The Sarmiento and Sigman groups continue to develop new models and tools for studying ancient changes in carbon fluxes and climate.

Carbon observing system

Much of the Pacala and Sarmiento Groups’ work with the Carbon Observing Network has focused on distinguishing between the natural variability of the carbon cycle and the anthropogenic transient. Importantly, the researchers have thus far been able to use models to identify processes that contribute to the elevated background variability. In applying this process understanding to the detection problem, it will be possible to reduce uncertainty in the detection of anthropogenic change. Work with the Carbon Observing Network will also continue with process-oriented studies of variability that are an integral part of the detection problem.


Temporal Changes in Terrestrial CO2 Fluxes

An important focus for future work will be on using the LM3V dynamical vegetation model (developed jointly by the Pacala Group at Princeton and GFDL) and statistical analysis to determine the cause of the apparent 1990 shift in the terrestrial carbon sink identified in work with the observations. In particular, the researchers will test the hypothesis that the increased terrestrial uptake is due to an observed increase in shortwave radiation and an intensification of the hydrological cycle in the early 1990’s. In addition, they will investigate the impact of other drivers of variability in the carbon cycle such as surface warming, land use change, fertilization of the terrestrial biosphere by increased atmospheric CO2 concentrations, fires, and natural variability of the climate system that occurs on interannual time scales. In support of this effort, work has begun with the LM3V dynamical vegetation model in stand-alone mode to test its sensitivity to a wide range of parameters. Not only does this constitute an important tuning exercise for developing a set of optimal parameter settings, but it also contributes scientifically to understanding ecosystems processes that could be important for climate studies.


Temporal Changes in Air-Sea CO2 Fluxes

Future efforts in the Sarmiento Group will focus on understanding the processes controlling interannual to decadal variations in air-sea fluxes of CO2 over the Southern Ocean. It has long been argued in the ocean carbon modeling community that variability in global air-sea CO2 fluxes is dominated by variations associated with El Niño in the Equatorial Pacific, but recent studies have argued that anthropogenic transients in the circulation of the Southern Ocean have driven significant perturbations in the exchange of carbon between the ocean and atmosphere in this region.

One component of this work will involve using models to understand mechanistically how the ocean carbon cycle may respond to an increase in the strength of the winds over the Southern Ocean in response to anthropogenic climate change. Böning et al. [2007] have argued that the current generation of ocean models used to model the carbon cycle may not adequately capture the eddy response to changes in surface winds over the Southern Ocean, since none of them have been run at eddy-permitting resolution and their parameterization of eddies in these models may not be adequate. To this end, a series of process-focused sensitivity studies will be performed with the MOM4-TOPAZ model forced with CORE reanalysis surface fluxes and a more sophisticated parameterization of eddy processes. These sensitivity studies will have as their goal not only a mechanistic understanding of the response of the ocean carbon cycle to perturbations, but also to an appropriate choice of eddy parameterization for the Southern Ocean.

As part of this work, the group will also develop an improved atmospheric boundary condition for CO2 concentrations for use with ocean modeling studies. Preliminary sensitivity studies indicate that the commonly used CO2 boundary condition used by the ocean modeling community may lead to spurious signals due to inconsistencies in the filtering of the temporal variations in the data.

Finally, in collaboration with Olivier Aumont at the IRD and Laurent Bopp at the LSCE, the team is currently evaluating simulations conducted with the same ocean circulation model (ORCA2- PISCES) that have been forced with two different reanalysis products, that of NCEP and ERA-40. Initial results indicate important differences in the globally integrated air-sea CO2 fluxes between the different model runs. In particular, decadal air-sea CO2 flux variability in the Equatorial Pacific is larger for ERA-40, with these variations sufficiently large to contribute to account for important differences in the globally integrated fluxes. Further analysis indicates that the dynamical variability in the ERA-40-forced run is reflected in a better simulation of sea surface height (SSH) in the Equatorial Pacific than is found for the NCEP-forced run. This serves to highlight that the Equatorial Pacific fluxes (and thereby the global fluxes) in the NOCES suite of NCEP-forced runs may share a bias towards being too small. As this analysis continues, it will be important to investigate the potential implications for estimates of not only global ocean carbon fluxes, but also for terrestrial carbon fluxes.


Land Modeling

Representation of nitrogen cycling in LM3 will continue to be improved by adding phosphorus feedbacks and assessing its impact on regional differences in carbon cycling. Both modeled disturbance experiments and the historical simulations show that feedbacks between carbon and nitrogen cycling in the model are strongest in extra-tropical regions. In contrast, carbon dynamics in tropical forests are almost unaffected by nitrogen feedbacks. However, lowland tropical forests often grow on deeply weathered soil and supply of soil derived nutrients – in particular phosphorus – might be more limiting than nitrogen. The researchers plan to amend the model with phosphorus feedbacks to assess if biomass increases throughout the tropics can be sustained under possible phosphorus constraints, and will further conduct research on how to produce a global data set that specifies external supply from weathering.


