Strong Nonlinearity of Land Climate-Carbon Cycle Feedback Under a High CO2 Growth Scenario

Projections of future climate change for given CO2 and other greenhouse gas emission scenarios depend on the response of global climate-carbon cycle feedback, which consists of carbon-concentration feedback (e.g., CO2 physiology effect on land carbon sink) and carbon-climate feedback (e.g., CO2 radiative effect on land carbon sink). Previous studies have assumed no significant interaction between these two feedbacks within the Earth system. This study quantifies the interaction of these two feedbacks, or the nonlinear feedback on land using the fully, biogeochemically, and radiatively coupled simulations under a 1% yr(-1) CO2 increase path from nine Earth system models of the Coupled Model Intercomparison Project Phase 6 (CMIP6). The results show that the nonlinear feedback is 1.64 +/- 2.92 x 10(-2) GtC ppm(-1) K-1 at the end of 140-year simulation with a quadrupling CO2 (4 x CO2), where its strength is 11% +/- 18% of the carbon-concentration feedback or -27% +/- 49% of the carbon-climate feedback on land. Compared to previous assumptions that did not consider this interaction, the nonlinear feedback contributes about 8% +/- 12% of the land carbon increase accumulated at the 4 x CO2. The nonlinear feedback largely results from the combined effect of increased CO2-induced additional fertilization effect on warming-induced additional leaf area index and vegetation productivity over the Northern Hemisphere. The magnitude of the nonlinear feedback on land decreases with an increase in atmospheric CO2 or warming under the high emission scenario. This study highlights the significance of land nonlinear climate-carbon cycle feedback in increasing land carbon sink and slowing down future climate change. Plain Language Summary The land biosphere absorbs a significant fraction of anthropogenic carbon emissions, and the rate of this carbon sink increases with atmospheric CO2 concentration-induced physiology effect on carbon sink (the so-called carbon-concentration feedback), but generally decreases with CO2-induced radiatively warming effect on land carbon sink (i.e., the so-called carbon-climate feedback). Quantifying the strength of these two feedbacks is one of the great challenges of climate change science today. Most previous studies examining these two feedbacks have ignored their nonlinear interaction, which may induce large uncertainty for future climate projections under high CO2 emission scenarios. To quantify this nonlinear feedback, we used simulations by nine Earth system models driven by a prescribed 1% per year increasing atmospheric CO2 concentration for 140 years. We find that the nonlinear feedback can provide a strong additional contribution (about 8% +/- 12%) to the global land carbon increase for a quadrupled CO2 concentration scenario. Further analysis shows that the combined effects of rising CO2 fertilization and warming-induced growing vegetation productivity over the Northern Hemisphere dominate the nonlinear feedback. This study based on Earth system models provides new insights into the interaction between climate and carbon cycle, and highlights the significant nonlinear effect in projecting future carbon mitigation.