 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
Geology, Vol. 29, NO. 4, pp. 295-298, 2001
Manuscript received August 10, 2000, revised manuscript received December 11, 2000, manuscript accepted December 21, 2000
Keywords: Cretaceous, Mesozoic, Cenozoic, paleo-oceanography, thermohaline circulation, numerical model
Introduction
Observations and ocean models suggest that the ocean circulation during warm Mesozoic-Cenozoic climates was dramatically different from its present-day pattern (Barron and Peterson, 1989; Bice et al., 1997; Gordon, 1973; Hay et al., 1981; Kennett, 1977; Kutzbach and Ziegler, 1993; Poulsen et al., 1998; Seidov, 1986; Shackleton and Kennett, 1975; see more in Barrera and Johnson, 1999). The warm, ice free Cretaceous period (65-135 m.y. ago) presents perhaps the most challenging problem, as no consensus exists on what climatic mechanisms could maintain warm polar climate with very small meridional and vertical thermal gradients in the world ocean.
Although meridional oceanic heat transport can be called upon to explain warm sea surface in the high latitudes (e.g., Schmidt and Mysak, 1996), keeping the high-latitude sea surface at about 10 to 15 ºC in both hemispheres would necessitate substantial equatorially symmetric oceanic poleward heat transport. The assumption of equatorially symmetric high-latitude sea surface temperatures (SST) is often used in atmospheric modeling (Sloan and Barron, 1992) and implicitly in data interpretation. However, there are indications that southern subpolar ocean was warmer than northern oceans (e.g., Huber and Sloan, 2000). Sloan et al. (1995) argued that increased poleward heat transport is difficult to accomplish in case of reduced oceanic thermal contrasts. They suggested that atmospheric feedbacks, in conjunction with the increased greenhouse gases, might be responsible for warming the poles. Of the greenhouse gases, higher CO2 level during the Cretaceous is often chosen to explain the warm equable Cretaceous-Eocene climate (Allen, 1997; Barron et al., 1995; Sloan and Rea, 1995; Sloan et al., 1995; Thomas et al., 2000). Although substantial increase of CO2 is the signature of warm climates (e.g., Barron and Washington, 1985) this increase could not be too strong without overheating the tropics, as the CO2-induced warming occurs everywhere (e.g., Sloan and Rea, 1995). The equatorial regions, however, either were as warm as today or even cooler (Crowley and Zachos, 2000; see also a review by Valdes, 2000). Even an eightfold CO2 increase appears to be insufficient to raise the Earth temperature to the Cretaceous mean values (e.g., Valdes, 2000).
The warm deep water is usually associated with high-latitude deep-water sources (see review in Valdes, 2000). However, because the sea-water density depends on both temperature and salinity, it may be questioned whether the deep-ocean water temperature (which is a direct geologic evidence) reflects the warm polar surface-ocean regions (a deduced supposition). De-densification of surface waters shuts off convection and reduces meridional overturning. Reducing the overturning leads to reducing poleward heat transport and vice versa, cooling and/or increase of salinity initiate or strengthen convection and lead to stronger overturning and stronger poleward heat transport. Thus, a compromise can be found in a scenario with relatively cool high-latitude surface at least in one hemisphere that would allow for a reasonable poleward heat transport and warm enough deep water produced in the opposite hemisphere.
The present-day deep-ocean circulation is driven by two deepwater sources - North Atlantic Deep Water (NADW) and Antarctic Bottom Water (AABW). Both water masses are characterized by their own distinct temperature and salinity. The NADW, which is slightly warmer, flows southward above the northward-flowing, cooler AABW, forming the well-known layered structure of the abyss. The structure of water masses during warm climates is less known. However, it may be assumed that the deep-ocean currents are driven similarly to their present-day analogue, as required by the Stommel and Arons (1960) theory. Therefore, it may also be assumed that past ocean circulation was sensitive to high-latitude salinity distribution (e.g., Bryan, 1986; Manabe and Stouffer, 1988; Stocker et al., 1992; see extended references in Seidov and Haupt, 1999)). Asynchrony of some of the glacial cycles of the Pleistocene implies that both the southern and northern deepwater sources could be affected by asynchronous meltwater events (Antarctica and North Atlantic show mixed record indicating that either synchrony, or asynchrony could occur, e.g., Blunier et al., 1998; Bond et al., 1997.). In some instances, the meridional overturning might have behaved as a bipolar seesaw with a periodicity of hundreds to a thousand years or more (e.g., Broecker, 1998; Seidov and Maslin, 2001; Stocker, 1998).
