Paleoceanography and Paleoclimatology Program at USC

Current research focuses on reconstructing the history of ocean/climate variability in the tropics during the past 70 thousand years. This information should provide insight that will lead to better assessments of climate may vary in the future .


Climate/Ocean History of the Western Tropical Pacific

With support from the National Science Foundation (NSF) we are investigating ocean and climate variability in the tropics. We are collaborating with a team of international colleagues in an effort to document the spatial and temporal patterns of ocean and climate variability within the tropics and to use the observations to test hypotheses about how tropical ocean/atmospheric variability has influenced Earth's large climatic changes in the past.

My group at the University of Southern California is currently attempting to reconstruct the history of ocean temperature and salinity variability within the western tropical Pacific, the so-called Pacific Warm Pool. This region is the heat engine that drives atmospheric convection. Today this region is the center activity for El Nino-La Nina, the interannual disruption of trade winds and ocean currents within the tropics that has wide spread influence on weather patterns. Our goal is to assess how the tropical ocean has changed on longer time scales, information that we believe will allow us to make better predictions of future variability, particularly as the oceans warm. To do this we use a variety of geochemical techniques (proxies) that extract temperature and salinity information from marine sediment archives. These studies have changed our views of how dynamic the tropical ocean is. It is now clear that the tropics undergo large and sustained changes in temperature and salinity and these changes occurred naturally on a variety of time scales. Below are a few excerpts from some of our studies.


Ongoing Research:

In 1998, in collaboration with an international team of scientists associated with IMAGES program, we participated in a coring cruise (IMAGES IV) in the Indonesian Archipelago that recovered 35 sediment cores from the western tropical Pacific. Samples from these cores are being analyzed in effort to reconstruct sea surface tempertature and salinity variability within the Pacific Warm Pool at a decadal resolution. This is possible because of the high sediment accumulation rates that characterize the Indonesian region and because of the excellent preservation of materials recovered from the cores..


The IMAGES ship Marion Dufresne (above left) and one of the long cores (above right) recovered from Indonesia in 1998


Tropical climate records with a resolution comparable to that of the polar ice cores are needed to accurately establish the nature of millennial-scale climate change in this region and to determine the phasing of climate/ocean changes between the low and high latitudes. The Indonesian cores provide this level of resolution (see figure below).

Figure above: 14C ages (year BP) for planktonic (surface-dwelling) planktonic foraminifers and benthic (deep water) foraminifers taken from discrete horizons of the MD81 core, which illustrates the high sediment accumulation rate that typifies these cores. The age offset between the planktonic and benthic foraminifers is a reflection of the age of deep waters at this equatorial site (approximately 1000 years old). The sediment accumulation rate estimated from these 14C results is between 50 and 80cm/ky.


We are particularly interested in reconstructing the history of SST variability and surface ocean salinity in the western tropical Pacific because this is the region of highest heat and moisture flux to the atmosphere, and is at the center of activity associated with the El Niño-Southern Oscillation (ENSO). The region is also the point of origin for the upper limb of the ocean thermohaline conveyor system, which transports heat and salt to the North Atlantic via the surface ocean to compensate for North Atlantic Deep Water (NADW) production. For these reasons, it is possible that the tropics may have contributed to or even caused abrupt climate changes that were felt world wide during the last glacial cycle.

We have several goals for this research:

1. To investigate how sea surface temperature and salinity varied in the western tropical Pacific during the past 70 thousand years.

2. To investigate whether there are recurring patterns of hydrographic variability in the warm pool.

3. To investigate how the tropical Pacific climate variability relates to larger, global-scale climate variability in an effort to determine cause and consequences of large

abrupt climate change.

In the plots below we show the oxygen isotopic record from GRIP, one of the Greenland Ice core records containing a record of abrupt climate variability during the last 80 thousand years. Below that we illustrate the portion of this record referred to as Marine Isotope stage 3 which contains a remarkable record of abrupt climate variability at high northern latitudes. One of the most surprising discoveries we have made thus far is that the same abrupt climate events are evident in the western tropical Pacific marine record!

Greenland Ice Core Climate Record

The figure on the left is the oxygen isotope record from the Greenland GRIP ice core. The interval between 80 and 20ka is highlighted in each panel to illustrate the abrupt nature of climate variability found in this time interval. The d18O is a record of air temperature over greenland. Up in the diagram reflects warming. Down reflects colder conditions. The magnitude of the climate variations reflected in the d18O excursions in this high latitude location were large, perhaps around 15oC during the largest interstadials.


















