In recent decades, many research efforts focused on global climate change, multidecadal, decadal, interannual variability, and the increasing extreme events of sea surface temperature. In contrast, the continuous evolution of the reference frame, the annual cycle of SST used to quantify the aforementioned variability and changes, has long been overlooked, resulting in difficulties in understanding the underlying physical mechanisms responsible for these variability and changes. In this study, we strive to bridge this gap on the phase changes in SST annual cycle. By devising a running correlation-based method, we can now quantify the non-sinusoidal shape of the evolving SST annual cycle, such as the advancing or delaying of summer and winter peaking times. It is revealed that the varying phases of summer or winter are more closely linked to multidecadal SST variability than to long-term climate change. Both the systematic shift of the phase and alterations in the annual cycle shape contribute to the phase changes, which explain 0.4~1.0 °C of monthly SST anomaly with respect to the climatological annual cycle in a multidecadal timescale. Furthermore, it is evident that the SST phases in historical simulations are better captured in winter than in summer and exhibit stronger variation compared with observation.
Read full abstract