| Summary: | The oxygen isotope composition of diatom silica (δ18Odiatom) provides valuable information for palaeoclimate studies where carbonate proxies are either rare or absent in many lakes and ocean sediments. Unfortunately, mineral and rock fragments found alongside diatoms in most sediments can be problematic as the method used will liberate oxygen from all components within the sediment, producing both high-frequency noise and low-frequency excursions that can resemble climate signals. Removal of mineral contamination to date has largely relied upon the combination of chemical leaching and physical separation techniques (e.g. sieving, density). This combination can be inefficient and often significant proportions of contaminants are present in the 'purified' diatom sample. Using electron optical imaging and whole-rock geochemistry on previously 'purified' diatom samples, a mass balance approach has been developed whereby the types and proportions of residual contaminants are identified. By integrating this information with measured oxygen isotope ratios of the contaminants, it is then possible to remove contamination effects from the δ18Odiatom record. Contamination effects relating to carbonates, tephra and silt are modelled for cores from Lake Tilo (Ethiopia) and Lake Baikal (Siberia). In both lakes the new modelled δ18Odiatom curves show less high-frequency noise, thus enabling better resolution of low-frequency climate signals. © Natural Environment Research Council (NERC) copyright 2008. Reproduced with the permission of NERC. Published by John Wiley and Sons, Ltd.
|
|---|