Western Tropical Pacific Marine Record

The figure above (right axis) shows the d18O (ratio of 18O to 16O in the calcium carbonate) of a species of planktonic foraminfiera (blue) from a core taken in the western tropical Pacific, near the island of Mindanao (MD81 upper figure) plotted with the ice core d18O record (red). Note how the Indonesian d18O record varies in close correspondance to temperature variations reflected in the d18O record from the Greenland ice core. These results indicate a pattern of millenial scale SST variability in the tropics that matches the so-called Dansgaard-Oeschger (D-O) cycles in Greenland. However, because d18O in the foraminiferal calcite varies in response to both temperature and to changes in 18O composition of local waters, it is impossible to say how much of the tropical variability is due to temperature based upon the d18O record alone. Fortunately, there are new techniques available that allow us to measure paleotemperatures from the same fossil carbonates. One of these techniques involves the ratio of Mg to Ca (measured in millimoles of Mg to moles of Ca), which varies systematically with calcite precipitation temperature. We can now measure d18O and Mg/Ca on carbonates from the same sample horizon and determine the temperature and the oxygen isotope composition of surface waters at the time the foram secreted it's calcite. This has important implications to tropical paleoceanography because the d18O of surface water varies systematically with salinity (the 18O/16O of ocean water increases as salinity increases). We are now able to reconstruct both temperature and salinity from fossil carbonates (see these papers, 1, 2, 3, 4).


In the figure above we have extracted the temperature component from the foraminiferal d18O signal in the top figure using the Mg/Ca temperatures in the top plot (A). The variations in d18O in the foraminiferal calcite is shown in top plot (A) together with the isotope record from the Greenland GISPII ice core, both record millenial variability that indicates a close coupling between tropical and high northern hemisphere climate changes during the last glacial..

The lower plot (B) is the d18O of the Hulu Cave in China (Wang et al., 2001, Science) compared with the reconstructed record of tropical Pacific surface water salinity varibility (Dd18O) and the d18O of the foraminiferal calcite. These are each plotted on the common time scale as constructed for the Hulu Cave using radiogenic isotope chronometry.. Since the local water d18O varies directly with salinity, the Dd18O is a reflection of changes in local surface water salinity after factoring out the ice volume component. We interepret the millenial fluctuations in Dd18O to reflect changes in precipitation over Indonesia that affected surface salinities. The close relationship between Dd18O and the GISP ice core temperature record suggests that the the tropics were involved in the large scale millenial climate changes.



A bi-polar signal recorded in the western tropical Pacific: Northern and Southern Hemisphere climate records from the Pacific warm pool during the last Ice Age



MD81 benthic vs Antarctic

Figure above, Byrd d18O and MD98-2181 U. hispida d18O plotted with respect to the GISP2 age scale of Blunier and Brook (2001)


The benthic δ18O record from MD98-2181 (above plot) documents upper Pacific DeepWater temperature and salinity variability (Figure below). Benthic d18O variations of 0.3–0.5&during MIS 3 indicate deep waters within the Pacific were varying by 1–1.5 degrees C, with the possibility that some of the variability was due to changing salinity and minor glacial–eustatic changes. The observed deep-water variability correlates to changes in Antarctic surface temperatures and thus reflects changes in Southern Ocean temperatures at the site of Pacific Deep Water formation. The combined planktonic and benthic records from MD98-2181 thus provide a northern and southern hemispheric climate record of anti-phased variability during MIS 3 as has been inferred previously fromice core records. Furthermore, the deep sea temperature excursions appear to have led millennial variations in atmospheric CO2 as recorded in the EDML ice core byw1 kyr.


Pacific Deep Water

Figure (above): (A) Map showing location of MD98-2181 and a N–S salinity transect. (B) Salinity of the Pacific Ocean along N–S transect and depth and location of MD98-2181.Water masses
Antarctic Intermediate Water (AAIW), Upper Circumpolar Deep Water (UCDW), Lower Circumpolar Deep Water (LCDW) and Antarctic Bottom Water (AABW) are shown at their
appropriate depths. SAF, Sub-Antarctic Front



The Glacial Termination--The High Southern Latitudes Warmed 1300 years before atm. CO2 began to rise  

The data obtained from high latitude ice cores establishes a close temporal relationship between varying concentrations of atmospheric CO2 and atmospheric temperatures during glacial terminations. However, uncertainty in the gas age chronologies and inadequate temporal resolution in many proxy climate reconstructions has hampered efforts to establish the exact phasing of events during glacial terminations, a necessary step in understanding the physical relationship between the CO2 forcing and climate change. We have established a chronology of high and low latitude climate change at the last glacial termination by 14C dating benthic and planktonic foraminiferal stable isotope and Mg/Ca records from a high deposition rate marine core from the western tropical Pacific. Benthic foraminiferal δ18O values from core MD98-2181 document the temperature and oxygen isotopic composition of Pacific deep water, properties that are established in the Southern Ocean, while planktonic Mg/Ca is a proxy for tropical sea surface temperatures.

Arguably, the most robust estimate of changes in mean global temperatures that accompany glacial terminations is the amount of heat stored in the oceans. The large size of the oceanic reservoir and the long residence time of deep waters mean that deep water temperatures reflect a globally averaged record of the Earth’s radiative heat balance(2). At present we cannot accurately quantify changes in ocean heat content at the last glacial termination. However, because deep waters form at high latitudes and carry with them the conservative properties of temperature and salinity, it is possible to establish what the relative timing of high latitude versus low latitude climatic change was at glacial terminations by dating co-occurring benthic (bottom-dwelling) and planktonic (surface-dwelling) foraminiferal paleotemperatures records in a marine core from a tropical location that has sufficient temporal resolution. Here we use 14C dating to establish the timing and the magnitude of the deep sea and the tropical surface ocean temperature changes during the last glacial termination and compare this history with the timing of CO2 change and deglacial warming in the southern high latitudes during the last glacial termination.

