THE LAST INTERGLACIAL SEA-LEVEL CHANGE - NEW EVIDENCE FROM THE ABROLHOS ISLANDS, WEST AUSTRALIA

U-series ages measured by thermal ionisation mass spectrometry (TIMS) are reported for a Last Interglacial (LI) fossil coral core from the T urtle Bay, Houtman Abrolhos islands, western Australia. The core is 33 .4 m long the top of which is approximately 5 m a.p.s.l. (above presen t sea level). From the Th-232 concentrations and the reliability of th e U-series ages, two sections in the core can be distinguished. Calcul ated U/Th ages in core section I (3.3 m a.p.s.l to 11 m b.p.s.l) vary between 124+/-1.7 ka BP (3.3 m a.p.s.l.) and 132.5+/-1.8 ka (4 m b.p.s .l. i.e. below present sea level), and those of section II (11-23 m b. p.s.l.) between 140+/-3 and 214+/-5 ka BP, respectively. The ages of c ore section T are in almost per feet chronological order, whereas for section II no clear age-depth relationship of the samples can be recog nised. Further assessments based on the partial derivative(234)U((T)) criteria reveal that none of the samples of core section II give relia ble ages, whereas for core section I several samples can be considered to be moderately reliable within 2 ka, The data of the Turtle Bay cor e complement and extend our previous work from the Houtman Abrolhos sh owing that the sea level reached a height of approximately 4 m b.p.s.l at approximately 134 ka BP and a sea level highstand of at least 3.3 m a.p.s.l. at approximately 124 ka BP. Sea level dropped below its pre sent position at approximately 116 ka BP. Although the new data are in general accord with the Milankovitch theory of climate change, a deta iled comparison reveals considerable differences between the Holocene and LI sea level rise as monitored relative to the Houtman Abrolhos is lands. These observation apparently add further evidence to the growin g set of data that the LI sea level rise started earlier than recognis ed. by SPECMAP chronology. A reconciliation of these con tradictionary observations following the line of arguments presented by Crowley (19 94) are discussed with respect to the Milankovitch theory.