Modification and preservation of environmental signals in speleothems

Ian Fairchild, Claire Smith, Andrew Baker, Lisa Fuller, C Spotl, D Mattey, F McDermott

Research output: Contribution to journalArticlepeer-review

563 Citations (Scopus)
1107 Downloads (Pure)

Abstract

Speleothems are primarily studied in order to generate archives of climatic change and results have led to significant advances in identifying and dating major shifts in the climate system. However, the climatological meaning of many speleothem records cannot be interpreted unequivocally; this is particularly so for more subtle shifts and shorter time periods, but the use of multiple proxies and improving understanding of formation mechanisms offers a clear way forward. An explicit description of speleothem records as time series draws attention to the nature and importance of the signal filtering processes by which the weather, the seasons and longer-term climatic and other environmental fluctuations become encoded in speleothems. We distinguish five sources of variation that influence speleothem geochemistry: atmospheric, vegetation/soil, karstic aquifer, primary speleothem crystal growth and secondary alteration and give specific examples of their influence. The direct role of climate diminishes progressively through these five factors. We identify and review a number of processes identified in recent and current work that bear significantly on the conventional interpretation of speleothem records, for example: 1) speleothem geochemistry can vary seasonally and hence a research need is to establish the proportion of growth attributable to different seasons and whether this varies over time. 2) whereas there has traditionally been a focus on monthly mean �´18O data of atmospheric moisture, current work emphasizes the importance of understanding the synoptic processes that lead to characteristic isotope signals, since changing relative abundance of different weather types might 1Corresponding author, fax +44(0)1214145528, E-mail: i.j.fairchild@bham.ac.uk control their variation on the longer-term. 3) the ecosystem and soil zone overlying the cave fundamentally imprint the carbon and trace element signals and can show characteristic variations with time. 4) new modelling on aquifer plumbing allows quantification of the effects of aquifer mixing. 5) recent work has emphasized the importance and seasonal variability of CO2-degassing leading to calcite precipitation upflow of a depositional site on carbon isotope and trace element composition of speleothems. 6) Although much is known about the chemical partitioning between water and stalagmites, variability in relation to crystal growth mechanisms and kinetics is a research frontier. 7) Aragonite is susceptible to conversion to calcite with major loss of chemical information, but the controls on the rate of this process are obscure. Analytical factors are critical to generate high-resolution speleothem records. A variety of methods of trace element analysis are available, but standardization is a common problem with the most rapid methods. New stable isotope data on Irish stalagmite CC3 compares rapid laser-ablation techniques with the conventional analysis of micromilled powders and ion microprobe methods. A high degree of comparability between techniques for �´18O is found on the mm-cm scale, but a previously described high-amplitude oxygen isotope excursion around 8.3 ka is identified as an analytical artefact related to fractionation of the laser-analysis associated with sample cracking. High-frequency variability of not less than 0.5o/oo may be an inherent feature of speleothem �´18O records.
Original languageEnglish
Pages (from-to)105-153
Number of pages49
JournalEarth Science Reviews
Volume75
DOIs
Publication statusPublished - 1 Mar 2006

Keywords

  • speleothems
  • palaeoclimate
  • time series analysis
  • karst
  • trace elements
  • stable isotopes

Fingerprint

Dive into the research topics of 'Modification and preservation of environmental signals in speleothems'. Together they form a unique fingerprint.

Cite this