In the first three years of the grant, CMI researchers partnered with the Geophysical Dynamics Laboratory to develop a state-of-the-art climate model. CMI researchers led development of the terrestrial and ocean ecosystem components, which are at the leading edge of those available in coupled climate models.


Dynamic global land model

A major achievement of Steve Pacala’s team has been the development of a new dynamic global land model, LM3, in collaboration with GFDL and USGS scientists. LM3 simulates the dynamics of vegetation and soil carbon pools, the state of the hydrological cycle, the exchange of water, CO2, and energy exchange between land, atmosphere and ocean. LM3 is designed to study biosphere-atmosphere interactions and feedbacks: effects of changes in vegetation and soil functioning on the atmospheric physics and chemistry, and, reciprocally, the implication of changing climate and CO2 concentration on the land surface, and the implications of direct anthropogenic changes (i.e. land use) on the fate of climate and the global carbon cycle.

The land model currently does not simulate the cycling of nitrogen or phosphate, but understanding the cycling of nutrients like nitrogen in natural and managed ecosystems is important for assessing the impact of global change on areas of vital environmental and economic concern including forest carbon balances, agricultural productivity, eutrophication of coastal ecosystems and nitrous oxide emissions to the atmosphere. The group has now developed a strategy for adding nitrogen cycling to the land model with both short-lived and long-lived soil organic matter pools that will be implemented in the near future.

Understanding the role of boreal forests is also key to accurately simulating the future behavior of carbon pools in the context of global warming, and a major factor that influences the structure and carbon dynamics of the boreal forest is fire. Cyril Crevoisier has developed an empirically-based fire model for boreal forest that will be implemented in LM3. The model relies on information from various origins to predict ignition: climate (air temperature, air humidity, precipitation), human impact (distance to the nearest roads and to the populated zones) and history (last fires in the area). The first results obtained are encouraging and give an ignition prediction error of about 20%. Progress is also being made on developing and analyzing a model for fire dynamics and suppression.


Ocean biology and chemistry modules

Over the course of the CMI grant, Jorge Sarmiento’s group has implemented a new ocean carbon model in the GFDL Earth System Model and two efforts are underway to develop alternative models. One model is based on observations of dissolved nutrient, carbon, silicic acid, and alkalinity in the ocean to infer the production of organic matter, opal, and CaCO3. The second model uses distributions of chlorophyll and carbon biomass inferred from satellite color observations to determine the biome structure of the ocean and how this is related to physical processes. In addition, through the use of observations of chlorofluorocarbons and radiocarbon, the ocean carbon modeling group has been evaluating how well ocean circulation models simulate processes of particular importance to ecosystem and carbon cycle dynamics in order to lay the groundwork for improving them (see “Evaluating ocean models with observations” page 37).

Other processes that were examined over the past year include the parameterization of air-sea gas exchange and the influence of phytoplankton on the radiative balance of the surface ocean. Finally, Sarmiento’s group has been contributing to the development of a flux-coupler for the GFDL Flexible Modeling System that will allow gas transfers between the land, atmosphere, and ocean components of the GFDL Earth System Model; and to the development of a coarse resolution earth system model required for longer term simulations, which are particularly important for studies of ocean carbon chemistry.