Evaluation and projection of global marine heatwaves based on CMIP6 models

Abstract

Marine heatwaves (MHWs) are extreme climatic events that last for days to months and can extend up to thousands of kilometers and cause substantial ecological, social, and economic impacts. Climate models are the key tool for studying and predicting MHWs. However, it continues to be challenging for climate models to accurately simulate MHWs. In this study, we evaluate 29 models from the Coupled Model Intercomparison Project Phase 6 (CMIP6) and 19 models from the Coupled Model Intercomparison Project Phase 5 (CMIP5) in terms of their capabilities to simulate MHWs by examining the spatial patterns and temporal variations. Then, we estimate future changes until the end of the 21st century under three shared socioeconomic pathways (SSPs) (e.g., SSP126, SSP245, and SSP585). The results show that the CMIP6 ensemble mean is more skillful in capturing the features of MHWs than that of the CMIP5. The biases of the CMIP6 models for the MHWs intensities are within ±0.5 °C over most of the oceans, except in the western boundary current regions and eastern tropical Pacific, where the modeled MHWs are up to 1.5 °C weaker than the observations. In comparison, the results from CMIP5 are greater than ±1.5 °C in most areas. Both the CMIP5 and CMIP6 models underestimate long-duration MHWs in the eastern tropical Pacific, where they are nearly 20 days shorter than the observations. In most areas, the CMIP5 models overestimate the MHWs durations (by over 25 days), while the biases of the CMIP6 models are within 10 days. The projected MHWs exhibit significant increases in the intensity and duration and reach maximum intensities of 4 °C. The largest changes are projected to occur in the tropics, North Pacific, and North Atlantic. When comparing the shared socioeconomic pathways for the increasing trend of MHWs, the most extreme MHWs occur under SSP585, with their intensities nearly doubling and a near-permanent MHWs state occurring by the 2070s.