New paper finds transient climate sensitivity to CO2 is ~35% less than IPCC claims

A paper under discussion for Earth System Dynamics finds the transient climate response [TCR] to CO2 is 1.3°C per doubling of CO2 levels, about 35% less than claimed by the IPCC mean estimate and the same as determined by another recent paper by Otto et al finding a TCR of 1.3°C. 

The authors

 "assess the origin of these differences [between the IPCC high TCR estimates and lower estimates from this paper and others] and highlight the inverse relation between the derived anthropogenic temperature trend of the past 30 years and the weight given to the [natural] Atlantic Multidecadal Oscillation (AMO) as an explanatory factor in the multiple linear regression (MLR) tool that is usually employed. We highlight that robust MLR outcomes require a better understanding of the AMO in general and more specifically its characterization. Our results indicate that both the high- and low end of the anthropogenic trend over the past 30 years found in previous studies are unlikely and that a transient climate response with best estimates centred around 1.3°C per CO2 doubling best captures the historic instrumental temperature record."

Earth Syst. Dynam. Discuss., 5, 529-544, 2014 www.earth-syst-dynam-discuss.net/5/529/2014/ doi:10.5194/esdd-5-529-2014 Impact of the Atlantic Multidecadal Oscillation (AMO) on deriving anthropogenic warming rates from the instrumental temperature record

G. R. van der Werf and A. J. Dolman
VU University Amsterdam, Faculty of Earth and Life Sciences, Amsterdam, the Netherlands
Abstract. The instrumental surface air temperature record has been used in several statistical studies to assess the relative role of natural and anthropogenic drivers of climate change. The results of those studies varied considerably, with anthropogenic temperature trends over the past 25–30 years suggested to range from 0.07 to 0.20 °C decade⁻¹. In this short communication we assess the origin of these differences and highlight the inverse relation between the derived anthropogenic temperature trend of the past 30 years and the weight given to the [natural] Atlantic Multidecadal Oscillation (AMO) as an explanatory factor in the multiple linear regression (MLR) tool that is usually employed. We highlight that robust MLR outcomes require a better understanding of the AMO in general and more specifically its characterization. Our results indicate that both the high- and low end of the anthropogenic trend over the past 30 years found in previous studies are unlikely and that a transient climate response with best estimates centred around 1.3 °C per CO₂ doubling best captures the historic instrumental temperature record.