U/Th-dating living and young fossil corals from the central tropical Paci¢c

This study evaluates the accuracy of U/Th dates for young (6 a few thousand years old) reef corals, both living and fossil, and explores strategies for refining those dates. The high precision of the U/Th method (0 1^2%) for dating young corals is well-established. Earlier studies have demonstrated the method’s accuracy for select samples of known age. However, these studies have focused on typical samples that have extremely low 232Th concentrations (tens of pg/ g). Here we study the dating systematics of young corals that have low but significant amounts of 232Th (up to 1000 pg/g), indicating the presence of small fractions of non-radiogenic 230Th (i.e. 230Th not generated by in situ U decay). We report U/Th ages for living and subaerially exposed fossil corals from Palmyra Island, located in the central tropical Pacific, that range from 50 to 700 yr old. The Palmyra corals contain varying amounts of 232Th and small fractions of associated non-radiogenic 230Th. Uncertainty associated with the correction for non-radiogenic 230Th can lead to significant errors in U/Th dates. We have characterized non-radiogenic 230Th/232Th values, (Th/Th)nr, as a means of minimizing this source of error. We calculate (Th/Th)nr values ranging from 0 to 2U1035 for the Palmyra living corals by comparing measured U/Th dates to absolute dates for the living coral, whose chronology is well-established. For the fossil corals, we employ three different approaches to arrive at (Th/Th)nr estimates. First, we compare measured U/Th dates to absolute dates in samples from a young fossil coral that overlaps the living coral. Next, we use the firm relative dating constraints imposed by five overlapping fossil corals from the 14th^15th centuries to calculate (Th/Th)nr values. Finally, we attempt to anchor the 14th^15th century floating coral chronology to an absolute chronology by correlating the climate signals in the coral records to those in absolutely dated climate proxy records. All lines of evidence point to a range of (Th/Th)nr for fossil corals that overlaps the range determined for the living coral, suggesting that most of the thorium is primary or is added while the coral is still alive. Our work also demonstrates the utility of multiple (Th/Th)nr estimates. Most importantly, we demonstrate a method by which accurate ( 0 5 yr) U/Th-based chronologies can be obtained for young fossil corals with significant 232Th concentrations. 0012-821X / 03 / $ ^ see front matter A 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0012-821X(03)00138-9 * Corresponding author. Present address: California Institute of Technology, MC 100-23, Pasadena, CA 91125, USA. Tel. : +1-626-395-8126; Fax: +1-626-683-0621. E-mail addresses: [email protected] (K.M. Cobb), [email protected] (C.D. Charles), [email protected] (H. Cheng), [email protected] (M. Kastner), [email protected] (R.L. Edwards). EPSL 6616 22-4-03 Cyaan Magenta Geel Zwart Earth and Planetary Science Letters 210 (2003) 91^103 www.elsevier.com/locate/epsl A 2003 Elsevier Science B.V. All rights reserved.