The Sarmiento and Pacala Groups have continued the development of a “Carbon Observing System” that incorporates observation data from various origins along with different kinds of models (atmosphere, ocean and land) to monitor both short and long time scale changes in the carbon cycle, and to provide predictions for the future. An important component of this work consists of identifying the anthropogenic transient signal against a background of elevated natural (interannual to decadal) variability in the carbon cycle using both models and observations. Recent analyses of observations as well as models forced with the observations suggest that there have been significant changes in oceanic circulation and a reduction in the oceanic carbon sink over the past 2 to 3 decades, and that the land biosphere may have undergone a major increase in uptake starting around 1990.


Impacts of a Doubling of CO2

A collaboration between the Pacala Group and GFDL has produced a result that provides important new understanding of the implications of doubling the pre-industrial concentration of atmospheric CO2 (572 ppm), a level that many view as an appropriate target for mitigation or as a likely outcome of an attempt to cap the concentration at 450 or 500 ppm. The new finding is made possible by a coupled ocean-land-atmosphere earth system model that is able to calculate the equilibrium climate associated with any given level of atmospheric CO2. Note that virtually all previous calculations with earth system models have been limited to transient runs in which the climate is still changing at the end of the run. The result is contained in a manuscript that has been submitted for publication: “Uncertainty in the Land-Carbon Uptake due to CO2 Fertilization under Climate Change,” by Elena Shevliakova, Ronald J. Stouffer, Lori T. Sentman, Stephen W. Pacala, Michael J. Spelman, and Sergey Malyshev.

The study focuses specifically on uncertainty associated with the magnitude of CO2 fertilization of the biosphere and contains a review of published empirical evidence that CO2 fertilization may not deliver a sustained terrestrial sink because of several mechanisms of down-regulation. Evidence of the possible failure of CO2 fertilization has been described in several previous CMI annual reports.

With sustained CO2 fertilization, the model predicts that a doubling of the pre-industrial concentration of CO2 would create a terrestrial sink that would eventually absorb over 200 Gt of carbon. This result is similar in magnitude to the values produced by the transient runs reported in the latest IPCC report. However, if CO2 fertilization is down-regulated, then the model predicts that the biosphere will emit over 400 Gt of carbon to the atmosphere, mainly from tropical rainforests and arctic soils. The difference between a gains and losses in these two runs is nearly as large as the total CO2 currently in the atmosphere.

If this extra CO2 is emitted sufficiently slowly, then the oceans will take up most of it. But if it is emitted quickly, then it would require that humanity either mitigate an additional 26 wedges worth of CO2, or allow the atmospheric concentration to exceed a doubling, which would cause more climate change and induce still more emissions from the biosphere. This implies that it might be impossible to maintain the concentration at a doubling without revolutionary new mitigation technology. The group is now planning to attempt the transient calculation to determine how fast the extra carbon would be emitted.


Temporal Shifts in the Sources and Sinks of Atmospheric CO2

The Sarmiento Group has discovered an apparent increase in the net land uptake of CO2 of 0.9 (0.8 to 1.0) Pg C yr-1 after 1990/91 based on a novel analysis of the atmospheric CO2 growth rate. In this approach, the net land flux is estimated as the balance of relatively well-known components of the carbon budget: fossil fuel emissions, the observed growth rate in the atmosphere, and the oceanic uptake from state of the art ocean models. Due to substantial uncertainties in the temporal variability of oceanic uptake, a suite of ocean models was used.

The predominant signal in the inferred net land flux (Figure 15) is the very large interannual variability. However, using a low pass Butterworth filter, it is possible to resolve a clear shift towards greater land uptake after 1990/91 than before it. The net land sink estimated using the six time-varying ocean models increases by an average of 0.9 (0.8 to 1.1) Pg C yr-1 after 1990/91. Such an acceleration of the net land carbon uptake had been noted previously for the decade of the 1990s compared with the decade of the 1980’s; and this post-1990/91 net land uptake estimate is consistent within uncertainty with the atmospheric oxygen based estimate of 0.51 ± 0.74 Pg C yr-1 for the period of 1993 to 2003. However, this analysis suggests a much greater persistence in time of this signal, including that the major Pinatubo anomaly of 1991 to 1993 can account for only 0.27 (0.20 to 0.31) Pg C yr-1 of the increase in long-term averages. An analysis of the atmospheric growth rate shows that this post-1990/91 shift in the net land uptake appears to find expression as modulations of the amplitude of the variability itself rather than a simple shift of the mean state.

One possible explanation for this increased terrestrial uptake is an observed increase in shortwave radiation accompanied by an intensification of the hydrological cycle in the early 1990’s. The group’s plan to test this hypothesis is described below in “Future Plans.”

Figure 15. Terrestrial carbon uptake inferred from estimates of fossil fuels, the observed atmospheric growth rate of CO2, and an ocean model [Le Quere et al., 2007], smoothed with a 12 month running mean (gray), a five year low pass Butterworth filter (red), and a ten year low pass Butterworth filter (blue). Positive values indicate uptake and negative values indicate emissions.

Nutrient feedbacks on the Terrestrial Carbon Cycle

Stefan Gerber, in collaboration with Lars Hedin, Michael Oppenheimer and Steve Pacala, has completed implementation of nitrogen dynamics in the Princeton-GFDL LM3V land model. The new model resolves processes, dynamics and feedbacks that have not previously been captured and represents a substantial improvement over all previous and existing dynamic global vegetation models. Early simulations with the model suggest that nitrogen limitation acts as brake on the CO2 fertilization effect, and that feedbacks between carbon and nitrogen cycling are strongest in extra-tropical regions.

