Human-induced global ocean warming on multidecadal timescales

Large-scale increases in upper-ocean temperatures are evident in observational records1. Several studies have used well-established detection and attribution methods to demonstrate that the observed basin-scale temperature changes are consistent with model responses to anthropogenic forcing and inconsistent with model-based estimates of natural variability2,3,4,5. These studies relied on a single observational data set and employed results from only one or two models. Recent identification of systematic instrumental biases6 in expendable bathythermograph data has led to improved estimates of ocean temperature variability and trends7,8,9 and provide motivation to revisit earlier detection and attribution studies. We examine the causes of ocean warming using these improved observational estimates, together with results from a large multimodel archive of externally forced and unforced simulations. The time evolution of upper ocean temperature changes in the newer observational estimates is similar to that of the multimodel average of simulations that include the effects of volcanic eruptions. Our detection and attribution analysis systematically examines the sensitivity of results to a variety of model and data-processing choices. When global mean changes are included, we consistently obtain a positive identification (at the 1% significance level) of an anthropogenic fingerprint in observed upper-ocean temperature changes, thereby substantially strengthening existing detection and attribution evidence