In cold climates, low sea-surface salinity (SSS) in the high latitudes is due
mainly to melting of sea ice or icebergs; poleward water vapor transport from
the tropics is an important but secondary factor. A high salinity signal is
due to freezing seawater that leads to salt brine rejection (e.g., Gill, 1982).
In warm ice free climates, poleward water vapor transport or river runoff can
be the only causes of a low salinity signal, whereas increased evaporation (unlikely
in high latitudes) might be a cause of increased high-latitude surface salinity.
However, in our approach, we can model an impact of temperature increase by
changing salinity, as our concern is sea surface density rather than temperature
or salinity themselves. A useful rule of thumb is that the same density increase
can be achieved by either increase of salinity by approximately 1 psu (practical
salinity unit) or by decrease of temperature by about -5 °C (Pond and Pickard,
1983).
Numerical Experiments
The Modular Ocean Circulation Model, version 2.2 (e.g., Cox, 1984; Pacanowski, 1996) is used here with a grid resolution of 4° latitude x 4° longitude and 16 unevenly spaced vertical layers with annual mean SST, SSS, and wind stress. All runs are 2000 model years long, with five-fold acceleration in the deep layers (this means that the deep ocean is effectively run for 10000 yr). A complete steady state is reached in all numerical experiments.
The thermohaline ocean circulation is determined by density distribution and thus is controlled by surface density, which depends on both temperature and salinity. In the Cretaceous scenarios we specify the sea-surface thermal conditions. This requires that we make specific assumptions about the equator-to-pole sea surface temperature gradients that cannot be changed in sensitivity tests. Here, in order to vary sea-surface density, we keep SST unchanged and vary SSS.
To test the role of hemispheric asymmetry of the Cretaceous climate, we explore two different climatic scenarios described by Poulsen et al. (1998) and Poulsen (1999). Two different types of Cretaceous surface climates (comprising SST, SSS, and wind stress) were computed in ocean model numerical experiments driven by GENESIS atmospheric model (Thompson and Pollard, 1997) in Poulsen (1999); model output was provided to us by Poulsen (1999, personal commun.). The land-sea distribution (Fig. 1a) and bottom topography are from Poulsen et al. (1998).
Although it is called an intermediate Cretaceous scenario, the scenario with a relatively cool subpolar ocean is still a very warm climate, if compared to today's. The principal difference of this scenario from the "warm Cretaceous" scenario, with up to 20 ºC in the subpolar regions, is that the intermediate scenario bears a noticeable south-north SST asymmetry, the northern subpolar ocean SST being "only" 6 ºC, whereas the southern ocean subpolar SST is 12 ºC. The equatorial SST is about 28 ºC in the intermediate and 31 °C in the warm scenario (Fig. 1b) shows zonally averaged SSTs, and therefore depicts a slightly lower maximum value).
To test sensitivity to the density changes in the high latitudes that are known
to be the strongest regulator of the ocean global thermohaline conveyor, salinity
anomalies are added to the SSS field as circumglobal bands between the Antarctica
and 60 ºS (crosshatched area in (Fig.
1b), the amplitude being varied in different runs from -1 to +1 psu. The
perturbed salinity was smoothed in 8º latitudinal bands to the north of
the anomaly edge by use of a low-pass filter (Shapiro, 1971) to blend the modified
SSS to the unchanged field. The wind stress and SSTs (different in the two scenarios)
were held the same as in the control runs in the corresponding scenarios in
order to modify only one variable at a time. Note, however, that we do not specify
what kind of high-latitude impact could change salinity. Our only goal is to
test whether the Cretaceous ocean circulation is sensitive to perturbations
of the sea-surface density, i.e., to examine how robust the deep-ocean circulation
patterns are, and whether they favor a warm or cold abyssal ocean if one of
the poles is relatively cool. Possibly an equivalent temperature change would
produce the same changes of the sea-surface density. As mentioned above, we
keep SSTs unchanged. Thus, we choose SSS to be the variable controlling density
in our idealized perturbation experiments.
Results
The oceanic heat transport (Fig. 2), if added to other impacts, such as increased CO2 and increased water vapor (e.g., Sloan et al., 1995), seems to be consistent with this moderately warm subpolar sea surface. This is because even moderately cooler high-latitude ocean surface in the intermediate scenario leads to substantial changes of the thermohaline conveyor operation, if compared to the warm case, with strongly intensified southern overturning (see below).
