The Sarmiento and Pacala Groups, joined by new CMI members Lars Hedin and David Medvigy and their colleagues, are working to continuously improve a “Carbon Observing System.” The effort uses observational data and models to monitor both short and long timescale changes in the land and ocean carbon sinks, and to provide predictions for the future.


Understanding temporal shifts in terrestrial uptake of atmospheric CO2

In previous work, Jorge Sarmiento and colleagues detected an abrupt increase in the net land uptake of CO2 of approximately 0.8 Pg C/yr in 1988. In this study, the net land flux was 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. A suite of ocean models was used to represent uncertainties in the magnitude and temporal variability of oceanic uptake.

Given the importance of such a shift to the climate system and carbon cycle, a better understanding of it is necessary. In collaboration with researchers at UCLA as part of a study supported by a NASA Carbon Cycle Science grant, the Sarmiento Group has analyzed satellite data of gross primary production, terrestrial respiration and net ecosystem exchange fluxes for the region north of 45°N. In preliminary analyses, the group has detected a corresponding shift in 1987 in annual and spring fluxes of gross primary production and respiration in this region. As a visual support, the fluxes and cumulative fluxes are presented in Figure 13. The shifts detected in satellite data agree with the shift in the net land uptake of carbon, but the magnitude is smaller (the net shift is approximately 0.2 Pg C/yr). Through partial support from NASA Carbon Cycle Science, Claudie Beaulieu in the Sarmiento Group is applying change point detection techniques to other key data sets and model outputs. An improved understanding of this shift and its causes is important for the prediction of future shifts in the carbon cycle.

Figure 13. Total fluxes of gross primary production (GPP) and terrestrial respiration (TER) above 45° North anomalies (first row) and cumulative anomalies (second row). The change of slope in 1987 indicates the possibility for a shift. The shift in 1997 is probably due to the change of satellite covering and was accounted for in the analysis.

Understanding CO2 fertilization of the biosphere

Ray Dybzinski and Stephen Pacala are in the middle of a new project on the future of CO2 fertilization of the biosphere. An ongoing mystery has been why some CO2 enrichment experiments result in a large carbon sink, while others do not. Existing global models predict that nitrogen (N) limitation will limit CO2 fertilization, whereas water limitation will enhance CO2 fertilization. However, recent data implies that the opposite is true. The researchers now think they can explain why.

New forest models developed by the Pacala Group predict that N-limited forests will become sinks under CO2 fertilization, but water-limited forests will not. The reason is game-theoretic – because root systems intermingle, both nitrogen and water are best viewed as commons, and so the economics of plant competition is the economics of commons exploitation. Physiologically, water is best thought of as a fuel – it must be evaporated and lost to run photosynthetic machinery. Nitrogen is a structural material, used to make photosynthetic machinery. It is the difference between a fuel and a structural material that makes water-limited plants waste extra production from CO2 fertilization on new short-lived fine roots that do not create a large or long-lived carbon sink. In contrast, nitrogen-limited plants devote the new production to wood, which creates a large carbon sink. We have a large amount of empirical support for this explanation. It is important because it means that the CO2 fertilization sink should continue. If it were to fail, as it does when water is limiting, then the amount of mitigation necessary to stabilize at 500 or 550 ppm would double.

In another project investigating the influence of nitrogen cycling on CO2 fertilization, postdoctoral researcher Duncan Menge initiated a set of experiments to determine the degree of regulation in many types of nitrogen-fixing plants. Plants that adjust nitrogen fixation to meet demand remove more carbon dioxide from the atmosphere, whereas plants that fix nitrogen at the same rate regardless of demand remove less carbon dioxide. Furthermore, plants that always fix at the same rate produce very nitrogen-rich conditions that lead to emissions of the potent greenhouse gas nitrous oxide from soils. Early results suggest that all herbaceous nitrogen fixers from temperate habitats fix nitrogen at the same rate regardless of demand, limiting the CO2 fertilization effect outside the tropics.


The future of the Amazon as a carbon sink

One of the most striking results from models of the Earth System’s response to CO2 pertains to tropical forest ecosystems, and, in particular, the rainforests of South and Central America. If atmospheric CO2 concentrations continue to increase, some of these computer models predict that increased temperatures and a modified soil moisture regime would cause a catastrophic collapse of the Amazon by the mid-to-late 21st century. Savanna ecosystems would replace rainforests, with possibly severe implications for the 40,000 plant species that currently inhabit the Amazon. Furthermore, 120 billion tons of carbon, currently stored in rainforest biomass, could be released to the atmosphere.

There has been vigorous scientific debate as to whether such outcomes are realistic. A major problem is that models have received limited validation against field data in the tropics, as ecological field studies sample much smaller spatial scales than those simulated by global climate models, and so it is difficult to use these data to evaluate models. To address this scale mismatch, David Medvigy and colleagues have developed a new type of structure ecosystem model that is capable of being directly validated by small-scale field studies, including those being carried out by Lars Hedin’s research group. This study thus represents a collaborative effort between modelers and field researchers.

This approach has particular bearing on the following questions: (i) Will constraints by nutrients or water diminish the strength of the tropical forest carbon sink over time? (ii) Will different tropical forest species respond differently to climate change? If so, does species diversity affect the resilience of tropical forests to environmental perturbances? This year, the Medvigy and Hedin Groups will address these questions by developing an empirically-tested framework for studying carbon-nutrient interactions for different tropical species, and by using this framework to evaluate the susceptibility of the tropical forest carbon sink to climate change.


Quantifying carbon sinks on managed lands

The Intergovernmental Panel on Climate Change (IPCC) defined Land Use fluxes as greenhouse gas emissions and removals on managed lands (i.e. “managed land proxy”), including direct anthropogenic effects (e.g., agricultural conversion and wood harvesting) and indirect anthropogenic effects (e.g., CO2 and climate change). But in the ongoing fifth Climate Model Intercomparison Project (CMIP5), no global Earth System Model (ESM) has been capable of calculating this flux for technical reasons, which obviously impedes the analysis of the coupled climate-carbon cycle system and constrains development of carbon policy grounded in science.

This year, Senior Researcher Elena Shevliakova implemented a modeling approach addressing such limitations and used NOAA/GFDL ESMs to demonstrate the ongoing carbon sink on managed lands (since the 1990’s) and unmanaged lands (throughout the 20th century and presently). The sink on managed lands is from substantial re-growth of the secondary forests (1.5 Gt/yr) in both tropics and temperate regions. The ongoing sink on unmanaged lands is substantially smaller in this model (1.6 Gt/yr) than estimated in other studies. The present day warming would be significantly higher if both managed and unmanaged lands had not became a sink of CO2 in the second part of the 20th century.