Energy Flow Diagnosis of ENSO from an Ocean Reanalysis

Abstract

A method is introduced for diagnosing the time evolution of wave energy associated with ENSO from an ocean reanalysis. In the diagnosis, time changes of kinetic and available potential energy are mainly represented by energy inputs caused by surface wind stress and horizontal energy fluxes for each vertically decomposed normal mode. The resulting time evolutions of the wave energy and vertical thermocline displacements in the 1997/98 and 2014–16 El Niño events are consistent with our previous knowledge of these events. Further, our result indicated that representation of several vertical modes is necessary to reproduce the broadly distributed downward thermocline displacements in the central to eastern equatorial Pacific, generated by a westerly wind event in the western equatorial Pacific (e.g., in March 1997), that are preconditioning for El Niño development. In addition, we investigated the wave energy budget, including the influence of data assimilation, on the complicated time evolution of equatorial thermocline displacements caused by repeated westerly and easterly wind events during the 2014–16 El Niño event. Our result suggests that noise from a momentum imbalance near the equator associated with data assimilation, which possibly affected the El Niño prediction failure in 2014, was much reduced by our developed ocean data assimilation system and reanalysis. This study, which provides a new connection between the theoretical works and reanalysis products that use sophisticated systems for synthesizing OGCMs and observations, should be useful for climate research and operational communities interested in ENSO. A method is introduced for diagnosing the time evolution of wave energy associated with ENSO from an ocean reanalysis. In the diagnosis, time changes of kinetic and available potential energy are mainly represented by energy inputs caused by surface wind stress and horizontal energy fluxes for each vertically decomposed normal mode. The resulting time evolutions of the wave energy and vertical thermocline displacements in the 1997/98 and 2014–16 El Niño events are consistent with our previous knowledge of these events. Further, our result indicated that representation of several vertical modes is necessary to reproduce the broadly distributed downward thermocline displacements in the central to eastern equatorial Pacific, generated by a westerly wind event in the western equatorial Pacific (e.g., in March 1997), that are preconditioning for El Niño development. In addition, we investigated the wave energy budget, including the influence of data assimilation, on the complicated time evolution of equatorial thermocline displacements caused by repeated westerly and easterly wind events during the 2014–16 El Niño event. Our result suggests that noise from a momentum imbalance near the equator associated with data assimilation, which possibly affected the El Niño prediction failure in 2014, was much reduced by our developed ocean data assimilation system and reanalysis. This study, which provides a new connection between the theoretical works and reanalysis products that use sophisticated systems for synthesizing OGCMs and observations, should be useful for climate research and operational communities interested in ENSO.