2006b

Climate-Carbon Cycle Feedback Analysis: Results from the C4MIP Model Intercomparison

P. Friedlingstein, et al.

Observation and Context

Atmospheric CO₂ concentration is a primary driver of twenty-first-century climate change. Historically, climate models used simple offline carbon calculations that ignored how climate variations dynamically affect the earth’s natural carbon sinks. Real-world observations show that natural carbon fluxes are highly sensitive to climate shifts, such as those caused by El Niño events or volcanic eruptions. To resolve these inconsistencies, the Coupled Climate-Carbon Cycle Model Intercomparison Project (C4MIP) organized a study to evaluate how climate change and the carbon cycle interact within fully coupled, complex models.

Hypothesis

If rising atmospheric carbon dioxide drives global warming, then this climate change will in turn reduce the efficiency of the land and ocean to absorb anthropogenic carbon, creating a positive feedback loop that accelerates the rate of CO₂ accumulation in the atmosphere.

Experiment and Methodology

Eleven global climate-carbon cycle models (including 3D general circulation models and models of intermediate complexity) participated in the intercomparison using a standardized protocol. The models were forced with historical emissions from 1850 to 2000, followed by the IPCC SRES A2 anthropogenic emissions scenario through 2100.

To isolate the specific impact of climate change on carbon absorption, each modeling group performed two separate simulations:

• Coupled Simulation: Radiatively active CO₂ changes the climate, allowing researchers to observe how warming, drying, and circulation shifts affect land and ocean carbon sinks.
• Uncoupled Simulation: CO₂ is treated as a nonradiatively active gas, meaning the carbon cycle responds only to rising carbon concentrations (fertilization) without any accompanying climate change.

Mathematical analysis was applied to calculate the feedback gain (g) and define the specific sensitivities of the land (γL) and ocean (γO) sinks to climate warming, using temperature change (ΔT) as a proxy.

Results and Data

By 2100, there was unanimous agreement among the eleven models that the climate-carbon cycle feedback is positive:

• Atmospheric CO₂: Accounting for climate change resulted in an additional atmospheric accumulation of carbon dioxide ranging between 20 ppm and 200 ppm, with most models projecting an extra 50 to 100 ppm. This additional CO₂ led to an extra warming of 0.1 °C to 1.5 °C.
• Land and Ocean Sensitivities: Every model simulated a negative sensitivity to climate warming for both land and ocean reservoirs. However, the magnitudes varied widely; land sensitivity ranged from −20 GtC K⁻¹ to −177 GtC K⁻¹, and ocean sensitivity ranged from −14 GtC K⁻¹ to −67 GtC K⁻¹.
• Attribution: Eight out of the eleven models attributed the majority of the carbon storage reduction to the land, particularly noting reduced carbon uptake in tropical regions. However, there was no consensus on whether this land-sink reduction was driven by decreased plant productivity (NPP) or increased soil respiration (Rh).

Conclusion and Climate Impact

The experiment supported the hypothesis. Future climate change will reduce the earth’s natural capacity to absorb human carbon emissions, leaving a larger fraction of CO₂ airborne in the atmosphere. This positive feedback loop significantly amplifies projected global warming. These findings demonstrate that carbon emission targets must account for this weakening of natural sinks to successfully prevent severe, runaway climate warming over the twenty-first century.