Weather and Climate Extremes (Dec 2023)

Predictability of Northwest Pacific marine heatwaves in summer based on NUIST-CFS1.0 hindcasts

  • Tong Zhang,
  • Haiming Xu,
  • Jing Ma,
  • Jiechun Deng

Journal volume & issue
Vol. 42
p. 100617

Abstract

Read online

Marine ecosystems in the Northwest Pacific (NWP) are being threatened by marine heatwaves (MHWs), which are abnormally high sea-surface temperature (SST) lasting for more than 5 days. However, there is limited research on the predictability of these MHWs. In this study, we investigate the predictability of NWP MHWs in summer using daily data from the coupled climate forecast system (version 1.0) of the Nanjing University of Information Science and Technology (NUIST-CFS1.0) for the period of 1984–2020. We demonstrate that the NUIST-CFS1.0 hindcasts can predict the spatial pattern of the total MHW days (HWD) over the NWP in summer with a lead time of up to eight months. Moreover, the linear trend and interannual variability can be predicted at lead times of up to nine and three months, respectively. We reveal that the interannual variability of HWD is strongly correlated with the preceding El Niño-Southern Oscillation (ENSO); and its predictability also shows significant interannual variability, with MHWs being more predictable in the decaying years of stronger El Niño events. In the El Niño decaying summers, there is an increase in the occurrence of MHWs; the spatial distribution of HWD can be predicted at lead times of up to 19 months, and its forecasting accuracy is far superior to that of the other years. The occurrence of MHWs during El Niño decaying summers is closely linked to strengthened NWP subtropical anticyclone (NWPSA), which increases solar radiation reaching the ocean surface and suppresses surface evaporation, leading to increased SST and thereby MHWs. The remarkable predictability of the MHWs in the NWP region during El Niño decaying summers can be attributed to the high skill at predicting ENSO and its associated NWPSA anomalies.

Keywords