ENSO Model Research Articles

A Physics-Informed Auto-Learning Framework for Developing Stochastic Conceptual Models for ENSO Diversity

Abstract Understanding ENSO dynamics has tremendously improved over past decades. The ENSO diversity in spatial pattern, peak intensity, and temporal evolution is however still poorly represented in conceptual ENSO models. In this paper, a physics-informed auto-learning framework is applied to derive ENSO stochastic conceptual models with varying degrees of freedom. The framework is computationally efficient and easy to apply. Once the state vector of the target model is set, causal inference is exploited to build the right-hand side of the equations based on a mathematical function library. Fundamentally different from standard nonlinear regression, the auto-learning framework provides a parsimonious model by retaining only terms that improve the dynamical consistency with observations. It can also identify crucial latent variables and provide physical explanations. This methodology successfully reconstructs the equations of a realistic six-dimensional reference ENSO model based on the recharge oscillator theory from its data. A hierarchy of lower-dimensional models is derived and their representation of ENSO (including its diversity) systematically assessed. The minimum model that represents ENSO diversity is four-dimensional, with three interannual variables describing the western Pacific thermocline depth, the eastern and central Pacific sea surface temperatures (SSTs), and one intraseasonal variable for westerly wind events. Without the intraseasonal variable, the resulting three-dimensional model underestimates extreme events and is too regular. The limited number of weak nonlinearities in the model are essential in reproducing the observed extreme El Niño events and the observed nonlinear relationship between eastern and western Pacific SSTs.

Southward-shift zonal wind patterns during ENSO in CMIP6 models

Southward-shift zonal wind patterns during ENSO in CMIP6 models

Evaluation and attribution of shortwave feedbacks to ENSO in CMIP6 models

Evaluation and attribution of shortwave feedbacks to ENSO in CMIP6 models

Evaluation of ENSO in CMIP5 and CMIP6 models and its significance in the rainfall in Northeast Thailand

Evaluation of ENSO in CMIP5 and CMIP6 models and its significance in the rainfall in Northeast Thailand

The Subsurface and Surface Indian Ocean Dipoles and Their Association with ENSO in CMIP6 models

The Subsurface and Surface Indian Ocean Dipoles and Their Association with ENSO in CMIP6 models

Investigating decadal variations of the seasonal predictability limit of sea surface temperature in the tropical Pacific

El Niño and the Southern Oscillation (ENSO) have a worldwide impact on seasonal to yearly climate. However, there are decadal variations in the seasonal prediction skill of ENSO in dynamical and statistical models; in particular, ENSO prediction skill has declined since 2000. The shortcomings of models mean that it is very important to study ENSO seasonal predictability and its decadal variation using observational/reanalysis data. Here we quantitatively estimate the seasonal predictability limit (PL) of ENSO from 1900 to 2015 using Nonlinear local Lyapunov exponent (NLLE) theory with an observational/reanalysis dataset and explore its decadal variations. The mean PL of sea surface temperature (SST) is high in the central/eastern tropical Pacific and low in the western tropical Pacific, reaching 12–15 and 7–8 months, respectively. The PL in the tropical Pacific varies on a decadal timescale, with an interdecadal standard deviation of up to 2 months in the central tropical Pacific that has similar spatial structure to the mean PL. Taking the PL of SST in the Niño 3.4 region as representative of the PL in the central/eastern tropical Pacific, there are clearly higher values in the 1900s, mid-1930s, mid-1960s, and mid-1990s, and lower values in the 1920s, mid-1940s, and mid-2010s. Meanwhile, the PL of SST in the Niño 6 region—whose average value is 7 months—is in good agreement with the PL of most regions in the western tropical Pacific, with higher values in the 1910s, 1940s, and 1980s and lower values in the 1930s, 1950s, and mid-1990s. In the framework of NLLE theory, the PL is determined by the error growth rate (representing the dissipation rate of the predictable signal) and the saturation value of relative error (representing predictable signal intensity). We reveal that the spatial structure of the mean PL in the tropical Pacific is determined mainly by the error growth rate. The decadal variability of PL is affected more by the variation of the saturation value of relative error in the equatorial Pacific, whereas the error growth rate cannot be ignored in the PL of some regions. As an important source of predictability in ENSO dynamics, the relationship between warm water volume and SST in the Niño 3.4 region has a critical role in the decadal variability of PL in the tropical Pacific through the error growth rate and saturation value of relative error. This strong relationship reduces the error growth rate in the initial period and increases the saturated relative error, contributing to the high PL.

