Assessment of the marine heatwaves prediction performance of the short-term climate prediction system FIO-CPS v2.0

Abstract

In recent years, the frequent occurrence of marine heatwaves (MHWs) has affected the ecological environment and caused considerable socioeconomic impact. Consequently, MHWs prediction has received increasing attention. This study aims to evaluate the short-term (months to interannual timescales) MHWs prediction skill of the First Institute of Oceanography-Climate Prediction System version 2.0 (FIO-CPS v2.0) by using three statistical metrics including symmetric extremal dependence index (SEDI), forecast accuracy (FA), and Brier skill score (BSS). The results revealed that FIO-CPS v2.0 can better predict MHWs in tropical regions, especially in the tropical central and eastern Pacific Ocean, in which the SEDI, FA, and BSS values reached 0.73, 0.92, and 0.27 at the 1-month lead time, respectively. However, the MHWs prediction ability of FIO-CPS v2.0 has spring prediction barriers owing to the driving factors of ENSO. Moreover, further analysis revealed a definite relationship between the ability to predict MHWs and the ability to predict MHWs duration. The prediction skill of FIO-CPS v2.0 appears to be better for long-duration MHWs than for short-duration MHWs. Under the influence of global warming, FIO-CPS v2.0 can reproduce the increase in MHWs duration observed in recent years, and the prediction skill of some regions has been relatively high in the last 15 years. This study deepens the understanding of the prediction ability of FIO-CPS v2.0, and provides an important reference for the application of short-term prediction of MHWs.