Documentation for Abrupt Change in Atmospheric CO2 During the Last Ice Age, Version 1

Back to Top

Collapse All / Open All

During the last glacial period atmospheric carbon dioxide and temperature in Antarctica varied in a similar fashion on millennial time scales, but previous work indicates that these changes were gradual. In a detailed analysis of one event, we now find that approximately half of the carbon dioxide increase that occurred during the 1500-year cold period between Dansgaard-Oeschger (DO) Events 8 and 9 happened rapidly, over less than two centuries. This rise in carbon dioxide was synchronous with, or slightly later than, a rapid increase of Antarctic temperature inferred from stable isotopes.

Detailed Data Description

The Siple Dome and Byrd ice core records cover the time period from 37 to ~47 ka, during which Antarctica experienced two major warming events (A1 and A2) and several abrupt warming/cooling DO events occurred in Greenland. To place the records on the same chronology, we synchronized the age scales using existing and new methane (CH4) data, assuming CH4 concentration variations in Greenland and Antarctica were synchronous (Blunier and Brook 2001). Refer to Figure 1. Although measurements of carbon dioxide (CO2) in Greenland ice cores might provide a more direct way to compare Greenland climate variations with CO2 changes, reconstruction of the CO2 history from Greenland ice cores is difficult due to high levels of impurities in ice (Stauffer 2006).

We obtained sub-centennial CO2data for the period from about 40 to 38 ka, refer to Figure 1, corresponding to the Greenland stadial between DO Events 8 and 9, and the time period of the A1 warming in Antarctica (Blunier and Brook 2001) or Antarctic Isotope Maximum 8 (AIM8) (EPICA Community Members, 2006). Heinrich Event 4, an abrupt glacial discharge event in the North Atlantic, occurred in the time interval and the stadial is referred to as Heinrich Stadial 4 (Skinner and Elderfield, 2007). In general, CO2 strongly covaries with the Antarctic stable isotope records on centennial time scales over this period. After the DO9 warming event in Greenland, both CO2 and δ18Oice in the Byrd and Siple Dome records started to increase at ~40.0 ka. The initial CO2 rise was somewhat gradual, but at 39.6 ka, a 10 ppm jump occurred over ~150 years in both Siple and Byrd records. Following that jump CO2 levels show two oscillations of smaller amplitude (5 ppm) before an additional abrupt increase at ~38.3 ka, synchronous with DO8 warming in Greenland. After the abrupt DO8 warming in Greenland, CO2 remained high for 700 to 800 years while Antarctica cooled as shown in both the Byrd and Siple Dome records, confirming the decoupling of CO2 and Antarctic temperature proxies following Antarctic warming observed in previous Byrd ice core results (Ahn and Brook 2008). We note that the CO2 in Byrd is uniformly 3 to 7 ppm higher than in Siple Dome for the studied age interval, which we believe is probably related to the long duration of storage for the Byrd core drilled in 1968, but the timing and patterns of CO2 change are similar. High resolution Ca2+ and non-sea-salt Ca (nssCa2+) concentration records from the Siple Dome ice core (Mayewski et al., 2009) do not show any significant correlation with the abrupt CO2 rise, indicating that the abrupt CO2 rise is not likely produced by carbonate-acid reaction in the ice. Similar data are not available for Byrd. Although the abrupt changes in our data are defined by only a few data points, they are reproducible in the two ice cores, and we further note that each plotted point is the mean of several replicate samples. The replicate Byrd ice samples were generally analyzed on different days over six months.

Note: Refer to Table 1 for a legend describing the letters A through J and the Yellow Box in Figure 1.

fig 1
Figure 1. Synchronization of Ice Core Records and Comparison with Marine Sediment Records

