U-Net Models for Representing Wind Stress Anomalies over the Tropical Pacific and Their Integrations with an Intermediate Coupled Model for ENSO Studies

El Niño-Southern Oscillation(ENSO)is the strongest interannual climate mode influencing the coupled ocean-atmosphere system in the tropical Pacific,and numerous dynamical and statistical models have been developed to simulate and predict it.In some simplified coupled ocean-atmosphere models,the relationship between sea surface temperature(SST)anomalies and wind stress(τ)anomalies can be constructed by statistical methods,such as singular value decomposition(SVD).In recent years,the applications of artificial intelligence(AI)to climate modeling have shown promising prospects,and the integrations of AI-based models with dynamical models are active areas of research.This study constructs U-Net models for representing the relationship between SSTAs andτanomalies in the tropical Pacific;the UNet-derivedτmodel,denoted asτUNet,is then used to replace the original SVD-basedτmodel of an intermediate coupled model(ICM),forming a newly AI-integrated ICM,referred to as ICM-UNet.The simulation results obtained from ICM-UNet demonstrate their ability to represent the spatiotemporal variability of oceanic and atmospheric anomaly fields in the equatorial Pacific.In the ocean-only case study,theτUNet-derived wind stress anomaly fields are used to force the ocean component of the ICM,the results of which also indicate reasonable simulations of typical ENSO events.These results demonstrate the feasibility of integrating an AI-derived model with a physics-based dynamical model for ENSO modeling studies.Furthermore,the successful integration of the dynamical ocean models with the AI-based atmospheric wind model provides a novel approach to ocean-atmosphere interaction modeling studies.

Comparison of Surface Wind Stress Anomalies over the Tropical Pacific Simulated by an AGCM and by a Simple Atmospheric Model

In this paper, surface wind stress anomalies over the tropical Pacific simulated by an AGCM and by a simple atmospheric model are compared with observed. The AGCM is the higher resolution global spectral model-COLA R40 model and the simple atmospheric model is the atmospheric component of the Cane-Zebiak coupled ocean-atmosphere model.The results show that the wind stress anomalies simulated by both the COLA R40 and the simple model have captured the main features of observation but the x component in the CZ model is closer to that in observation than that in the COLA model, and the correlation coefficients between simulated SSTA from the CZ model and observed for Nino indices are higher than those from the COLA model.

Low Frequency Characteristics of Tropical Pacific Wind Stress Anomalies in Observations and Simulations from a Simple and a Comprehensive Models

Low frequency characteristics of tropical Pacific wind stress anomalies in observation and simulations from the CZ simple atmospheric model and COLA R15 AGCM are analyzed.The results show that ENSO event may be a multi-scale process,that is,ENSO time scale has the period longer than three years; biennial oscillation and annual variability.Dynamical characteristics are involved in the evolution process of wind stress anomaly with ENSO time scale: 1) the development and eastward movement of a cyclonic anomaly circulation in subtropical northwestern Pacific and weakening of Southern Oscillation result in the eastward propagation of westerly anomaly along the equator,therefore,interactions between flows in subtropics and in tropics play an important role in the evolution of wind stress anomaly with ENSO time scale; 2) easterly and westerly anomalies with ENSO time scale are one kind of propagating wave,which differs from Barnett's (1991).It is interesting that the evolution of observed and simulated wind stress anomalies with biennial time scale bears a strong resemble to that with ENSO time scale although their period is different.Observed annual variability is weak during 1979-1981 and intensified after 1981,especially it reaches to maximum during 1982-1984,and the spatial structure of the first mode is the ENSO-like pattern.

Roles of Wind Stress and Subsurface Cold Water in the Second-Year Cooling of the 2017/18 La Nina Event

