2024 global temperature record is consistent with model-predicted warming

Abstract

We employ a semiempirical approach combining climate model simulations and observational temperatures to assess the likelihood of recent global temperature records. Monte Carlo simulations are used to generate global temperature series consistent with combined estimates of forced (anthropogenic + natural) and internal variability derived from observations and CMIP6 multimodel simulations. We find that the El Niño-boosted 2024 global temperature record had a ~12% likelihood of occurrence (a one-in-eight-year event), similar to the prior (also El Niño-boosted) record year 2016 (~14% likelihood). Of the records set during the past three decades, only 1998 is found to have been truly anomalous, with a ~2.5% likelihood of occurrence. Each of these records is found to have been nearly impossible in the absence of human-caused warming.