GFDL - Geophysical Fluid Dynamics Laboratory

Skip to main content

Trajectory sensitivity of the transient climate response to cumulative carbon emission

Key Findings

  • This study quantifies the sensitivity of the climate response to cumulative carbon emissions trajectory – something previously assumed to be a small source of uncertainty.
  • The Transient Climate Response to cumulative carbon Emissions (TCRE) is largest for both very low emission rates (e.g., 2 GtC/yr) and very high (e.g., 25 GtC/yr) emission rates.
  • TCRE is lowest for emission rates that are comparable to the present day (5 to 10 GtC/yr).
  • Structural model uncertainty (i.e. differences in model parameterizations) is larger than emissions pathway uncertainty for TCRE.
  • The TCRE signal is difficult to detect under low emissions and on short time scales.
  • The climate response to zero-emissions depends on prior emission rates.

J. P. Krasting, J. P. Dunne, E. Shevliakova, R. J. Stouffer. Trajectory sensitivity of the transient climate response to cumulative carbon emission. Journal: Geophysical Research Letters. DOI: 10.1002/2013GL059141.


The insensitivity of the global climate response once emissions cease in simplified coupled climate models has been used to argue for lack of committed warming based on past carbon emissions. Additional studies have also demonstrated that the cessation of carbon emissions results in a stabilization or decrease of global mean surface air temperature. Such studies generally assume either a 1%/year increase or an instantaneous doubling/quadrupling of atmospheric CO2.

For this study, the ESM2G coupled climate-carbon cycle model was used to evaluate Transient Climate Response to cumulative carbon Emissions (TCRE) by forcing the model with 7 constant carbon emission rates (2, 3, 5, 10, 15, 20 and 25 GtC/yr), including low emission rates that have been largely unexplored in previous studies. The range of TCRE resulting from these varying emission pathways is 0.76 to 1.04 ºC/TtC.

TCRE has a complex relationship with emission rates; TCRE is largest for both very low (2 GtC/yr) and very high (25 GtC/yr) emissions and smallest for present-day emissions (5-10 GtC/yr). We demonstrate that natural climate variability hinders precise estimates of TCRE for periods shorter than 50 years for emission rates near or smaller than present day values. If carbon emissions were to suddenly cease, the prior emissions pathways would affect the future climate responses.

This study confirms the robustness of the proportionality between surface air warming and ongoing cumulative carbon emissions. This study, however, demonstrates a larger response dependence on carbon emissions pathway when forced by constant "low" carbon emissions that take many centuries to double atmospheric CO2. This trajectory dependence is related, in part, to oceanic heat and carbon uptake. These processes are dependent on ocean mixing and the equilibrium climate sensitivity, which vary from model to model. The results from these experiments also indicate that the mechanisms that give rise to the proportionality depend of the emissions pathway and possibly depend on processes other than oceanic heat and carbon uptake, as demonstrated here by the varying terrestrial response to different carbon emission rates.

TCRE is appealing to both the scientific and policy communities since it provides a single metric that integrates both climate and carbon cycle sensitivities. Based on the results of this study, the additional uncertainty related to emissions trajectory should be taken into account when interpreting TCRE in policy-making timescales (i.e. multi-decadal and longer). TCRE, however, has the potential to be a useful metric in bench-marking coupled climate-carbon-cycle models.

Ensemble mean (solid line) and range (envelope).
a. cumulative carbon emissions (GtC) vs. time (note the different number of years on the x-axis), dashed lines represent the zero emissions experiments;
b. Global surface air temperature response (C);
c. Model-predicted atmospheric CO2 concentration (ppm);
d. 20-yr smoothed TCRE (C/TtC);
e. 20-yr smoothed atmospheric fraction of carbon;
f. cumulative land carbon uptake (GtC);
g. cumulative ocean carbon uptake (GtC). Dots indicate the time of atmospheric CO2 doubling.