Research ArticleCLIMATOLOGY

Slow climate mode reconciles historical and model-based estimates of climate sensitivity

See allHide authors and affiliations

Science Advances  05 Jul 2017:
Vol. 3, no. 7, e1602821
DOI: 10.1126/sciadv.1602821
  • Fig. 1 Evolution of top-of-atmosphere energy flux as a function of global temperature.

    Annual average (brown dots) and 10-year running average values (black line) of top-of-atmosphere energy flux (TOA flux or H) and global-average temperature (T) are shown for a representative CMIP5 GCM (NorESM1-M) along with draws from the Bayesian fits (pink lines) and their median (red line). The distributions of radiative forcing (green line along y axis) and ECS (purple line along x axis) are estimated as part of the Bayesian fit. Curvature in the fit can be compared against a constant equilibrium feedback parameter that connects median radiative forcing to median warming (dashed gray line), where convexity indicates a decrease in the magnitude of λ(t) with time. Note that the line of constant λ used here connects median TOA and ECS values, whereas a regression line from the Gregory method (10) is used elsewhere. Also, note that both T and H are halved to facilitate comparison of these results using quadrupled CO2 forcing with the standard definition of ECS using doubled CO2. Ticks at top of graph indicate simulation year. NorESM1-M was chosen on the basis of having the minimum total deviation from the ensemble median in ECS, ICS, and F (see figs. S4 to S7 for the other 23 GCMs.)

  • Fig. 2 Evolution of fast and slow modes under historical forcing.

    Response of the CMIP5 GCMs to AR5 historical forcing from 1750 to 2011 (brown) and to an abrupt increase in 1750 to the 2011 forcing value of 2.2 W/m2 (blue). Responses are derived from the medians of our Bayesian fit, which permits for parsing ultrafast (mode 1, solid lines), fast (mode 2, dashed line), and slow (mode 3, dash-dot lines) contributions. Equilibrium contributions (green lines), referred to as committed warming in this context (43, 44), are also parsed according to mode for a net forcing of 2.2 W/m2. Mode 3 accounts for 44% of equilibrium warming but only 3% of present warming (see Table 1 for the median and 5 to 95% range for relevant parameters).

  • Fig. 3 Equilibrium and instantaneous climate sensitivity distributions.

    (A) Distribution of ECS from 5000 posterior draws of our Bayesian fit to each of 24 GCMs (indicated by colors). Aggregating across the posterior draws for all GCMs yields a median of 3.4°C and a 5 to 95% CI of 2.2° to 6.1°C. (B) Similar to (A) but for instantaneous climate sensitivity. ICS is obtained by applying AR5 historical forcing to our Bayesian fits and has a median of 2.5°C and a 5 to 95% CI of 1.6° to 4.2°C. The range of ICS values estimated in historical studies (vertical dashed black lines) (8) bracket the most likely GCM values, demonstrating consistency between observational and GCM results when they are appropriately compared.

  • Table 1 Contributions of fast and slow modes to equilibrium and historical warming.

    The posterior median and 5 to 95% CI values for the time scale, τ, and feedback factors, λ, of individual eigenmodes are provided, along with the fraction of warming contributed in equilibrium and from historical forcing as of 2011.

    Eigenmode parametersMedian5%95%
    τ1 (years)0.80.22.6
    λ1 (W/m2 per°C)1.60.44.0
    Contribution to inferred equilibrium warming24%10%40%
    Contribution to historical warming47%26%75%
    τ2 (years)9437
    λ2 (W/m2 per°C)1.40.82.8
    Contribution to inferred equilibrium warming32%15%48%
    Contribution to historical warming49%24%72%
    τ3 (years)350180960
    λ3 (W/m2 per°C)0.80.31.6
    Contribution to equilibrium warming44%28%66%
    Contribution to historical warming3%1%7%

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/7/e1602821/DC1

    Supplementary Text

    fig. S1. Structure of residuals.

    figs. S2 to S5. Figure 1 continued.

    fig. S6. Spatial projection of the eigenmodes.

    fig. S7. Historical temperature anomalies.

    fig. S8. Autocovariance of temperature residuals.

    fig. S9. Autocovariance of energy flux residuals.

    table S1. Posterior parameter values for each GCM.

    data file S1. Ensemble of posterior draws.

    References (4952)

  • Supplementary Materials

    This PDF file includes:

    • Supplementary Text
    • fig. S1. Structure of residuals.
    • figs. S2 to S5. Figure 1 continued.
    • fig. S6. Spatial projection of the eigenmodes.
    • fig. S7. Historical temperature anomalies.
    • fig. S8. Autocovariance of temperature residuals.
    • fig. S9. Autocovariance of energy flux residuals.
    • table S1. Posterior parameter values for each GCM.
    • References (49–52)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement: