Abrupt Decrease in Tropical Pacific Sea Surface Salinity at End of Little Ice Age

AIMS Coral Drilling Rig Abrupt Decrease in Tropical Pacific Sea Surface Salinity at End of Little Ice Age
Science, v.295(5559), pp.1511-1514, February 22, 2002

E. J. Hendy,1* M.K. Gagan, 1 C.A. Alibert,1 M.T. McCulloch,1 J.M. Lough,2 P.J. Isdale2

1 Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia

2Australian Institute of Marine Science, Townsville, Queensland 4810, Australia.

*To whom correspondence should be addressed.
E-mail: Erica.Hendy@anu.edu.au

Researchers from the Australian Institute of Marine Science collecting a core from a massive Porites coral colony at Britomart Reef, on the Great Barrier Reef, Australia. This colony is possibly over 300 years old. The 8 coral cores used in our study spanned 120 to 420 years of continuous growth, the longest reaching back to 1565. The Australian Institute of Marine Science drilled the cores between 1984 and 1988. These photos are from a recent sampling trip to update these records by retrieving cores that also contain the last 20 years growth. [Photo: AIMS]


ABSTRACT:
A 420-year history of strontium/calcium, uranium/calcium, and oxygen isotope ratios in eight coral cores from the Great Barrier Reef, Australia, indicates that sea surface temperature and salinity were higher in the 18th century than in the 20th century. An abrupt freshening after 1870 occurred simultaneously throughout the southwestern Pacific, coinciding with cooling tropical temperatures. Higher salinities between 1565 and 1870 are best explained by a combination of advection and wind-induced evaporation resulting from a strong latitudinal temperature gradient and intensified circulation. The global Little Ice Age glacial expansion may have been driven, in part, by greater poleward transport of water vapor from the tropical Pacific.


Great Barrier Reef Coral Location Map The 8 coral cores used in our study were collected at seven reefs, 12 to 120 km apart, from the central Great Barrier Reef, Australia. The use of multiple cores allowed us to test the fidelity of individual tracers over century time scales and to establish regional-scale proxy climate signals. This map also highlights Southwest Pacific coral records from other researchers (also available from the World Data Center for Paleoclimatology website). These extra records have enabled us to trace the climate signals seen in the Great Barrier Reef records across the Pacific. [Figure: E. Hendy]

Download data from this study from WDC Paleo as an Excel file or in Text format plus Data Description.
Download additional coral data from the WDC Paleo archive.

Fig. 1. Composite GBR coral records of (A) Sr/Ca, (B) U/Ca and (C) d18O at pent-annual resolution. The composite reconstructions for each of the three proxies were calculated by normalizing all records relative to the longest continuous record, and then taking the average. The number of records averaged at each pent-annual interval for the d18O reconstruction is shown in the bar graph (D) and is based on all 8 coral cores. The U/Ca and Sr/Ca composite records are constructed from 7 of the 8 cores. The solid line in graphs (A), (B) and (C) is the composite reconstruction normalized to the period 1860-1985, plotted as ratios (left axis) and SSTA (right axis). The 95% confidence intervals calculated for each 5-year period are shown by the dotted lines surrounding the solid line of the reconstruction. SSTA conversions are -61.5 mol/mol per C for Sr/Ca and -46.5 nmol/mol per C for U/Ca from previously published high-resolution SST-slope calibrations of Alibert & McCulloch (Paleoceanography,12:345,1997) and Min et al. (Geochim. Cosmochim. Acta, 59: 2025, 1995) respectively. The horizontal dashed lines define a 1C band according to these slope calibrations. Download the Sr/Ca, U/Ca and d18O data for the composite records and 95% confidence envelopes in (A), (B) and (C), and the number of coral cores that contributed to each proxy composite record through time. Fig. 1 Composite Coral Isotope and Element Data

Fig. 2 Comparison of Coral and Instrumental Data Fig. 2. Comparison of coral and instrumental temperature records. GBR composite SSTA reconstructions derived from coral Sr/Ca (A) and U/Ca (B), converted to SSTA using slope calibrations from Alibert & McCulloch (Paleoceanography,12:345,1997) and Min et al. (Geochim. Cosmochim. Acta, 59: 2025, 1995) respectively. (C) UKMO composite SSTA records for the Tropical Western Pacific (20N-20S, 120E-170W, GOSTA). (D) The Rarotonga Sr/Ca-SST record (21.5S, 159.5W) of Linsley et al. (Science, 290:1145, 2000) is plotted using the Sr/Ca-SST slope calibration given in the reference. (E) and (F) are proxy-temperature (land and SST) compilations for the Southern and Northern Hemispheres, respectively from Jones et al (The Holocene, 8:455, 1998). Jones et al caution that the Southern Hemisphere reconstruction (E) is a poor representative of temperature because it is constructed from only 7 paleotemperature records. All records plotted for comparison are resampled to equivalent 5-year averages, and all series are normalised to the common period 1860-1985.

Fig. 3, Regional Coral Isotope Records

Fig. 3. Comparison of regional coral d18O records and proxies of Southern Hemisphere ocean-atmosphere circulation. The composite GBR d18O record (A) is reduced to a residual d18O record (B) by subtracting the temperature signal from the GBR Sr/Ca record. Propagation of total uncertainties gives an average 95% for (B) of 0.15 from 1700-1780 AD, and 0.11 from 1780-1985 AD. The three southwestern Pacific d18O records are: (C) Abraham Reef, GBR (22S, 153E) of Druffel and Griffin (J. Geophys. Res., 98:20,1993), (D) Amede Lighthouse, New Caledonia (22S, 166E) of Quinn et al. (Paleoceanography, 13:412,1998), and (E) Espirtu Santos, Vanuatu (15S, 167E) of Quinn et al. (Quat. Sci. Rev. 12:407,1993). All records are resampled to equivalent 5-year averages and each d18O record is normalised to the average of the whole record. (F) The annual Abraham Reef d14C record of Druffel and Griffin (1993) is calculated as an anomaly relative to the expected ocean mixed layer estimates of d14C (values reported by Druffel & Griffin in Fig. 1B). The d14C estimates were modeled from 20-yr tree-ring (atmospheric) d14C values and an ocean box-diffusion model. (G) Number of large particles (diameter > 1.59 m) in the Quelccaya ice core (14S, 71W) of Thompson et al. (Science, 234: 361, 1986) plotted as an anomaly from the mean (1560-1980). The * highlights the extreme dust concentrations corresponding to the local 1600 AD eruption of Huaynaputina, Peru.


To read or view the full study, please visit the Science website. It was published in Science Vol. 295, Issue 5559, pp. 1511-1514, February 22, 2002.

Contact Us
National Oceanic and Atmospheric Administration
21 February
2002