81 Changes in Atmospheric Trace Gases and Aerosols on Glacial-interglacial Time Scales: A Key Challenge for Earth System Models

Wednesday, 30 July 2003
8:00 AM - 12:00 PM

A MODEL STUDY ON HOLOCENE ATMOSPHERIC CO2 VARIATIONS

Paper number: 81-5
Program: SYMPOSIA
Time: 9:10 AM – 9:30 AM

Authors

Abstract

The mechanisms governing atmospheric CO2 and 13CO2 variations during the Holocene are investigated by comparing model results and ice core data. Between 11 and 8 ka BP, atmospheric CO2 fell from 265 to 260 ppm, then it increased again towards the pre-industrial level of about 280 ppm. Indermuehle et al. (1999) suggest that the observed CO2 variations were caused mainly by uptake and subsequent release of terrestrial carbon. An alternative explanation is that the steady CO2 rise after 8 ka is due to the CaCO3 compensation mechanism, responding to the earlier extraction of carbon from the atmosphere by terrestrial uptake during the last deglaciation (Broecker et al., 2001). Uncertainties in the existing 13CO2 data do not allow us to discriminate reliably between the two hypotheses based on observations alone.

The LPJ dynamical global vegetation model coupled to the atmosphere-ocean-sediment carbon cycle component of the Bern CC model was forced by a 21 ka-long time series of snapshot simulations from the Hadley Center climate model (Kaplan et al.,2002). Atmospheric CO2 was prescribed during the deglaciation (21-10.5 ka BP) and simulated thereafter.

The model results matched the Holocene CO2 ice core record within a few ppm. A large part of the CO2 rise after 8 ka BP is explained by the CaCO3 compensation mechanism in response to earlier terrestrial uptake. A terrestrial carbon uptake of 680 GtC is simulated during the transition, in agreement with current reconstructions. Storage increased by 170 GtC during the early Holocene and varied little after 5 ka BP. There was no period of sustained terrestrial carbon loss as required by the hypothesis of Indermuehle et al. Increasing sea surface temperature contributed with about 4 ppm to the modeled CO2 rise after 7 ka BP.

Modeled atmospheric d13C increased by about 0.2 permil from early to mid-Holocene due to biospheric uptake. After 5 ka atmospheric d13C decreased by about 0.03 ppm until 1800 AD. These variations are relatively small compared to the uncertainties in the available ice core data.

In conclusion, the observed atmospheric CO2 variations can be explained by initial terrestrial carbon uptake during the deglaciation and the early Holocene followed by a slow release of CO2 from the ocean due to CaCO3 compensation and sea surface warming.