Record temperature streak bears anthropogenic fingerprint

We use a previously developed semiempirical approach to assess the likelihood of the sequence of consecutive record-breaking temperatures in 2014-2016. This approach combines information from historical temperature data and state-of-the-art historical climate model simulations from the Coupled Model Intercomparison Project Phase 5 (CMIP5). We find that this sequence of record-breaking temperatures had a negligible (< 0.03%) likelihood of occurrence in the absence of anthropogenic warming. It was still a rare but not implausible event (roughly 1-3% likelihood) taking anthropogenic warming into effect. The probability that three consecutive records would have been observed at some point since 2000 is estimated as similar to 30-50% given anthropogenic warming and < 0.7% in its absence. The likelihood of observing the specific level of record warmth recorded during 2016 is no more than similar to one-in-a-million neglecting anthropogenic warming, but as high as 27%, i.e., a nearly one-in-three chance of occurrence taking anthropogenic warming into account.