In a Carbon Cycle Data Assimilation System (CCDAS) one infers the values of the parameters controlling the function of a process model using various observations. One can then calculate quantities of interest from the optimized parameters and the model. One can also calculate the uncertainties on the parameters and propagate these to uncertainties of the calculated quantities. In Rayner et al. [2005] we assimilated atmospheric observations over two decades, into a terrestrial model and calculated fluxes over this period. Here we extend this work by calculating the response of the calibrated terrestrial biosphere to a GCM simulation of future climate. Using this combination we are able to comment on the fate of terrestrial carbon pools and fluxes under climate change, calculate the uncertainties of the response, and determine which parameters in the model are responsible for this uncertainty. We include an extra parameter that scales the climate change signal from the GCM projection. We thus extend the sensitivity and uncertainty analysis to include the climate sensitivity.

Future climate warming is expected to enhance plant growth in temperate ecosystems and to increase carbon sequestration. But although severe regional heatwaves may become more frequent in a changing climate, and their impact on terrestrial carbon cycling is unclear. Europe experienced a particularly extreme climate anomaly during 2003, with July temperatures up to 6°C above long-term means, and annual precipitation deficits up to 300 mmy^-1, that is 50% below the average. We used the 2003 heatwave as a ‘laboratory assistant’ to estimate the impact on terrestrial carbon cycling.

The mixing ratios of atmospheric CO₂ were observed at the summit of Mt. Fuji by using a system for continuous measurements during September 2002-February 2003 and May 2003-May 2004. The observed CO₂ variations at Mt. Fuji showed a seasonal cycle of the background level with a maximum around April and a minimum around August. A lot of episodic events with a large enhancement of CO₂ were found, and the episodic enhancements of CO₂ at Mt. Fuji were well associated with increased CO peaks observed at the same time. The enhancement ratios of CO to CO₂ mixing ratios (ΔCO/ΔCO₂) mainly showed lower values of less than 0.03 due to urban/industrial sources, while relatively higher ΔCO/ΔCO₂ ratios up to 0.08 were also found for the episodic events due to the biomass burning emissions. Three-dimensional transport model simulations of CO suggested that the major contributions for the increased events at Mt. Fuji were from China (~50%) and the other major regions were Southeast Asia and South Asia (~10%).

This study examines the potential for feedbacks between the carbon cycle, atmospheric carbon dioxide (CO₂) increases and climate change to affect the anthropogenic emissions that are required to stabilize future levels of CO₂ in the atmosphere. Using a coupled climate-carbon cycle model, I found that positive carbon cycle-climate feedbacks reduced allowable emissions by an amount that varied with the model’s climate sensitivity.  Emissions were further reduced if CO₂ fertilization was assumed to be inactive in the model, as this removed an otherwise important negative feedback on atmospheric CO₂.

Increasing ocean stratification associated with global warming has been posited to serve as a positive feedback on global warming, reducing the oceanic uptake of anthropogenic carbon dioxide. We suggest that a poleward shift of westerly winds combined with future increases in atmospheric carbon dioxide may drive an increase in the CO₂ uptake in the Southern Ocean, representing a negative feedback on atmospheric anthropogenic CO₂.

In recent years attention has focused on the role of terrestrial biosphere dynamics in the climate system, and the possibility of large land-atmosphere carbon cycle feedbacks under human-induced future climate warming. During the 1990s rapid development of Dynamic Global Vegetation Models (DGVMs) led a growing community to soon recognize the need for model evaluation and intercomparison. In Cramer et al. 2000 six DGVMs were run using identical forcing data based on the HadCM2 GCM climatology (1860-2100) and the IS92a emission scenario.

The terrestrial ecosystem carbon cycle model, Sim-CYCLE, was combined with the CCSR/NIES/FRCGC AGCM5.7b (including a land surface model: MATSIRO). That coupled model shows a reasonable distribution of the LAI, NPP and other carbon storages after the 1000yrs spin-up run. This presentation introduces the preliminary results of the coupled run in 20th century.

Here we use the severe heat and drought event in Europe from summer 2003 as a natural experiment to study the impact of a climatic extreme event on ecosystem physiology and its feedback to the atmosphere. The combination of continuous eddy covariance and tree growth measurements at two nearby located deciduous forests showed a large reduction in carbon uptake during the drought (-30%) and a strong carry-over effect into the next year. Both forests, however, responded differently, although climatic forcing was almost identical. Species composition and site condition of the ecosystems seemed to play a major role in the ecosystems response to the drought.

Different CO₂ stabilization scenarios and CO₂ emission scenarios have been carried out with an earth system model to investigate feedbacks between future climate change and carbon cycle. The model predicts a sensitivity of 1.6±0.1 K for an increase of 280 ppm in atmospheric CO₂ concentration. The decrease of the thermohaline circulation is predominantly controlled by an enhanced atmospheric moisture transport to high latitudes by global warming. Overall, the simulated effect of atmospheric CO₂ concentration on climate change reduces the total carbon uptake of the ocean and the land is reduced by 24-29%.