Figure 2 shows noticeable cross-equatorial heat transport in the intermediate scenario, whereas there is almost no cross-equatorial heat transport in the warm scenario. Importantly, in this scenario the southern and northern deepwater sources trade places in driving the conveyor. The warmer Southern Ocean is saltier and therefore denser than the northern subpolar oceans, and the Drake Passage is nearly closed (200 m deep in this geometry; Poulsen, 1999). At the same time, the northern basins are colder and fresher, and the Tethys passageway prevents formation of an analogue to the present-day Gulf Stream in the Northern Hemisphere (compare the land-sea distribution in Fig. 1a with the present-day ocean geometry).
Meridional temperature sections (Fig. 3) along 120ºW in the Pacific Ocean (along the meridional arc in Fig. 1a) show the thermal structure in the two scenarios. The intermediate scenario, though less sensitive to additional density changes in the subpolar surface oceans, provides a realistic combination of relatively warm bottom water and warm southern but cold northern subpolar surface ocean. The warm Cretaceous scenario does provide even warmer bottom water but at the expense of unrealistically warm surface ocean in the high latitudes. Thus, the intermediate scenario is consistent with the Southern Hemisphere being warmed by southward cross-equatorial oceanic heat transport, and with the warm deep ocean coexisting with a reasonable oceanic heat transport amount. In other words, we argue that a combination of warm deep ocean and cool high-latitude ocean surface in one of the hemispheres is possible without sacrificing the main assumption that the Cretaceous ocean was warm in the abyss.
Although in the warm scenario the ocean is warm in the abyss (18-21 ºC bottom water, Fig. 3b), this scenario presents the major problem for explaining the warm sea surface in the polar regions (surface temperature varies from approximately 30 ºC at the equator to 20 ºC at the poles; Fig. 1b). Since the equator-to-pole and top-to-bottom thermal gradients are both very small, the meridional overturning (schematically shown by arrowed lines in Fig. 4), though not that small, produces too low poleward oceanic heat transport, which would be insufficient to maintain polar oceans that warm were the subpolar SST allowed to drift from the specified values (e.g., in an ocean-atmosphere model).
The intermediate scenario has a robust circulation pattern undisturbed by the
perturbations of the surface salinity, the heat transport being virtually unchanged
(Fig. 2), whereas the circulation in the
warm scenario indicates a substantial change of the meridional overturning and
shows a northward cross-equatorial heat transport, which would have produced
an asymmetry in surface climatology, where the model would allow SST to drift
away from the specified values (e.g., in a coupled ocean-atmosphere model).
Discussion and Conclusions
The Cretaceous ocean thermohaline conveyor, though different from its present-day analogue, might have operated similarly to the present-day mode, the only difference being that the amplitude might have been lower than today (because of smaller equator-to-pole density gradients), and the driving deepwater sources may have switched roles, AABW taking the NADW role in driving the abyssal ocean currents. However, it could also be that geometry-induced north-south asymmetry was a permanent feature responsible for warming the abyssal ocean (Fig. 4).
We emphasize that our experiments do not target the details of the Mesozoic-Cenozoic
ocean circulations; they are not designed to explain how such warm climates
could exist in the first place. This far more complex problem can be addressed
only by using an advanced coupled ocean-atmosphere model. There must be some
value of SSS perturbations that would cause bifurcation and reversing of the
conveyor that would lead to cooling of the deep ocean. However, within a reasonable
interval of SSS variances, such rebounds were not achieved in the experiments
shown here, implying that SST variations must be added for the conveyor to be
bifurcated and for warm deep water to be replaced by colder water. Showing,
as we have here, that moderately warm deep ocean can easily coexist with moderately
warm, or even relatively cool, subpolar oceans removes the restriction on poleward
heat transport and allows for the warm deep ocean. Thus, it removes the major
paradox of too-warm high latitudes having no plausible physical mechanisms existing
that could maintain such excessive high-latitude warmth.
Acknowledgments
We thank Chris Poulsen for providing us with the Cretaceous surface boundary
conditions and paleobathymetric data; Lisa Sloan for discussions and her help
in writing the manuscript; and Eric Barron for support and valuable comments.
The editor, Dr. B.A. van der Pluijm, and two reviewers provided very useful
comments that helped to substantially improve the manuscript. This study is
supported by the National Science Foundation (project 9975107).
