MESH - Marine aspects of Earth System History

On interannual timescales, the El Niño/Southern Oscillation (ENSO) causes large changes in salinity over the equatorial Pacific as the warm, low-salinity waters from the western tropical Pacific (WTP) are advected east into the central Pacific. This redistribution of warm waters along the Equator also alters the locus of atmospheric convection and seems to enhance the transport of heat from the tropics to higher latitudes. On longer timescales it is possible that other modes of variability have existed that can only be ascertained through analysis of long proxy records.

Figure 1. Mean annual sea-surface salinity in the tropical Pacific. Location of marine sediment cores used to determine sea-surface temparature and salinity in the past.

We reconstructed WTP sea surface temperature (SST) and the stable oxygen isotope composition of surface water (d18Osw) variability through the Holocene from three marine sediment cores (Fig. 1) as a method of determining how the freshwater flux and ocean dynamics varied in the past and how that variability affected salinity gradients in the tropics. In the WTP, sea surface salinity (SSS) and d18Osw are correlated, reflecting the balance between the freshwater flux (evaporation-precipitation) and salt advection by means of ocean transport. A change in the d18O of fresh water or the amount of fresh water falling on the ocean relative to inputs of salty water by means of ocean dynamics changes the salinity and also d18Osw. Groundwater d18O values of early, middle and late Holocene age from sites in the WTP all have the same value as present-day monsoon rainwater, indicating there has not been a change in the isotopic value of fresh water falling on the ocean over this period. We therefore expect WTP d18Osw values to have been controlled during the Holocene primarily by the combined effects of changing freshwater flux (the amount effect) and the flux of salty water through the WTP by means of ocean dynamics. Today these two processes result in a surface water d18O/salinity relationship that has a slope of between 0.3 and 0.4 per mil per practical salinity unit (p.s.u.).

Figure 2. Stacked records of d18Osw (left) and sea-surface temperature (right) from the western tropical Pacific.

The stacked data reveal that since approximately 8 kyr ago, d18Osw values have decreased in the WTP by 0.5 per mil (Fig. 2). On the basis of the modern d18O–salinity relationship, a decrease in d18Osw of 0.5‰ reflects a decease in surface salinity of between 1 and 1.5 p.s.u. This observation implies that in the early Holocene WTP surface water was as salty as surface water in the south central equatorial Pacific today (more than 35.2 p.s.u.). This observation further implies that if the salinities in the central and eastern equatorial Pacific have not changed in association with those in the WTP, the salinity gradient that exists today across the tropical Pacific would have been absent or significantly reduced in the early to middle Holocene.

With only a limited number of high-resolution Holocene d18Osw records available from other parts of the global ocean it is not possible to assess whether the observed changes in the WTP reflect a larger, basin-scale change in the isotopic composition of the ocean. However, numerical model simulations have suggested that sustained shifts in the location of the Intertropical Convergence Zone (ITCZ) over the tropical Atlantic Ocean would probably affect the vapour flux to the Pacific, and change the fresh water as well as the isotopic balance over the oceans. A persistent displacement of the ITCZ to more northerly latitudes in summer would act to trap isotopically light vapour within the Atlantic basin and decrease the vapour gain in the Pacific. The salinity and d18O changes we document from the WTP occurred in close association with the precessional cycle and with enhanced solar heating in the northern tropics during the early Holocene, which would have tended to pull the ITCZ north off its present summer latitude. Over the course of millennia a northerly bias in the latitude of the ITCZ and reduced vapour transport between the oceans could have affected the isotopic composition of Pacific surface waters. At present there are no continuous SSS records from the tropical Atlantic and eastern Pacific that span the entire Holocene and provide the same temporal resolution that is available for the WTP. Nonetheless, previous studies have documented an early Holocene pluvial over North Africa and a stronger Indian Ocean summer monsoon in response to Earth’s precessional cycle. Cariaco Basin sediments also contain evidence of higher rainfall in northern South America during the early Holocene, with increasingly arid conditions developing during the past 5,000 yr. These data all point to tropic-wide changes in the hydrological cycle that have been attributed to a more northerly position of the ITCZ during the early Holocene in response to changes in solar radiation associated with the precessional cycle. Data from other parts of the Pacific and Atlantic will now be required for an assessment of whether the changes in the hydrologic cycle affected the salinity gradient between the Pacific and Atlantic Oceans. If so, millennial to centennial scale changes in Holocene ocean thermohaline circulation would be directly affected by ocean–atmosphere processes that have occurred in the tropics.


	© 2005 MESH