Figure 2 : Empirical EUN superimposed on various forced signals (black lines) over the 20th and 21st centuries.

From: Comparing the model-simulated global warming signal to observations using empirical estimates of unforced noise

Figure 2

(a), Forced signal originating from the multi-model mean of the CMIP5 ‘historical experiment’ from 1900 to 2005 and RCP 6.0 from 2005 to 2013. (b), Hypothetical forced signal represented by a linear trend fit to observations during 1900 to 2013. (c), (d) and (e), Same as (a) but with forced signals corresponding to the RCP 4.5, RCP 6.0 and RCP 8.5 emissions scenarios extended to 2050. Uncertainty in the CGCM-produced forced signals is represented by ±2 standard errors calculated across all CGCM realizations (light blue shading surrounding the black line) and the EUN ranges are expanded by this uncertainty. Note that the light blue shading does not represent the uncertainty in the radiative forcings themselves but rather the uncertainty in the forced GMT signal given an assumed radiative forcing trajectory. Three arbitrary realizations of GMT from the ESRUN (superimposed on the forced signals) are shown in each panel (labeled Treali 1, Treali 2 and Treali 3). In all plots the observed GMT55 (as well as its 2σ uncertainty), is shown in yellow. All anomalies are defined relative to the average from 1961–1990.