References
Allen, P.A., 1997, Earth surface processes: London, Blackwell Science, 404 p. Barrera, E., and Johnson, C.C., 1999, Evolution of the Cretaceous ocean-climate system: Boulder, Colorado, Geological Society of America, 445 p. Barron, E.J., and Peterson, W.H., 1989, Model simulations of the Cretaceous ocean circulation: Science, v. 244, p. 684-686. Barron, E.J., and Washington, W.M., 1985, Warm Cretaceous climates: High atmospheric CO2 as a plausible mechanism, in Sundquist, E.T., and Broecker, W.S., eds., The carbon cycle and atmospheric CO2: Natural variations Archean to present: American Geophysical Union Monograph, 32, p. 546-553. Barron, E.J., Fawcett, P.J., Peterson, W.H., Pollard, D., and Thompson, S.L., 1995, A "simulation" of mid-Cretaceous climate: Paleoceanography, v. 10, p. 953-962. Bice, K.L., Barron, E.J., and Peterson, W.H., 1997, Continental runoff and early Cenozoic bottom-water sources: Geology, v. 25, p. 951-954. Blunier, T., Chappellaz, J., Schwander, J., Daellenbach, A., Stauffer, B., Stocker, T.F., Raynaud, D., Jouzel, J., Clausen, H.B., Hammer, C.U., and Johnsen, S.J., 1998, Asynchrony of Antarctic and Greenland climate change during the last glacial period: Nature, v. 394, p. 739-743. Bond, G., Showers, W., Cheseby, M., Lotti, R., Almasi, P., deMenocal, P., Priore, P., Cullen, H., Hajdas, I., and Bonani, G., 1997, A pervasive millennial-scale cycle in North Atlantic and glacial climates: Science, v. 278, p. 1257-1366. Broecker, W.S., 1998, Paleocean circulation during the last deglaciation: A bipolar seesaw?: Paleoceanography, v. 13, p. 119-121. Bryan, F., 1986, High-latitude salinity effects and interhemispheric thermohaline circulations: Science, v. 323, p. 301-304. Cox, M., 1984, A primitive equation, 3-dimensional model of the ocean: Princeton, University, Ocean Group, Geophysical Fluid Dynamics Laboratory, Technical Report 1, 250 p. Crowley, T.J., and Zachos, J.C., 2000, Comparison of zonal temperature profiles for past warm time periods, in Huber, B.T., et al., eds., Warm climates in earth history: Cambridge, UK, Cambridge University Press, p. 50-76. Gill, E.G., 1982, Atmosphere-ocean dynamics: San Diego, Academic Press, 666 p. Gordon, W.A., 1973, Marine life and ocean surface currents in the Cretaceous: Journal of Geology, v. 81, p. 671-684. Hay, W.W., Barron, E.J., Sloan, J.L., and Southam, J.R., 1981, Continental drift and the global pattern of sedimentation: Geologische Rundschau, v. 70, p. 302-315. Huber, M.A., and Sloan, L.C., 2000, Climatic responses to tropical sea surface temperature changes on a "greenhouse" Earth: Paleoceanography, v. 15, p. 443-450. Kennett, J.P., 1977, Cenozoic evolution of Antarctic glaciation, the circum-Antarctic Ocean, and their impact on global paleoceanography: Journal of Geophysical Research, v. 82, p. 3843-3860. Kutzbach, J.E., and Ziegler, A.M., 1993, Simulation of Late Permian climate and biomes with an atmosphere-ocean model: Comparisons with observations: Royal Society of London Philosophical Transactions, Ser. B, v. 341, p. 327-340. Manabe, S., and Stouffer, R.J., 1988, Two stable equilibria of a coupled ocean-atmosphere model: Journal of Climate, v. 1, p. 841-866. Pacanowski, R.C., editor, 1996, MOM-2, Documentation, user's guide and reference manual: Princeton, New Jersey, Geophysical Fluid Dynamics Laboratory/NOAA, 329 p. Pond, S., and Pickard, G.L., 1983, Introductory dynamical oceanography: Oxford, UK, Pergamon Press, 329 p. Poulsen, C.J., 1999, The mid-Cretaceous ocean circulation and its impact on Greenhouse climate dynamics: University Park, Pennsylvania, Earth System Science Center/College of Earth & Mineral Sciences, 219 p. Poulsen, C., Seidov, D., Barron, E., and Peterson, W., 1998, The impact of paleogeographic evolution on the surface oceanic circulation and the marine environment within the mid-Cretaceous Tethys: Paleoceanography, v. 13, p. 546-559. Schmidt, G.A., and Mysak, L.A., 1996, Can increased poleward oceanic heat flux explain the warm Cretaceous