Re -writing History: the Role of Forcing in Surge and Hiatus Events over the 20th and 21st Centuries

The CMIP5 archive and CMIP6 protocol uses only one historical scenario pathway, representing the best estimate of radiative forcing evolution over 1850 to near-present. However, several components of the total effective radiative forcing such as aerosols have time varying uncertainties, and it is possible that some of the observed temperature record over the 20th and 21st centuries may be due to decadal changes in forcing rather than internal variability. The standard single historical pathway is not able to sample this variation, and we are limited to assessing uncertainty due to model differences and internal variability driven with the same forcing (or emissions or concentrations) time series.

Using distributions of climate sensitivity and radiative forcing guided by the IPCC Fifth Assessment Report, we generate a 100,000 member ensemble of forcing evolution and temperature response using the FaIR simple climate model. Twelve representative pathways are selected from this large ensemble, half of which represent the 5th percentile and the other half the 95th percentile of present-day forcing but with very different relative strengths of greenhouse gas, ozone, aerosol, solar and volcanic forcing. All twelve pathways are consistent with the uncertainty in 20th-century temperature observations. These pathways, plus one representing the best estimate, are run in the HadGEM3-GA7.1 climate model, using a scaled historical time series of concentrations or emissions from CMIP6.

The evolution in effective radiative forcing across the 13-member ensemble closely follows temperature "surges" (1910-1940 and 1970-1998) and "hiatuses" (1940-1970 and 1998-2012) seen in the historical record. This suggests that a substantial portion of the most recent hiatus may have been externally forced rather due to internal variability.