At the last glacial termination benthic δ18O values in the deep Pacific decreased by 0.5‰ between 19 and 17 years ago (ka) whereas SSTs did not begin to warm until 17 ka (see figure below). The benthic δ18O record indicates that a substantial amount of heat was added to the deep ocean before the tropics began to warm and before atmospheric CO2 began to rise. Climate model simulations utilizing the history of greenhouse gases, ice-sheet orography and orbital forcing demonstrate that austral spring insolation combined with sea ice-albedo feedbacks were key factors responsible for this warming. Atmospheric CO2 was a contributing factor but would not have been the trigger that initiated deglacial warming.



Figure above: The figure shows the Mg/Ca-based SST reconstruction from three sites located in the western tropical Pacific between about 5oS and 8oN. The tropical SSTs did not begin to warm at the last glacial termination until 17kyBP, in association with rising concentrations of atmospheric CO2. However, benthic d18O values began to decrease (indicating warming) at 18.5, more than 1000 years before the rise in atmospheric CO2. The deep water warmed by approximately 2oC prior to 17kyBP and this implies that virtually all of the deglacial warming of the deep ocean had been completed before the CO2 began to rise. This has important implication for hypothesis that have called upon CO2 itself to explain the deglacial warming. Clearly CO2 forcing was a contributing factor to the warming, but was not responsible for much of the warming that took place globally. We hypothesize that the high southern latitudes began to warm due to increasing spring-time solar insolation (see figure above), which promoted the melting of sea ice around the Antarctic (Stott et al., Science, 2007).

Marine sediment core MD98-2181 from the Morotai Basin (Figure above ) is ideally suited for documenting the timing of both deep sea and tropical surface watertemperature change during the last glacial termination because terrigenous sediments shed off of the Island of Mindanao accumulated at this site with an admixture of planktonic and benthic foraminiferal carbonate at rates of 50-80cm/ky over the past glacial/interglacial transition (3) . Sampling this core at centimeter scale provides a temporal resolution for each sample of approximately 25-50 years. This core was recovered from a depth of 2114 meters, where it was bathed in circumpolar deep water, a water mass whose temperature and salinity are acquired within the Southern Ocean. Furthermore, since the site is also located within the Pacific warm pool, the δ18O and Mg/Ca of planktonic foraminifera from this core reflect the sea surface temperature and salinity in western tropical Pacific.


The Holocene

Much of our current effort is directed towards the development of a composite record of sea surface temperature and salinity for the entire warm pool during the Holocene (past 10,000 years).. To accomplish this we are applying measurements of Mg/Ca and d18O to samples from three marine sites that provide a north-south transect through the warm pool (see figure below).


The salinity and temperature variability documented above for the Pleistocene continued into the Holocene

The three sites from Indonesia (on the left in the above figure) are shown with a record from the eastern tropical Pacific (data on right from Koutavas et al. Science, 2003). Note how the d18O values become progressively lighter through the Holocene whereas the temperature show a maximum in the early Holocene and decline (in the western Pacific) towards the present. The trend in d18O during the Holocene reflects the progressive freshening of the western tropical Pacific during the mid to late Holocene.

The freshening of sea surface salinities in the Pacific (upper panel Stott et al. 2004) were associated with reduced precipitation over the tropical Atlantic (lower panel above, Data from Haug et al., 2001, Science, Vol. 293) . Note also the millennial-scale variability that is superimposed on the longer-term trend.

We attribute the freshening of sea surface salinities in the Pacific during the Holocene to a shift in the ITCZ (Intertropical Convergence Zone) in response to the change in precessional forcing of solar radiation. As the ITCZ migrated southward from the early-mid Holocene the vapor transport between the Atlantic and Pacific increased.


The most recent manifestion of millennial-scale variability occurred in conjunction with the Medieval Warm Period/Little Ice Age in Europe.


Low-High Latitude Coupling

Shown in the figure above is the sea surface salinity record (blue) from Indonesia site MD81 vs the oxygen isotope record of temperature variability over Greenland from the N. Grip record (red). The long term trends have been removed revealing a persistent correspondence between the millennial-scale events in the Pacific and in the Greenland ice core records. The comparison suggests there is a persistent and recurring relationship between lower salinities in the warm pool and colder conditions over Greenland during the Holocene.


We believe these centennial to millennial scale climate oscillations could reflect a pattern of high-low latitude coupling that is similar to that seen in historic records for the past 100 years (Hoerling et al., Science 2002). The oscillation seen in the past 1000 years that included the Medieval Warm Period-Little Ice Age oscillation (see figure above) is the the most recent of these oscillations. We further believe that these recurrent oscillations provide an important perspective for evaluating how unique the 20th century warming in the Northern Hemisphere is.