Received: 08 Aug 2014 – Accepted for review: 21 Aug 2014 – Discussion started: 29 Aug 2014
Abstract. The influence of prognostic and prescribed atmospheric CO2 concentrations ([CO2]) on the carbon uptake and temperature is investigated using all eight Earth System Models (ESMs) with relevant output variables from the Coupled Model Intercomparison Project Phase 5 (CMIP5). Under the RCP8.5 scenario, the projected [CO2] differences in 2100 vary from −19.7 to +207.3 ppm in emission-driven ESMs. Incorporation of the interactive concentrations also increases the range of global warming, computed as the 20 year average difference between 2081–2100 and 1850–1869/1861–1880, by 49% from 2.36 K (i.e. ranging from 3.11 to 5.47 K) in the concentration-driven simulations to 3.51 K in the emission-driven simulations. The observed seasonal amplitude of global [CO2] from 1980–2011 is about 1.2–5.3 times as large as those from the eight emission-driven ESMs, while the [CO2] seasonality is simply neglected in concentration-driven ESMs, suggesting the urgent need of ESM improvements in this area. The temperature-concentration feedback parameter α is more sensitive to [CO2] (e.g. during 1980–2005 versus 2075–2100) than how [CO2] is handled (i.e. prognostic versus prescribed). This sensitivity can be substantially reduced by using a more appropriate parameter α' computed from the linear regression of temperature change versus that of the logarithm of [CO2]. However, the inter-model relative variations of both α and α' remain large, suggesting the need of more detailed studies to understand and hopefully reduce these discrepancies.
How to cite: Shao, P., Zeng, X., and Zeng, X.: Differences in carbon cycle and temperature projections from emission- and concentration-driven earth system model simulations, Earth Syst. Dynam. Discuss., 5, 991–1012, https://doi.org/10.5194/esdd-5-991-2014, 2014.