After the strong 2015/16 El Nino event,cold conditions prevailed in the tropical Pacific with the second-year cooling of the 2017/18 La Ni?a event.Many coupled models failed to predict the cold SST anomalies(SSTAs)in 2017.By using the ERA5 and GODAS(Global Ocean Data Assimilation System)products,atmospheric and oceanic factors were examined that could have been responsible for the second-year cooling,including surface wind and the subsurface thermal state.A time sequence is described to demonstrate how the cold SSTAs were produced in the central-eastern equatorial Pacific in late 2017.Since July 2017,easterly anomalies strengthened in the central Pacific;in the meantime,wind stress divergence anomalies emerged in the far eastern region,which strengthened during the following months and propagated westward,contributing to the development of the second-year cooling in 2017.At the subsurface,weak negative temperature anomalies were accompanied by upwelling in the eastern equatorial Pacific,which provided the cold water source for the sea surface.Thereafter,both the cold anomalies and upwelling were enhanced and extended westward in the centraleastern equatorial Pacific.These changes were associated with the seasonally weakened EUC(the Equatorial Undercurrent)and strengthened SEC(the South Equatorial Current),which favored more cold waters being accumulated in the central-equatorial Pacific.Then,the subsurface cold waters stretched upward with the convergence of the horizontal currents and eventually outcropped to the surface.The subsurface-induced SSTAs acted to induce local coupled air–sea interactions,which generated atmospheric–oceanic anomalies developing and evolving into the second-year cooling in the fall of 2017.

Numerical simulation of the tropical Pacific response to equatorial wind stress anomalies

The results of the tropical Pacific response to the sudden onset of the equatorial wind stress anomalies are discussed. The ocean model is a barotropic, non-linearized one that includes reduced-gravity and an equation for the temperature of the ocean mixed-layer. The experiments are based on a state of equilibrium reached through a long running under the action of annual mean wind stress. There are two kinds of westward wind intensity regions: the whole tropical Pacific and the western tropical Pacific, which are all between latitude 6. 8癗 and 6. 8癝.In these cases, the results show that the positive sea surface temperature (SST) anomalies in the Eastern Pacific and the negative SST anomalies in the Western Pacific are produced, and the positive SST anomalies propagate eastward, just as those observed during the actual El Nino phenomena. The propagations of the Kelvin waves and Rossby waves in the ocean are discussed.Another experiment is also carried out in simulating the process of the decay of El Nino event after the weakened Equatorial Pacific eastern winds returned to normal. The results are similar to the observations, too.

RECONSTRUCTION OF WIND STRESS ANOMALIES SIMULATED BY AN AGCM USING SVD TECHNIQUE

Systematic errors in the COLA R15 AGCM are analyzed by the SVD technique.In order to remove or reduce this kind of error source and reduce climate drift in coupled runs a way in which the wind stress anomalies simulated by an AGCM are reconstructed is proposed by using SVD anal- ysis.Experimental results show that not only wind stress anomalies simulated by an AGCM are obviously improved by reconstructed wind stress anomalies but also this reconstruction has a func- tion of a low-pass time filter,as a result,response of the ZC ocean model to reconstructed wind stress anomalies is more realistic than that of simulated wind stress anomalies by an AGCM.In this paper,application of the hybrid couple ocean-atmosphere model is further discussed.

On the Second-Year Warming in Late 2019 over the Tropical Pacific and Its Attribution to an Indian Ocean Dipole Event

After its maturity,El Niño usually decays rapidly in the following summer and evolves into a La Niña pattern.However,this was not the case for the 2018/19 El Niño event.Based on multiple reanalysis data sets,the space-time evolution and triggering mechanism for the unusual second-year warming in late 2019,after the 2018/19 El Niño event,are investigated in the tropical Pacific.After a short decaying period associated with the 2018/19 El Niño condition,positive sea surface temperature anomalies(SSTAs)re-intensified in the eastern equatorial Pacific in late 2019.Compared with the composite pattern of El Niño in the following year,two key differences are evident in the evolution of SSTAs in 2019.First,is the persistence of the surface warming over the central equatorial Pacific in May,and second,is the re-intensification of the positive SSTAs over the eastern equatorial Pacific in September.Observational results suggest that the re-intensification of anomalous westerly winds over the western and central Pacific,induced remotely by an extreme Indian Ocean Dipole(IOD)event,acted as a triggering mechanism for the second-year warming in late 2019.That is,the IOD-related cold SSTAs in the eastern Indian Ocean established and sustained anomalous surface westerly winds over the western equatorial Pacific,which induced downwelling Kelvin waves propagating eastward along the equator.At the same time,the subsurface ocean provided plenty of warm water in the western and central equatorial Pacific.Mixed-layer heat budget analyses further confirm that positive zonal advection,induced by the anomalous westerly winds,and thermocline feedback played important roles in leading to the second-year warming in late 2019.This study provides new insights into the processes responsible for the diversity of El Niño evolution,which is important for improving the physical understanding and seasonal prediction of El Niño events.