Paleoprecipitation record from coral Sr/Ca and δ18O during the mid Holocene in the northern South China Sea

Coupled high-resolution Sr/Ca and delta O-18 records of a modern and a mid-Holocene coral from Sanya in the southern Hainan Island, northern South China Sea (SCS), were reported and the residual delta O-18 (Delta delta O-18) were calculated to indicate precipitation change in this region. Unlike other paleoclimate studies, this study focused on changes of precipitation time rather than precipitation amount. As negative Delta delta O-18 peaks in coral generally correspond to peak precipitations or rainy seasons in the surrounding region, the time offsets between negative Delta delta O-18 peaks and other seasonal indicators, such as sea surface temperature (SST), can well indicate the time of rainy seasons, and the precise time offsets can be estimated by the method of cross spectral analysis. The results of the modern coral indicate that the variation of the coral Delta delta O-18 lags that of the instrumental measured precipitation by about 2 months, and about 3 months to the SST derived from coral Sr/Ca ratios. This agrees well with the modern observation that the salinity change in the southern coastal regions generally lags that of the precipitation in Hianan Island by about 2 months, and the precipitation change lags about 1 month behind the SST in this region. Thus, coral Delta delta O-18 records can be a reliable proxy for the change of rainy seasons in this region. The results of the mid-Holocene coral show about 2.5 months' leading of the Delta delta O-18 variation ahead of the SST. By compensating the approximate 3 months' lag of the Delta delta O-18 variation behind the SST in modern time, the occurrence of rainy seasons during the mid Holocene may have advanced about 5-6 months. In detail, it may start around December and end around April to May with maximum occurring around February. Therefore, rainy seasons mainly occur in winter through early spring during the mid Holocene, compared with that from May through October in modern times. Such precipitation patterns appear to agree with the mid-Holocene pollen records in this region. Variations of large-scale circulation may possibly result in such a different precipitation pattern. Further studies, in particular climate model studies collaborated with meteorologists, are required for a better understanding of the mechanism.