Figure One Legend

Table 1. Legend for Figure 1
Letter Reference
Data Description
Greenland isotopic temperature (Blunier and Brook 2001). Blue numbers indicate the timing of DO events. Red arrows are age tie points.
Byrd ice core CH4 records (Blunier and Brook 2001). Yellow circles are new data for this study.
Byrd ice core CO2 records (Ahn and Brook 2008). Red circles are new data for this study.
Byrd ice core δ18Oice temperature proxy (Blunier and Brook 2001).
Siple Dome ice core CH4 records (Brook et al. 2005). Yellow circles are new data for this study.
Siple Dome ice core CO2 records for this study.
Siple Dome δDice for a temperature proxy (Brook et al. 2005).
Benthic foraminifera δ13C from Iberian margin sediment core (Margari et al. 2010). The δ13C is used for proxies for deep water sources that are indicated with a dark yellow arrow, NADW, North Atlantic Deep Water.
Opal flux in the Southern Ocean as a proxy for upwelling (Anderson et al.). AABW, Antarctic Bottom Water.
Speleothem δ18O records from Chinese Hulu (green and gray) [Wang et al.] and Pacupahuain Caves (blue) [Kanner et al.], proxy for precipitation. More negative δ18O indicates increased rain out. Proxy ages are synchronized with GISP2 δ18O at age tie points (red arrows). Ages between the tie points are linearly interpolated.
Yellow Box
Indicates the time interval between two abrupt DO 8 and 9 warming events, for which high-resolution CO2 records were obtained. Dotted vertical line points timing of abrupt CO2 changes.

Data are provided in Microsoft Excel (.xlsx) format.

Background color on
File and Directory Structure

Data are available on the FTP site in the directory. Within this directory, there is one Microsoft Excel file: Ahn_et_al_2012_GRL_data_NSIDC_20130412.xlsx. Within this excel file, there are six worksheets:

  • Byrd_CO2
  • Sipledome_CO2
  • Sipledome_CH4
  • Byrd CH4
  • Taylor Dome CH4
  • GISP2_CH4.
Background color on
File Size

42 KB

Background color on

42 KB

Background color on
Spatial Coverage

Siple Dome Ice Core:
Southernmost Latitude: -81.66°
Northernmost Latitude: -81.66°
Westernmost Longitude: -148.82°
Easternmost Longitude: -148.82°

Byrd Ice Core:
Southernmost Latitude: -80.01°
Northernmost Latitude: -80.01°
Westernmost Longitude: -119.83°
Easternmost Longitude: -119.83°

Spatial Resolution

Siple Dome Ice Core:
Min Depth: 819 m
Max Depth: 905 m

Byrd Ice Core:
Min Depth: 1685.1 m
Max Depth:1814.5 m

Background color on
Temporal Coverage

Paleo Temporal Coverage: 37 ka to 47 ka
Phanerozoic  Cenozoic  Quaternary

Temporal Coverage: The data were collected between 2009-01-01 to 2012-12-31.

Background color on
Parameter or Variable

Ice Core Records - Carbon Dioxide (CO2)
Ice Core Records - Methane (CH4)

Sample Data Record

The following Sample Data Record is of the Byrd_CO2 data file. Note that this data record is a .xlsx file, and this is Worksheet Number 1 in the Excel file.

sample data

Background color on

Software and Tools

Background color on

Data Acquisition and Processing

Theory of Measurements

Details of CO2 analysis at Oregon State University (OSU) are described in Ahn et al. (2009). Siple Dome ice samples were analyzed from 58 depths from 819-905 m depth (37.1-52.4 ka) and 105 Byrd ice samples from 40 depths from 1685.1-1814.5 m (36.8-42.9 ka). Carbon dioxide concentrations are reported on the WMOX2007 CO2 mole fraction scale. We utilized nitrogen isotope data from Brook et al. (2005)Sowers et al. (1992), and Bender et al. (1995) for corrections of 0.8-0.9 and 1.0-1.5 ppm for gravitational fractionation in Siple Dome and Byrd ice, respectively.

CH4 analysis was performed at OSU using methods described by Mitchell et al. (2011). We analyzed 62 new samples from 40 depths of the Siple Dome core, 87 samples from 48 depths of the Byrd core, 56 samples from 34 depths from the Taylor Dome core, and 88 samples from 44 depths of the GISP2 core. Data are reported on the NOAA04 CH4 concentration scale (Dlugokencky et al. 2005).

Background color on

References and Related Publications

Contacts and Acknowledgments

Edward J. Brook
Oregon State University
Department of Geosciences
104 Wilkinson Hall
Corvallis, OR 97331-5506 

Jinho Ahn
School of Earth and Environmental Sciences
Seoul National University
599 Kwanak-ro, Kwanak-gu, Seoul 151742, South Korea


This research was supported by NSF OPP Grant Number 0944764.

Document Information


August 2013