Drivers of multi-century trends in the atmospheric CO₂ mean annual cycle in a prognostic ESM

Abstract. The amplitude of the mean annual cycle of atmospheric CO₂ is a diagnostic of seasonal surface–atmosphere carbon exchange. Atmospheric observations show that this quantity has increased over most of the Northern Hemisphere (NH) extratropics during the last 3 decades, likely from a combination of enhanced atmospheric CO₂, climate change, and anthropogenic land use change. Accurate climate prediction requires accounting for long-term interactions between the environment and carbon cycling; thus, analysis of the evolution of the mean annual cycle in a fully prognostic Earth system model may provide insight into the multi-decadal influence of environmental change on the carbon cycle.

We analyzed the evolution of the mean annual cycle in atmospheric CO₂ simulated by the Community Earth System Model (CESM) from 1950 to 2300 under three scenarios designed to separate the effects of climate change, atmospheric CO₂ fertilization, and land use change. The NH CO₂ seasonal amplitude increase in the CESM mainly reflected enhanced primary productivity during the growing season due to climate change and the combined effects of CO₂ fertilization and nitrogen deposition over the mid- and high latitudes. However, the simulations revealed shifts in key climate drivers of the atmospheric CO₂ seasonality that were not apparent before 2100. CO₂ fertilization and nitrogen deposition in boreal and temperate ecosystems were the largest contributors to mean annual cycle amplification over the midlatitudes for the duration of the simulation (1950–2300). Climate change from boreal ecosystems was the main driver of Arctic CO₂ annual cycle amplification between 1950 and 2100, but CO₂ fertilization had a stronger effect on the Arctic CO₂ annual cycle amplitude during 2100–2300. Prior to 2100, the NH CO₂ annual cycle amplitude increased in conjunction with an increase in the NH land carbon sink. However, these trends decoupled after 2100, underscoring that an increasing atmospheric CO₂ annual cycle amplitude does not necessarily imply a strengthened terrestrial carbon sink.