Premium
Improving ENSO prediction in a hybrid coupled model with an embedded entrainment temperature parameterisation
Author(s) -
Zhu Jieshun,
Zhou GuangQing,
Zhang RongHua,
Sun Zhaobo
Publication year - 2013
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.3426
Subject(s) - anomaly (physics) , sea surface temperature , climatology , el niño southern oscillation , environmental science , entrainment (biomusicology) , ocean general circulation model , atmospheric model , coupling (piping) , general circulation model , forcing (mathematics) , madden–julian oscillation , atmosphere (unit) , meteorology , geology , climate change , physics , oceanography , mechanical engineering , condensed matter physics , rhythm , acoustics , engineering , convection
Improving El Niño/Southern Oscillation (ENSO) forecast remains a great challenge in the climate‐predicting community. Previously, an improved solution to sea surface temperature (SST) anomaly simulations in the tropical Pacific was obtained by explicitly embedding into an ocean general circulation model (OGCM) a separate SST anomaly submodel with an empirical parameterisation for the temperature of subsurface water entrained into the ocean mixed layer ( T e ). In the present work, the benefit of the approach is explored and demonstrated in terms of ENSO prediction. A hybrid coupled ocean‐atmosphere model (HCM) is utilized to perform two retrospective ENSO forecasts, differing in the way SST anomaly fields are taken for their coupling to the atmosphere, one directly from the OGCM (referred to as a standard coupling, HCM std ), and another from the embedded SST anomaly submodel with optimized T e parameterisation (referred to as an embedded coupling, HCM embed ). The results indicate that ENSO forecasts can be effectively improved using the embedded approach; the predicted Niño‐3.4 SST anomaly correlation is higher by 0.1–0.2 at a 12‐month lead time in the HCM embed than in the HCM std , and the corresponding root‐mean‐square (RMS) error is lower by 0.1–0.2 °C. Further improvements and applications are discussed. Copyright © 2012 Royal Meteorological Society