A novel approach for construction of radiocarbon-based chronologies for speleothems

摘要

Robust chronologies are crucial for the correct interpretation of climate proxy records and for detailed reconstructions of palaeoclimate. Stalagmites have garnered strong interest as recorders of past climate in part due to their amenability to U-series dating. However, many stalagmites are not dateable using this technique due to low U-238 and/or high detrital Th concentrations (e.g., many tropical cave systems (Adkins et al., 2013)), and occasionally these issues affect stalagmites across wide geographical regions (e.g., large parts of Australia (Green et al. 2013)) complicating the use of stalagmites in these areas. Radiocarbon (C-14) offers an alternative method of dating stalagmites, but issues associated with the 'dead carbon fraction' (DCF) have historically hindered this approach. Here, a novel C-14-based method for dating stalagmites is presented and discussed. The technique calculates a best-fit growth rate between a time-series of stalagmite C-14 data and known atmospheric C-14 variability. The new method produces excellent results for stalagmites that satisfy four requirements: i) the absence of long-term secular variability in DCF (i.e., stalagmite DCF varies around a mean value with no long-term trend), ii) stalagmite growth rate does not vary significantly (the technique identifies stalagmites with substantial growth rate variability), iii) the stalagmite record is long enough that measurable C-14 decay has occurred, and iv) one 'anchor' point exists where the calendar age is known. The model produces good results for a previously U-Th dated stalagmite from Heshang Cave, China, and is then applied to an undated stalagmite from southern Poland. The new method will not replace high-precision U-Th measurements, because the precision of the technique is difficult to quantify. However, it provides a means for dating certain stalagmites undateable by conventional U-Th methods and for refining coarse U-Th chronologies. (C) 2016 Elsevier B.V. All rights reserved.