Field Measurements

Impacts of Ocean Acidification

The project dealing with the impacts of ocean acidification on phytoplankton will be widened to examine the conditions of nitrogen and phosphorus limitation in addition to that of iron limitation. Continuous cultures have been initiated to quantify the possible effects of increasing CO2 on the C:N ratio of model phytoplankton under N-limited conditions. New laboratory experiments will also examine the effect of decreasing pH on the utilization of organic phosphate. Both of these studies will be complemented by field experiments as described below.


Field Measurements

The ongoing laboratory studies of the impacts of ocean acidification on phytoplankton will be complemented by field experiments taking advantage of vessels of opportunity. In a cruise to the iron-limited region of the North Pacific (Canadian P-line cruises), new field incubations with a redesigned experimental protocol will be carried out in an attempt to quantify the effect of pH on of Fe uptake and Fe use efficiency in situ. On board experiments with varying pCO2 will be performed during cruises off the California coast in N-limited regions of the Pacific. The effect of pH on the use of organic phosphorus by the ambient flora will be tested in the North Atlantic at Bermuda time series station in the North Atlantic.



Extending the Ice Core Record of Paleoclimate

Michael Bender’s group is reconstructing polar climates and the history of atmospheric greenhouse gas concentrations by analyzing trapped gases in ice cores. Ph. D. theses completed in 2007 and 2008 developed the definitive timescales for the Vostok ice core (back to 400 ka, i.e., 400,000 years before present) and the Dome C ice core (back to 800 ka). The researchers’ current focus is to obtain older ice by drilling in locations where ice appears to be preserved for millions of years because of anomalous conditions in the local climates or bedrock topography. In Mullins Valley, in the Dry Valleys region of Antarctica, their collaborator David Marchant has drilled cores reaching 16-26 m depth underlying volcanic ash deposits dated to 1.5 – 4 Ma (million years). Bender and colleagues will date the trapped gases by their Ar isotope composition using a method recently developed in Bender’s lab; the basis is that the abundance of 40Ar relative to other Ar isotopes increases with time because 40Ar in Earth’s crust, produced by 40K radiodecay, continuously outgases to the atmosphere. They will then tackle the problem of greenhouse gas concentrations in these cores. Ar isotope dating also enables a series of related studies (on the age of permafrost, on old ice in other settings, and on the “dirty ice” at the base of the Greenland glacier) which will give new information about climate history and glacier dynamics.


A Proxy for Southern Ocean Winds

The Sarmiento Group will continue to pursue a series of oceanic and atmospheric model simulations that suggests that physical processes over the Southern Ocean may have first order importance in determining the latitudinal gradient of atmospheric radiocarbon. Therefore, atmospheric radiocarbon from direct observations and tree rings could provide valuable insight into processes controlling ocean circulation over the last 1,000 years, a period for which few paleo-proxies are available.

Measurements of radiocarbon in tree rings over the last 1000 years indicate that there was a preindustrial latitudinal gradient of atmospheric radiocarbon of 3.9-4.5‰ and that this gradient has temporal variability on the order of 6‰. Previous efforts to explain the variability in the latitudinal gradient have suggested that it is caused by changes in the frequency of ENSO in the tropics. An alternative hypothesis has been tested here, namely that the natural latitudinal gradient of radiocarbon is primarily controlled by ventilation of the Southern Ocean using fluxes from a suite of models based on the Modular Ocean Model version 3, which are used to force an atmospheric transport model. The results from this suite of simulations suggest that the atmospheric latitudinal gradient of radiocarbon is sensitive to wind stress and wind speed in the Southern Ocean. Increased wind stress in this region leads to greater upwelling of strongly radiocarbon depleted Circumpolar Deep Water (CDW) to the surface, leading to a strong decoupling of the air-sea fluxes of 12CO2 and 14CO2 in this region. This decoupling is due to the fact that the 14DIC in CDW is depleted relative to 12DIC due to radioactive decay. As a result, increased wind speed leads to stronger gas exchange and therefore stronger 14CO2 uptake as well as increased 12CO2 outgassing in this region. Taken together, these effects result in a decrease in the Δ14C of atmospheric CO2.

These dynamical perturbations to the Southern Ocean are much more efficient than dynamical perturbations in the tropics or the North Atlantic in changing the atmospheric radiocarbon signal. Perturbations of amplitudes similar to those of observed decadal trends in Southern Ocean winds for the NCEP reanalysis (about 25%) are sufficient to account for changes in the latitudinal gradients in atmospheric radiocarbon from the tree-ring proxy records over the last 1000 years.

Thus the first stage of this work will focus on demonstrating that pre-anthropogenic atmospheric Δ14C serves as a proxy for past changes in Southern Ocean winds. This first step, which is strongly supported by preliminary results, will emphasize Δ14C as a tracer of physical state variables. The second stage of this work will have as its goal to connect this with variability in the carbon cycle. Radiocarbon has now been included in the biotic ocean biogeochemistry module developed by Eric Galbraith in the Sarmiento group, and this tool will allow us to identify quantitatively the relationship between the responses of 12CO2 and 14CO2 fluxes to wind perturbations over the Southern Ocean. Thus this work that relates Δ14C to the carbon cycle will be conducted in close collaboration with Joe Majkut in his study of the effect of ocean eddies in modulating the response of the carbon cycle to wind perturbations over the Southern Ocean.