Holocene Carbon-cycle Dynamics
Based on CO2 Trapped in Ice at Taylor Dome, Antarctica

A. Indermühle*, T. F. Stocker*, F. Joos*, H. Fischer˛, H. J. Smith˛, M. Wahlen˛, B. Deck˛, D. Mastroianni˛, J. Tschumi*, T. Blunier*, R. Meyer* & B. Stauffer*

*  Climate and Environmental Physics, Physics Institute, University of Bern, Sidlerstrasse 5, CH-3012 Bern, Switzerland
˛  Scripps Institution of Oceanography, University of California San Diego, La Jolla, California 92093-0220, USA


Links to Paper Sources:
Published: © 1999, Macmillan Magazines Ltd
Nature, Vol. 398, 11 March 1999

Download Taylor Dome Holocene CO2 Data and Data Description
Download .pdf of article from NOAA Paleoclimatology Publication Site

Get Acrobat Reader Here!! To view this .pdf document you will need to download Adobe Acrobat Reader. Acrobat Reader can be obtained for no cost from Adobe's Web Site or by clicking the icon at right.

Also available are Indermühle et al 60-20K and Smith et al 27-11 KYrBP Taylor Dome CO2 data.


Abstract:
A high-resolution ice-core record of atmospheric CO2 concentration over the Holocene epoch shows that the global carbon cycle has not been in steady state during the past 11,000 years. Analysis of the CO2 concentration and carbon stable-isotope records, using a one-dimensional carbon-cycle model, suggests that changes in terrestrial biomass and sea surface temperature were largely responsible for the observed millennial-scale changes of atmospheric CO2 concentrations.

Figure 1: (below)
Mean CO2 concentrations from ice cores. a, Open diamonds, mean of five to six samples from Taylor Dome ice core measured in Bern. Open triangles, mean of three to six samples measured in San Diego. Open squares, data from Byrd Station ice core5. Error bars, 1 sigma of the mean. The brackets indicate samples where natural artefacts cannot be excluded. Inset, as a but for the interval where the Byrd Station and Taylor Dome data overlap. Using a Monte Carlo method, 1,000 time series of the CO2 record from Byrd Station were generated and filtered in order to take into account the different enclosure processes at Byrd Station and Taylor Dome, respectively. Shaded area, 1 sigma band; solid lines, 2 sigma band. b, Expanded-scale view of the data for the past millennium (shaded area in a). Open diamonds and triangles, mean CO2 concentrations from Taylor Dome; dots, CO2 concentrations from Law Dome4.

Figure 1: Click for larger previewing image.

 

Figure 2:
CO2 concentrations and stable-isotope ratios, Taylor Dome. a, CO2 concentrations. Open diamonds and triangles, mean CO2 concentration. Error bars,1 sigma of the mean. Solid line, spline fit48 of the CO2 results used as input for the single and double deconvolution. This spline fit method acts as a low-pass filter; parameters were selected to obtain a cut-off period of about 8,000 yr which is consistent with the low resolution of the delta13C record. b, Measured and calculated delta13C. Filled circles, delta13C values. Error bars, estimate of the reproducibility of the delta13C values. Solid line and shaded area, input values of delta13C for the Monte Carlo simulation. Dashed line, delta13C calculated with the single deconvolution31 (land-biosphere-only hypothesis, h1; see text).

 

Figure 2: Click for larger previewing image.
Figure 3:
Comparison of our hypotheses. Dashed lines, land-biosphere-only hypothesis (h1). Solid lines, land biota-SST hypothesis (h2). The shaded areas indicate the 1 sigma confidence interval of a Monte Carlo analysis, taking into account the uncertainty of the ice-core data. Dash-dotted lines, land biota-calcite hypothesis (h3). a, Ocean uptake. b, Cumulative biospheric release. c, Required change in SST.

 

Figure 3: Click for larger previewing image.

NOAA NOAA Paleoclimatology Program / WDC-A for Paleoclimatology
29 March 99