Climate models are essential for investigating and predicting marine heatwaves (MHWs), but their inherent biases pose challenges for accurately analyzing and projecting MHWs. As global warming intensifies, MHWs are expected to become more extreme, emphasizing the need for precise model evaluations and deeper understanding of the underlying causes of bias to improve MHW simulations. In this study, we evaluated the MHW intensity simulated by 30 Coupled Model Intercomparison Project Phase 6 (CMIP6) models in the North Pacific Ocean by comparing them with the optimum interpolation sea surface temperature (OISST) dataset over a 33 year period (1982–2014). Over 80% of the CMIP6 models showed a dipole-pattern bias in the mean MHW intensity over the Kuroshio Extension, characterized by overestimation to the north and underestimation to the south. Hierarchical clustering based on the spatial pattern of the intensity bias in the Kuroshio Extension identifies four distinct clusters: two showing a dipole-pattern bias with slightly different magnitudes and the other two showing weaker or stronger intensity. The separation latitude based on the maximum surface current revealed that the models with the dipole bias pattern simulated the northward overshooting of the Kuroshio Current. This northward overshooting likely enhanced the transport of warmer sea surface temperatures to higher latitudes, thereby intensifying the MHWs. These findings highlight the critical role of ocean currents in the simulation of MHW in the western boundary current regions, including the Kuroshio region.

Read original article

Read original article