Dependence of Climate Sensitivity Estimates on Internal Climate Variability During 1880-2020

Abstract

Abstract Observed rates of global-average deep-ocean and surface warming during 1880–2020 are matched with a 1D forcing-feedback model of vertical energy flow departures from assumed energy equilibrium driven by both anthropogenic and natural forcings. The monthly time resolution model ocean has a mixed layer, a second layer to 2,000 m depth, and a third layer extending to the ocean bottom. The model mixed layer temperature is radiatively forced with estimates of anthropogenic, volcanic, and direct solar forcings since 1765, and radiatively and non-radiatively forced with the observed history of El Niño Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO) activity since 1880. Model adjustable parameters are chosen to match observed sea surface temperature (SST) and deep ocean (0-2000 m) temperature trends during 1960–2020, as well as twenty years of lag regression relationships between sea surface temperature and satellite longwave and shortwave flux anomalies at the top of the atmosphere. The results support the dominant role of anthropogenic greenhouse gas forcing in ocean warming since 1880. Without ENSO and PDO effects, diagnosed climate sensitivity is 1.8 and 2.2 deg. C for two sea surface temperature datasets, respectively. Inclusion of ENSO and PDO improves agreement between model and observations, from 81–92% explained variance, and reduces the diagnosed sensitivity to 1.5 to 1.8 deg. C, depending on the surface temperature dataset. The global warming slowdown during 1998–2012 is also well matched after inclusion of ENSO and PDO effects.