During the Paris conference in 2015, nations of the world strengthened the United Nations Framework Convention on Climate Change by agreeing to holding ‘the increase in the global average temperature to well below 2 °C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5 °C’ (ref. 1). However, ‘pre-industrial’ was not defined. Here we investigate the implications of different choices of the pre-industrial baseline on the likelihood of exceeding these two temperature thresholds. We find that for the strongest mitigation scenario RCP2.6 and a medium scenario RCP4.5, the probability of exceeding the thresholds and timing of exceedance is highly dependent on the pre-industrial baseline; for example, the probability of crossing 1.5 °C by the end of the century under RCP2.6 varies from 61% to 88% depending on how the baseline is defined. In contrast, in the scenario with no mitigation, RCP8.5, both thresholds will almost certainly be exceeded by the middle of the century with the definition of the pre-industrial baseline of less importance. Allowable carbon emissions for threshold stabilization are similarly highly dependent on the pre-industrial baseline. For stabilization at 2 °C, allowable emissions decrease by as much as 40% when earlier than nineteenth-century climates are considered as a baseline.
We thank K. Cowtan for making his code and results available and for help in their use and S. Phipps for CSIRO-Mk3L-1.2 model data. A.P.S., G.C.H. and S.F.B.T. were supported by the ERC-funded project TITAN (EC-320691) and A.P.S. and G.C.H. by NERC under the Belmont forum, grant PacMedy (NE/P006752/1), G.C.H. and S.F.B.T. were supported by NCAS (R8/H12/83/029) and G.C.H. was further funded by the Wolfson Foundation and the Royal Society as a Royal Society Wolfson Research Merit Award (WM130060) holder. E.H. and G.C.H. were supported by the NERC-funded SMURPHS project (NE/N006038/1) and E.H. by a NERC Fellowship (NE/I020792/1) and NCAS. M.E.M. acknowledges support for this work from the P2C2 programme of the National Science Foundation (grant ATM-1446329). We acknowledge the World Climate Research Programme’s Working Group on Coupled Modelling, which is responsible for CMIP, the climate modelling groups for producing and making available their model output, the US Department of Energy’s Program for Climate Model Diagnosis and Intercomparison, and the Global Organization for Earth System Science Portals for Earth System Science Portals. We thank F. Joos for discussion of causes of the CO2 increase since the Little Ice Age.