climate?: Paleoceanography, v. 11, p. 579-593. Seidov, D.G., 1986, Numerical modeling of the ocean circulation and paleocirculation, in Hsu, K., ed., Mesozoic and Cenozoic oceans: Geodynamics Series, v. 15, American Geophysical Union, p. 11-26. Seidov, D., and Haupt, B.J., 1999, Last glacial and meltwater interbasin water exchanges and sedimentation in the world ocean: Paleoceanography, v. 14, p. 760-769. Seidov, D., and Maslin, M., 2001, Atlantic Ocean heat piracy and the bi-polar climate sea-saw during Heinrich and Dansgaard-Oeschger events: Journal of Quaternary Science, v. 16, (in press). Shackleton, N.J., and Kennett, J.P., 1975, Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: Oxygen and carbon isotope analysis in DSDP Sites 277, 279 and 28, in Kennett, J.P., C. C. von der Borch, et al., ed., Initial Reports of the Deep Sea Drilling Project, Volume 29: Washington, D.C., U.S. Government Printing Office, p. 743-756. Shapiro, R., 1971, The use of linear filtering as a parameterization of atmospheric diffusion: Journal of Atmospheric Sciences, v. 28, p. 523-531. Sloan, L.C., and Barron, E.J., 1992, A comparison of Eocene climate model results to quantified paleoclimatic interpretations: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 93, p. 183-202. Sloan, L.C., and Rea, D.K., 1995, Atmospheric carbon dioxide and early Eocene climate: A general circulation modeling sensitivity study: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 119, p. 275-295. Sloan, L.C., Walker, J.C.G., and Moore T.C., Jr., 1995, Possible role of oceanic heat transport in early Eocene climate: Paleoceanography, v. 10, p. 347-356. Stocker, T.F., 1998, The seesaw effect: Science, v. 282, p. 61-62. Stocker, T.F., Wright, D.G., and Broecker, W.S., 1992, The influence of high-latitude surface forcing on the global thermohaline circulation: Paleoceanography, v. 7, p. 529-541. Stommel, H., and Arons, A.B., 1960, On the abyssal circulation of the world ocean, I, Stationary planetary flow patterns on a sphere: Deep Sea Research, v. 6, p. 140-154. Thomas, E., Zachos, J.C., and Bralower, T.J., 2000, Deep-sea environments on a warm earth: latest Paleocene-early Eocene, in Huber, B.T., et al., eds., Warm climates in earth history: Cambridge, UK, Cambridge University Press, p. 132-160. Thompson, S.L., and Pollard, D., 1997, Greenland and Antarctic mass balances for present and doubled atmospheric CO2 from the GENESIS version-2 global climate model: Journal of Climate, v. 10, p. 871-900. Valdes, P., 2000, Warm climate forcing mechanisms, in Huber, B.T., et al., eds., Warm climates in earth history: Cambridge, UK, Cambridge University Press, p. 3-20.
a)
b) 
Figure 1. a: Reconstruction of land-sea distribution for mid-Cretaceous
time. Crosshatched area around Antarctica is circumglobal band of surface ocean
with salinity anomalies added in sensitivity tests (see text). Solid line shows
position of sections in Figure 4. b: Zonally
averaged sea-surface temperature representing two surface climates: intermediate
(solid line) and warm (dashed line) Cretaceous climate scenarios.
Figure 2. Global northward heat flux in PW (1 PW = 1015 W): intermediate
Cretaceous scenario (control run and experiment with low-salinity anomaly (-1
psu) in Southern Hemisphere); warm Cretaceous scenario (control run and experiment
with low-salinity anomaly (-1 psu) in Southern Hemisphere).
a)
b)
Figure 3. Meridional temperature sections in Pacific Ocean at 120°W
(see Fig. 1). a: intermediate Cretaceous
scenario; b: warm Cretaceous scenario.
a)
b)
c)
Figure 4. Sketch of bipolarity of deep-ocean dynamics. a: Present-day ocean; NADW = North Atlantic Deep Water, AABW = Antarctic Bottom Water. b: Intermediate Cretaceous ocean; NHW = Northern Hemisphere water, SHW = Southern Hemisphere water. c: Warm Cretaceous ocean. Note that NA present-day overturning is shown whereas during Cretaceous (b, c) the global overturning can be only estimated. The arrowed lines show schemes of the global meridional overturning.