunchUrlCrossRef ↵ Zhang, J. & Quay, P. D. (1997) Deep-Sea Res. Part II 44 , 2163-2190. LaunchUrl ↵ Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linden, P. J. & Xiaosu, D. (2001) Climate Change, 2001: The Scientific Basis—Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) (Cambridge Univ. Press, Cambridge, U.K.). ↵ Blunier, T. & Brook, E. J. (2001) Science 291 , 109-112. pmid:11141558 LaunchUrlAbstract/FREE Full Text ↵ Hardjono, Asikin, T. S., Purnomo, J. & Rau, J. L. e. (1998) Thirty-Third Annual Session of the Coordinating Committee for Coastal and Offshore Geoscience Programmes in East and Southeast Asia 33 , Part II, 56-83. LaunchUrl ↵ Neben, S., Hinz, K. & BeiersExecuterf, H. (1998) in Gas Hydrates—Relevance to World Margin Stability and Climatic Change, eds. Henriet, J.-P. & Mienert, J. (The Geological Society, LonExecuten), Vol. 137, pp. 255-265. LaunchUrl ↵ Dansgaard, W., Johnsen, S. J., Clausen, H. B., Dahl-Jensen, D., Gundestrup, N. S., Hammer, C. U., Hvidberg, C. S., Steffensen, J. P., SveinbjörnsExecutettir, A. E., Jouzel, J., et al. (1993) Nature 364 , 218-220. LaunchUrlCrossRef
Like (0) or Share (0)