Fig. 10 shows coral S18Oc time series for the last two decades of the twentieth century for our coral sites in the tropical northwest Pacific. Peak matching between S18Oc and instrumental SST with simple interpolation was employed to construct an age model of each coral record. In the Japanese and Philippine coral records, significant decreasing trends in 818Oc (-0.24% to -0.42%o/10yr) are evident, even over the relatively short period between 1980 and 2000, which may suggest an SST rise and/or surface-water freshening related to global warming. Any decreasing trend in S18Oc recorded by the Micronesian corals may have been masked by the recent large number of ENSO events, which caused local cooling and drought (S18Oc increase) during the latter part of the period. Warming or freshening trends have been found generally in coral S18Oc records from several sites in the tropical Pacific and Indian oceans (Gagan et al., 2000).
In Fig. 10, coral S18Oc curves are also matched to instrumental SSTs based on the S18Oc-SST relation obtained for each coral site. We have shown that coral records from Micronesia, and possibly the Philippines, are influenced by seawater S18Ow variation. Therefore, the coral-based SSTs from these sites should be interpreted as "apparent" SSTs (Gagan et al., 2000) because the influence of seawater S18Ow variations is not taken into account. It is also assumed that all changes in S18Oc can be attributed to temperature variation, but salinity changes also influence S18Oc values. Nevertheless, the general trends in instrumental and coral-derived "apparent" SSTs for each coral site are in good agreement, suggesting that coral records have potential as an SST proxy when discussing long-term trends. High-latitude sites, such as Ishigaki Island, showed a more rapidly increasing SST trend during the last two decades of the twentieth century. Although the long-term instrumental SST data available since 1914 for Ishigaki Port showed a distinct warming trend of 0.12 °C/10yr throughout the monitoring period (Suzuki et al., 2003), the observed SST increase during the last two decades (0.57 ° C/10 yr) is large compared with the earlier part of the century.
Comparison of "apparent" and instrumental SSTs is also a useful tool for identifying El Nino-related drought events in the western Pacific. "Apparent" SSTs during the year 1998 for the Philippine and Micronesian sites were low relative to the actual SSTs (Fig. 10). As discussed previously, the deviation of the Chuuk Atoll S18Oc curve from the instrumental SST curve in 1998 can be attributed to increased seawater S18Ow (higher salinity), corresponding to the El Nino-induced drought (Fig. 7). The Philippines experience more severe drought conditions compared with the Micronesian islands, as seen in the pattern of annual precipitation anomalies associated
I El Nino
I El Nino
Figure 10: Comparison of instrumental SST (gray curves) and Porites coral S18Oc values (curves with circles) from Ishigaki Island, Japan; Bicol, Philippines; and Chuuk Atoll and Pohnpei Island, Micronesia, in the northwestern Pacific. Regional weekly SSTs (IGOSS NMC SSTs) are shown for Bicol, Chuuk, and Pohnpe while the 10-day mean SST record of the JMA Ishigaki-jima Meteorological Observatory is shown for Ishigaki Island. Coral 8 Oc curves are related to "apparent" SSTs based on the 818Oc-SST relations: Ishigaki (IU96-07), SST - -1.4-5.57 818Oc; Bicol (SWGM01-01), SST - 0.57-5.58 818Oc; Chuuk (CHU99-01), SST - 8.1-3.60 818Oc; and Pohnpei (POH99-02), SST - 10.3-3.43 818Oc. "Apparent" SSTs derived from the coral 818Oc values include offsets due to changes in seawater 818Ow in each region. The gray and solid lines show simple regressions for with strong El Nino events (Fig. 2B), and, indeed, coral records from the Philippines showed the largest deviation among the three sites examined.
Was this article helpful?