The correlation between interannual variations in carbon dioxide uptake by vegetation and interannual variations in precipitation as simulated by a coupled model of climate and terrestrial ecosystems. Carbon uptake is enhanced in wet years (red areas) since most of the world's vegetation is water-limited. Only in cold regions do thermal limitations dominate (blue areas).

Starley Thompson, Philip Duffy, Jose Milovich, Bala Govindasamy, Peter Eltgroth, and Art Mirin, Lawrence Livermore National Laboratory
Aaron Herrnstein, University of California, Davis
Christine Delire, University of Wisconsin

Research Objectives
Our overall goal is to produce the first comprehensive coupled climate/carbon cycle model in the U.S. This research will allow better predictions of future climate, because feedback effects of climate change on absorption of carbon by the ocean and terrestrial biosphere—which are ignored in present U.S. climate models—will be taken into account. This model will be more useful to policymakers than present climate models because it will use CO₂ emission rates, rather than atmospheric CO₂ concentrations, as the fundamental input.

Computational Approach
All the computational models we employ use a Eulerian discretization approach. The POP ocean model uses finite differences on a structured 3D mesh. Its discretization allows the use of any locally orthogonal horizontal grid, which, for example, allows the Arctic ocean to be resolved without the standard problems of convergence of meridians at the North Pole. The CCM3 atmospheric model implements a spherical harmonics-based spectral formulation with a latitude/longitude transform grid and uses a terrain-following coordinate in the vertical dimension. We are currently using two ocean biogeochemistry (OBGC) models, the LLNL OBGC model and the PICSES OBGC model, as well as the IBIS terrestrial biospheric model.

Accomplishments
High-resolution global climate simulations: We are running three experimental simulations at very high spatial resolution. The first is a simulation of the period 1979-1989 performed at T239 spectral truncation (50 km resolution). This is the highest-resolution long-term global climate simulation ever attempted. The second is a pair of simulations at slightly coarser resolution (T170 truncation; 75 km resolution) for a doubled-CO₂ climate and a control simulation of the present climate. These are the highest-resolution runs ever performed of anthropogenic climate change. Preliminary results show that in simulations of the present climate, some features become more realistic as spatial resolution becomes finer; but the simulated increase in temperature in response to a given increase in atmospheric CO₂ does not seem to be sensitive to model resolution.

Coupled climate and terrestrial biosphere simulations: The Community Climate Model 3 (CCM3) coupled to the Integrated Biosphere Simulator (IBIS 2) was used to perform a 16-member ensemble of present day simulations with observed sea surface temperatures (SSTs) for the period 1979-1992. Simulated interannual variations in terrestrial carbon uptake have good positive correlation with inferred uptake from observations. Nearly 65% of interannual variability is caused by unforced climate variability not related to variations in SSTs. This unforced interannual variability in the uptake is caused primarily by the unforced variability in net primary productivity, which is driven by variability in precipitation and temperature.

Significance
A comprehensive coupled climate/carbon cycle model will allow us to directly assess the climatic impact of specified rates of burning of fossil fuels, because it will use CO₂ emission rates (not atmospheric CO₂ concentrations) as its fundamental input variable. The model will also allow us to improve our fundamental understanding of the coupling between climate change and Earth's carbon cycle, since it is known that both the marine and terrestrial components of the carbon cycle are sensitive to climate change.

Publications
J. C. Bergengren, S. L. Thompson, D. Pollard, and R. M. DeConto, "Modeling global climate-vegetation interactions in a doubled CO₂ world," Climatic Change 50, 31 (2001).

B. Govindasamy, P. B. Duffy, and K. Caldeira, "Land use change and Northern Hemisphere cooling," Geophys. Res. Lett. 28, 291 (2001).

P. B. Duffy, M. Eby, and A. J. Weaver, "Climate model simulations of effects of increased atmospheric CO₂ and loss of sea ice on ocean salinity and tracer uptake," J. Climate (in press).

http://www-pcmdi.llnl.gov/cccm/