Table 2.
Emergent Constraints for S Based Upon Present‐Day Climate System Variables and CMIP Models
| Category | Present‐day climate system variable | Reference | Authors' statements about S | Central estimate of S (K) from ordinary linear regression | Central estimate of λ (W m−2 K−1) from ordinary linear regression |
|---|---|---|---|---|---|
| Low cloud | Boundary layer cloud amount response to SST variations in subtropical stratocumulus regions (after removing the stability contribution) | Qu et al. (2014) | No statement | 3.74 | −1.03 |
| Seasonal response of boundary layer cloud amount to SST variations in oceanic subsidence regions between 20°and 40° latitude | Zhai et al. (2015) | Models consistent with observation “have S higher than the multi‐model mean with “an ensemble mean S of 3.9 K and a standard deviation of 0.45 K.” | 4.13 | −0.82 | |
| Fraction of tropical clouds with tops below 850 hPa whose tops are also below 950 hPa | Brient et al. (2016) | Models consistent with observation “have S between 2.4 and 4.6 K.” | 3.06 | −1.20 | |
| Sensitivity of cloud albedo in tropical oceanic low‐cloud regions to present‐day SST variations | Brient and Schneider (2016) | “Most likely S estimate around 4.0 K; an S below 2.3 K becomes very unlikely (90% confidence).” | 3.68 | −0.92 | |
| General cloud | Difference between tropical and Southern Hemisphere midlatitude total cloud fraction | Volodin (2008) | An estimate of S is “3.6 ± 0.3” (1‐sigma). | 3.63 | −0.97 |
| Extent to which cloud albedo is small in warm SST regions and large in cold SST regions | Siler et al. (2017) | A likely value of S is “3.68 ± 1.30 K (90% confidence).” | 3.55 | −0.97 | |
| Humidity | Southern Hemisphere zonal average midtropospheric relative humidity in dry zone between 8.5°S and 20°S | Fasullo and Trenberth (2012) | “Many models, particularly those with low S, … are identifiably biased.” | 4.12 | −0.96 |
| Tropical zonal average lower tropospheric relative humidity in moist convective region | Fasullo and Trenberth (2012) | “Only a few models, generally of lower sensitivity, are identifiably biased.” | 3.42 | −1.06 | |
| Tropospheric zonal average relative humidity vertically and latitudinally resolved between 40°N and 40°S | Su et al. (2014) | “Models closer to the satellite observations tend to have S higher than the multi‐model mean.” | 3.85 | −0.90 | |
| Strength of resolved‐scale humidity mixing between the boundary layer and the lower troposphere in tropical East Pacific and Atlantic | Sherwood et al. (2014) | No specific statement | 4.13 | −0.76 | |
| Strength of small‐scale humidity mixing between the boundary layer and the lower troposphere in tropical convective regions | Sherwood et al. (2014) | No specific statement | 3.26 | −1.14 | |
| Sum of Sherwood resolved‐scale and small‐scale humidity mixing | Sherwood et al. (2014) | “Observations at face value implies a most likely S of about 4 K, with a lower limit of about 3 K.” | 4.07 | −0.83 | |
| Precipitation | Strength of model's precipitation bias in the “double‐ITCZ” (Intertropical Convergence Zone) region | Tian (2015) | “S might be in the higher end of its range (~4.0 K).” | 4.02 | −0.87 |
| Radiation | Net top‐of‐atmosphere radiation averaged over the Southern Hemisphere | Trenberth and Fasullo (2010) | “Only the more sensitive [higher S] models are in the range of observations.” | 3.53 | −1.05 |
| Temperature | Amplitude of seasonal cycle of surface temperature | Covey et al. (2000) | No specific statement | 3.23 | −1.16 |
| Strength of global average surface temperature interannual variations and their temporal autocorrelation | Cox et al. (2018) | The emergent constraint “yields a central [S] estimate of 2.8 K with 66% confidence limits … of 2.2–3.4 K.” | 2.91 | −1.22 | |
| Circulation | Latitude of the southern edge of the Hadley cell in austral summer | Lipat et al. (2017) | Models “closer to the observations … tend to have smaller S values.” | 2.80 | −1.23 |
| Average | 3.60 ± 0.42 | −1.01 ± 0.14 | |||
Note. Emergent constraints are categorized by the type of present‐day climate system variable (Columns 1 and 2) with the reference for each constraint in Column 3. Column 4 reports the authors' statements about S quoted directly from the cited reference. Column 5 reports a central estimate of S from each constraint calculated from the ordinary least squares linear regression of S on the present‐day climate system variable evaluated at its observed value. The data used in these calculations are taken from that compiled by Caldwell et al. (2018). Column 6 reports a central estimate for λ calculated in the same manner as Column 5. The last row reports the averages and standard deviations of the data in Columns 5 and 6.