The Carbon Science group works to explain historical changes in atmospheric carbon dioxide levels, the nature and variability of carbon sources and sinks, and the feasibility and impacts of large-scale carbon mitigation. During the past 3 years, advances have been made on all fronts.


Interannual Variability of CO2

Michael Bender’s group has been making observations of atmospheric composition to constrain the magnitude of carbon sinks and their interannual variability

Sampler Deployment

Since the inception of the grant, the group has built seven automated sampling devices that are significantly increasing the accuracy of atmospheric argon and oxygen records. Two of the samplers have already been deployed in the Equatorial Pacific and Tasmania, and the remaining five will be deployed around the world in the coming year.

Carbon Sink Estimates

Better measurements allow the group to make increasingly reliable estimates of the size and variability of carbon sinks. The latest estimates for the annual carbon sinks are 1.1 ± 0.7 billion tons on land versus 1.8 ± 0.5 billion tons in the ocean. These refined estimates are slightly lower than, but in rough agreement with, previous calculations. Measurements also indicate that, from 1996 to 2001, almost all interannual variability in CO2 was due to the land sink, and that the land biosphere is a large source of CO2 during El Niño events.

Model-Data Comparison

The new measurements are also allowing the group to test ocean model predictions. Their work has revealed a substantial difference in timing of seasonal ocean gas fluxes between models and observations. This discrepancy in upper ocean dynamics could impact the models’ ability to predict biological productivity and carbon fluxes, so the observations are driving a simultaneous effort to improve simulation of mixed layer physics.

In a second case, new measurements lend some support to a model prediction. Computer simulations have long indicated that the equatorial Pacific ocean should be a net source of oxygen to the atmosphere, emitting more than it absorbs. The group’s preliminary measurements show a peak in atmospheric oxygen concentrations in the region, but its amplitude is much smaller than models predict. A more comprehensive study is currently underway.

Biological productivity estimates

Bender’s group is also currently working on a new technique to take the pulse of micro-organisms at the ocean surface. By measuring biological oxygen fluxes in surface waters and combining that information with oxygen isotope data, the group is currently measuring the variability of marine biological activity in space and time in two oceanographically interesting regions, the equatorial Pacific and the Southern Ocean


CO2 and glacial cycles

Daniel Sigman and colleagues are working to explain the long-term changes in atmospheric CO2 content between ice ages and interglacial periods. The group has proposed that parts of the ocean were more stratified during glacial times, which prevented CO2-rich deep waters from mixing upward and expelling their CO2 into the atmosphere.

Because upwelling of such deep waters also impacts surface nutrient concentrations, the team is analyzing evidence of ancient nitrogen levels in surface waters to test their theory. Since the inception of the CMI grant, they’ve developed a new method for analyzing minute quantities of organic nitrogen from the skeletons of microscopic organisms preserved in deep-sea sediments. The new data indicate that upwelling rates in two critical polar regions, the Antarctic and the Subarctic North Pacific, were indeed lower during the last ice age.

In addition to supporting the hypothesis of polar stratification during ice ages, the new data suggest that natural changes in iron input have had a significant impact on biological productivity in polar ocean waters. This result, which represents geologic validation of the “iron hypothesis,” was completely unexpected on the basis of previous studies of bulk sediments.

Since most short-term computer simulations predict greater stratification with climatic warming, increased stratification in a cold climate runs counter to expectations. The Sigman group is now working to place these short-term model experiments into a broader framework of how climate and ocean circulation interact.

SEM image of diatom fragments from Antarctic deep-sea sediments that are being analyzed to assess the nutrient status of glacial surface waters. (Image ~80 μm across)


The Future of the Ocean Sink

Jorge Sarmiento and colleagues are working on a variety of projects that investigate the ocean’s ability to act as a sink for carbon.

Deep-Sea Injection

One project completed during the first three years is a study of the potential of deep-sea injection as a sink for CO2. Results from a suite of general circulation model simulations indicate that at least 70% and up to 93% of the carbon injected below 3000 meters water depth remains in the oceans after 500 years. These results indicate that deep-sea injection would be effective in storing CO2, but possible environmental impacts raised by other researchers need to be addressed before deep-sea injection could be implemented.

Ocean Fertilization

Another early study investigated the possibility of fertilizing the ocean to enhance phytoplankton uptake of CO2 and the delivery of this carbon to the deep ocean via an increased flux of organic matter. The group’s model simulations suggest that only 2-10% of additional carbon flux to the deep ocean would come from the atmosphere, and that fertilization might eventually decrease biological production and impact fisheries. It thus seems that ocean fertilization would be less effective than originally envisioned and could ultimately backfire by decreasing ocean biological productivity in the future.

Climate Change Impacts on Ocean Chemistry & Biology

A longer timescale study on ocean chemistry suggests that the ocean sink of carbon dioxide will shrink in the future, and that this decline should be considered in specifying leakage limits for underground CO2 storage. The group’s ocean model simulations indicate that the carbonate buffer that now allows the ocean to absorb large quantities of atmospheric carbon will become saturated in a few centuries.

A study carried out this year provides new estimates of recent ocean uptake of carbon dioxide. The team analyzed a suite of global ocean models and selected only those that correctly predicted distributions of CFC’s and radiocarbon derived from the atmosphere. Using these models, the team calculated the size the ocean carbon dioxide sink in the 80’s and 90’s. Their estimates confirm previous observational work and narrow the uncertainty in ocean CO2 uptake.

Also in this year, the group has linked chlorophyll and primary production models to a 3-D ocean model to estimate changes in biological productivity with climatic warming. Their results indicate that the total warming of the climate between the beginning of the Industrial Revolution to 2050 could be accompanied by a 0.7 to 8 % increase in primary productivity. Even in the most extreme case considered, this change would have only a modest effect on export production and atmospheric CO2 levels.


The Future of the Terrestrial Sink

Steve Pacala’s group has been investigating the human influence on land-based carbon sinks and the impacts of renewable energy production on the environment.

Land Use vs. Fertilization

Since the inception of the grant, the group has been investigating the source of the large land-based carbon sink in the coterminous United States in the 20th century. There are two competing theories for the origin of the large sink. One proposes that increasing levels CO2 had a fertilizing effect on land plants, increasing growth rates and causing heightened carbon uptake. The second suggests that land-use changes encouraged regrowth of forests that stored large amounts of CO2. The group’s most recent data show that forest growth rates in Wisconsin have actually decreased with rising carbon dioxide concentrations since the 1960’s, suggesting that the big U.S. carbon sink is more likely due to land use changes than fertilization.

Because the terrestrial biosphere is largely recovered from past land use, this finding indicates that the terrestrial carbon sink is likely to shrink in the future. Such a decrease would require steeper cuts in emissions than currently anticipated to stabilize atmospheric carbon dioxide levels.


Impacts of Renewable Energy

Pacala’s group has also been studying potential impacts of large-scale renewable energy production on the environment.

Impacts of Wind Turbines

A preliminary study incorporating the effects of wind turbines on climate suggests that wind energy production would have minimal impact on global average surface temperature, even if wind power was scaled up to produce 20 trillion watts of electric power. On regional scales, however, simulations indicate that temperature changes of up to half a degree Celsius and small but significant changes in precipitation might occur.

Ozone precursors

Another project on renewables analyzed the impacts of forest plantations on air quality. Previous work had indicated that certain trees are a significant source of volatile organic carbon species (VOC’s) that contribute to tropospheric ozone formation and smog. The new work reveals that while reforestation by natural succession tends to decrease forest VOC emission, “managed” reforestation increases the level of VOC’s released to the atmosphere and has a negative impact on air quality. The group’s study suggests that changes in current forest management practices to allow more natural succession could bring significant benefits for air quality and human health.


Earth System Model

In the first three years of the grant, CMI researchers partnered with the Geophysical Dynamics Laboratory in developing a state-of-the-art climate model. This year, the team succeeded in linking completely new components for the atmosphere, ocean, sea ice, land, terrestrial biosphere and ocean biosphere. CMI researchers led development of the terrestrial and ocean ecosystem components of the ESM, which are at the leading edge of those available in coupled climate models. Simulations incorporating these modules will be carried out in Year 4.

The new model has a climate comparable to those of the best coupled models previously available, and it is the first coupled model of its type to produce a realistic El Niño event in a long model run. This new tool should increase confidence in future climate predictions, and substantially increase our insight into carbon cycle processes.