Modelling the 14C bomb-pulse in young speleothems using a soil carbon continuum model

The 'bomb-pulse' method is a chronological approach to further constrain the age of speleothems that grew between 1950 CE - present. Establishing dependable chronological constraints is crucial for modern calibration studies of speleothems to instrumental climate records, which provides the basis for paleoclimate interpretations. However, a large unknown is how C-14 is transferred from the atmosphere to any individual speleothem owing to the site-specific residence times of organic matter above cave systems. Here, we employ the bomb-pulse method to build chronologies from C-14 measurements in combination with a new unsaturated zone C model which considers C decomposition as a continuum, to better understand unsaturated zone C-14 dynamics. The bomb-pulse curves of eight speleothems from southern Australia in three contrasting climatic regions; the semi-arid Wellington Caves site, the mediterranean Golgotha Cave site and the montane Yarrangobilly Caves site, are investigated. Overall, the modelled C-14 bomb-pulse curves produce excellent fits with measured C-14 speleothem data (r(2) = 0.82-0.99). The C modelling reveals that unsaturated zone C is predominately young at the semi-arid site, with a weighted-mean residence time of 32 years and that tree root respiration is likely an important source of vadose CO2. At the montane site, similar to 39% of C is young (<1 years), but the weighted-mean C ages are older (145-220 years). The mediterranean site has very little contribution from young C (<12%: 0-1 years), with weighted-mean ages between 157 and 245 years, likely due to greater adsorption of organic matter in the upper vadose zone during matrix flow, and remobilisation of C from young syngenetic karst. New end members for low speleothem Dead Carbon Proportion (DCP) are identified (2.19% and 1.65%, respectively) for Australian montane and semi-arid zone speleothems, where oversupply of modern CO2 in the vadose zone leads to lower DCP. It is also demonstrated that DCP can be quite variable over small time scales, that processes may be difficult to untangle and a constant DCP assumption is likely invalid. DCP variability over time is mainly controlled by the changes vadose zone CO2, where vegetation regeneration, wild-fires and karst hydrology play an important role. Crown Copyright (C) 2019 Published by Elsevier Ltd. All rights reserved.