4:30 PM - 4:45 PM
[MGI25-05] Sea surface salinity and trade wind behavior recorded in 70-year Tuvalu coral annual bands
Keywords:Tuvalu, coral annual band, oxygen isotope ratio, sea surface salinity, trade winds, multidecadal-scale variation
The trace chemical composition of coral carbonate skeleton provides alternative information for the tropical ocean environment. In particular, the coral oxygen isotope ratio (d18Ocoral), which is determined by the SST and the oxygen isotope ratio of seawater (d18Osw), is useful as a paleo-thermometer. The oxygen isotope ratio of seawater is considered as a proxy of SSS, which is determined by run-offs from rivers, precipitation-evaporation processes, and advection. However, the contribution of SST and SSS to d18Ocoral is not constant, but varies depending on the oceanic region, geography, and climatic conditions. Reconstructions of SST and SSS from coral annual bands have been attempted in various oceanographic regions. We present the first 70-year long d18Ocoral record in the northern edge of the diagonal SPCZ, obtained from a coral core in Tuvalu (8S, 179E). Due to its location at the bifurcation of the WPWP-ITCZ and the WPWP-diagonal SPCZ, the Tuvalu δ18Ocoral is expected to be particularly sensitive to changes in the distribution of SPCZ freshwater pools, salinity fluctuations, and trade wind behavior.
Tuvalu is an atoll island with an elevation of 3-4 m and no rivers. The climatology of Funafuti Atoll, the capital of Tuvalu, has an average SST of 29.5 degree celsius and a small annual cycle with the maximum of 29.9 degree celsius and the minimum of 29.1 degree celsius. The annual precipitation is 3,000-4,000 mm, and the rainy season is from December to March with the maximum in January, while the dry season is from May to October with the minimum in September. The easterly trade winds prevail from March to September, whereas westerly winds are observed from October to February.
A coral core (98 cm) collected from Funafuti Atoll showed 70-yr continuous annual bands (1940-2009) and was analyzed for monthly d18O. The d18O variability not only shows seasonal variations in SST, but also decadal and multi-decadal-scale variations. The d18O variability range over the past 70-yr is 0.98 permil (0.6permil for 1978-2009). In the last 30 years since 1977-78, when more instrumental records were available, no high correlation was found between d18Ocoral and observed SST (r=-0.45). Given that the annual SST range is within 1 degree celsius (equivalent to 0.2permil of d18O), the d18Ocoral variation (0.6permil) in Tuvalu during the last 30 years (1978-2009) can be regarded as a mixture of SST variability of 0.2permil and d18Osw variability of 0.4permil, which is equivalent to 1.8psu in salinity.
From a comparison of d18Ocoral anomalies during the rainy season (December-March) since the 1980s with rainfall records and reanalyzed SSS, we found a low correlation with rainfall (r=-0.30), while the very high correlation with SSS (r=0.76). Furthermore, a correlation analysis between d18Ocoral anomalies during the rainy season (December-March) and the reanalyzed sea surface height around Tuvalu suggests that the northwestward surface geostrophic current is enhanced, facilitating the advection of high salinity water masses from the subtropical South Pacific. Therefore, the d18Ocoral in Tuvalu may reflect SSS anomalies caused by the northwestward advection linked to the intensity of the trade winds.
The reconstructed SSS for the 70-yr shows multidecadal-scale variations in d18Ocoral with higher values between the 1950s to 1977/78. A wavelet spectrum analysis of the 70-yr d18Ocoral record reveals significant spectra at both long-period and short-period, but both spectra became shorter in period throughout the 70-yr. The present results suggest that SSS variability could modulate these multidecadal ocean-atmosphere coupled phenomena.