An urgent question with respect to global change is whether terrestrial systems may act as a significant sink for anthropogenic carbon via a CO2 fertilization response. To test whether nitrogen limitation would restrict the biosphere’s ability to sequester carbon in such a scenario, Gerber and colleagues performed simulations from 1500 to 2000 AD that included past changes in atmospheric CO2, land-use transitions and recent climate change. The model with the new nitrogen formulation predicts a terrestrial carbon source of ~11 GtC over the last 150 years (Figure 16), which is in broad agreement with estimations based on ocean inventory and budgeting approaches. Conversely, when feedbacks between terrestrial carbon and nitrogen cycling were neglected, the model predicted a cumulative sink of about 15 GtC. Sensitivity testing showed that nitrogen cycling hampers carbon sequestration in the model by preventing uptake under CO2 increase but also reduces carbon losses during land-use transitions.

Figure 16. Cumulative change in terrestrial carbon inventory over the past 500 years as simulated for different scenarios (CO2, LU, CO2&LU, Dep). CO2: changes in atmospheric CO2 are accounted for to simulate potential vegetation. LU: Scenario where land‐use transitions only were considered while CO2 was kept constant at preindustrial levels. CO2&LU: Both CO2 increase and LU transitions are accounted for. Dep: Nitrogen deposition is held constant at natural background levels. Solid lines represent simulations carried out with the coupled carbon‐nitrogen land model, while the dashed lines show model realizations where carbon‐nitrogen feedbacks were neglected.

The model is also able to reproduce global patterns of terrestrial carbon storage over individual decades. While the terrestrial biosphere appeared to be a source of about 0.5 GtC yr-1 over most of the historic period, the team found a shift from source to sink around 1970. Comparison of simulated carbon fluxes for the 1980s and 1990s against results from budgeting approaches and inverse modeling results shows good agreement for both decades. Increased carbon uptake in the 1990s appears to occur mostly in the tropics, while extra-tropical regions (mostly Northern Hemisphere) sequester carbon at a similar rate in both decades.


A Wintertime Window for CO2 Uptake in the North Pacific

To evaluate the pathways and timescales associated with the uptake of anthropogenic CO2 over the North Pacific, the Sarmiento Group forced an ocean model with NCEP reanalysis fluxes over 1948-2003. The model revealed that there are two principal regions of uptake, the first along a band between 35–45°N and 140–180°E, and the second along a band between 10-20°N and between 120°W and 180°E (Figure 17a). For both of these regions, the dominant timescale of variability in uptake is seasonal, with maximum uptake occurring during winter and uptake being close to zero or slightly negative during summer when integrated over the basin.

The model results indicate a decadal trend toward increased uptake of anthropogenic CO2 (Figure 17), but the trend is due largely to modulations of the uptake maximum in winter. This implies that, for detection of anthropogenic changes in CO2 uptake, in situ measurements will need to resolve the seasonal cycle in order to capture decadal trends in ΔpCO2. As uptake of anthropogenic CO2 occurs preferentially during winter, observationally-based estimates which do not resolve the full seasonal cycle may results in underestimates of the rate of uptake of anthropogenic CO2.

Figure 17. Uptake of anthropogenic carbon over the North Pacific for the ocean model. (a) Air‐sea CO2 fluxes integrated over 1993‐2001 (moles/m2) with positive values indicating a flux into the ocean. The figure reveals a clear uptake maximum in the Kuroshio Extension region to the east of Japan, and a secondary maximum to the southeast of Hawaii. (b) The temporal behavior of the basin‐integrated uptake of anthropogenic CO2 (Pg C/yr) over the North Pacific, with maximum values in winter and minimum in summer.


Using Satellite Measurements to Detect Anthropogenic Change in Ocean Chemistry

One of the principal challenges in detecting anthropogenic change with a Carbon Observing System is that there is elevated natural variability in the Earth’s climate cycle. As a result, detection must be understood in part as a signal-to-noise problem. A key component of separating the anthropogenic perturbation signal from the natural variability signal involves developing a mechanistic understanding of the natural background variability in the carbon cycle. To this end, observations and ocean models have been used to identify mechanisms driving large seasonal to interannual variations in dissolved inorganic carbon (DIC) and dissolved oxygen (O2) in the upper ocean. The Sarmiento Group began with observations linking variations in upper ocean DIC and O2 inventories with changes in the physical state of the ocean. Models were subsequently used to address the extent to which the relationships derived from short-timescale (6 months to 2 years) repeat measurements are representative of variations over larger spatial and temporal scales.

Figure 18. Changes in measured column inventory of DIC (black, mol/m2), O2 (green, mol/m2), and SSH (red, cm) from TOPEX altimetry data along 80°E in the Indian Ocean between March 1995 and September 1995. The close correspondence between the changes in SSH and the changes in DIC inventories indicates that local convergence and divergence of water associated with the passage of ocean (Rossby) waves is driving the DIC inventory variations.

The main new result with this work is that local redistribution of water associated with the passage of planetary waves in the upper ocean (in particular through the action of baroclinic Rossby waves) can make a first-order contribution to the natural variability of DIC and O2 in the upper ocean. This results in a close correspondence between (natural) DIC and O2 column inventory variations and sea surface height (SSH) variations over much of the ocean (Figure 18).

The close correspondence between SSH and both DIC and O2 column inventories for many regions suggests that SSH changes (inferred from satellite altimetry) may prove useful in reducing uncertainty in separating natural and anthropogenic DIC signals (using measurements from CLIVAR’s CO2/Repeat Hydrography program).