A New Way of Evaluating ENSO in Climate Models: The CLIVAR ENSO Metrics Package

A New Way of Evaluating ENSO in Climate Models: The CLIVAR ENSO Metrics Package

ENSO Dynamics in the E3SM-1-0, CESM2, and GFDL-CM4 Climate Models

AbstractThis study examines historical simulations of ENSO in the E3SM-1-0, CESM2, and GFDL-CM4 climate models, provided by three leading U.S. modeling centers as part of the Coupled Model Intercomparison Project phase 6 (CMIP6). These new models have made substantial progress in simulating ENSO’s key features, including: amplitude; timescale; spatial patterns; phase-locking; spring persistence barrier; and recharge oscillator dynamics. However, some important features of ENSO are still a challenge to simulate. In the central and eastern equatorial Pacific, the models’ weaker-than-observed subsurface zonal current anomalies and zonal temperature gradient anomalies serve to weaken the nonlinear zonal advection of subsurface temperatures, leading to insufficient warm/cold asymmetry of ENSO’s sea surface temperature anomalies (SSTA). In the western equatorial Pacific, the models’ excessive simulated zonal SST gradients amplify their zonal temperature advection, causing their SSTA to extend farther west than observed. The models underestimate both ENSO’s positive dynamic feedbacks (due to insufficient zonal wind stress responses to SSTA) and its thermodynamic damping (due to insufficient convective cloud shading of eastern Pacific SSTA during warm events); compensation between these biases leads to realistic linear growth rates for ENSO, but for somewhat unrealistic reasons. The models also exhibit stronger-than-observed feedbacks onto eastern equatorial Pacific SSTAs from thermocline depth anomalies, which accelerates the transitions between events and shortens the simulated ENSO period relative to observations. Implications for diagnosing and simulating ENSO in climate models are discussed.

The Importance of Central Pacific Meridional Heat Advection to the Development of ENSO

AbstractThe relationship between the Warm Water Volume (WWV) ENSO precursor and ENSO SST weakened substantially after ~2000, coinciding with a degradation in dynamical model ENSO prediction skill. It is important to understand the drivers of the equatorial thermocline temperature variations and their linkage to ENSO onsets. In this study, a set of ocean reanalyses is employed to assess factors responsible for the variation of the equatorial Pacific Ocean thermocline during 1982-2019. Off-equatorial thermocline temperature anomalies carried equatorward by the mean meridional currents associated with Pacific Tropical Cells are shown to play an important role in modulating the central equatorial thermocline variations, which is rarely discussed in the literature. Further, ENSO events are delineated into two groups based on precursor mechanisms: the western equatorial type (WEP) ENSO, when the central equatorial thermocline is mainly influenced by the zonal propagation of anomalies from the western Pacific, and the off-equatorial central Pacific (OCP) ENSO, when off-equatorial central thermocline anomalies play the primary role. WWV is found to precede all WEP ENSO by 6-9 months, while the correlation is substantially lower for OCP ENSO events. In contrast, the central tropical Pacific (CTP) precursor, which includes off-equatorial thermocline signals, has a very robust lead correlation with the OCP ENSO. Most OCP ENSO events are found to follow the same ENSO conditions, and the number of OCP ENSO increases substantially since the 21st century. These results highlight the importance of monitoring off-equatorial subsurface preconditions for ENSO prediction and to understand multi-year ENSO.

Improving forecasts of El Ni\xf1o diversity: a nonlinear forcing singular vector approach

Observations indicate that two types of El Niño events exist: one is the EP-El Niño with a warming center in the eastern tropical Pacific, and the other is the CP-El Niño with large positive SST anomalies in the central tropical Pacific. Most current numerical models are not able to accurately identify the different types of El Niño. The present study examines the dynamic properties of the ENSO forecast system NFSV-ICM which combines an intermediate-complexity ENSO model (ICM) with a nonlinear forcing singular vector (NFSV)-based tendency perturbation forecast model. This system is able to distinguish the different types of El Niño in predictions. Hindcasts show that the NFSV-ICM system is able to capture the horizontal distribution of the SST anomalies and their amplitudes in the mature phase of not only EP-El Niño events but also CP-El Niño events. The NFSV-ICM is also able to describe the evolution of SST anomalies associated with the two types of El Niño up to at least two-season lead times, while the corresponding forecasts with the ICM are limited to, at most, one-season lead times. These improvements are associated with the modifications of the atmospheric and ocean processes described by the ICM through the NFSV-based tendency perturbations. In particular, the thermocline and zonal advection feedback are strongly modified, and the conditions of the emergence of both EP- and CP-El Niño events are improved. The NFSV-ICM therefore provides a useful platform for studying ENSO dynamics and predictability associated with El Niño diversities.

The role of natural factors (part 2): Indian summer monsoon in climate change period\u2014observation and CMIP5 models

This study discusses the role of natural factors and related teleconnections for Indian summer monsoon (ISM) with a special emphasis on later two decades of the last century. The combined influence of the sun and volcanos on ISM is examined using observational data as well as CMIP5 model outputs. Possible mechanisms relating to a disruption of the usual ENSO-ISM teleconnection for those decades are explored. Observation suggested that the regional Hadley circulation, via the NAO in the northern hemisphere and Indian Ocean Dipole in the southern hemisphere, may have a role in the change in ISM behaviour. Such features though captured well in the observation are shown missing in models. Additionally, it indicates that differences among models mainly originate in a regional level, which could be due to inconsistency in representing regional teleconnection features. Interestingly, all models perform reasonably well in terms of global thermodynamic scaling arguments. The overall study underpins important areas, where natural factors influence regional climate, but models miss out and suggest discrepancies among each other. Such knowledge has major implications in regional as well as global scale. The modelling community will also greatly benefit by an improved representation of ENSO and ISM in models.

Linear or Nonlinear Modeling for ENSO Dynamics?

The observed ENSO statistics exhibits a non-Gaussian behavior, which is indicative of the presence of nonlinear processes. In this paper, we use the Recharge Oscillator Model (ROM), a largely used Low-Order Model (LOM) of ENSO, as well as methodologies borrowed from the field of statistical mechanics to identify which aspects of the system may give rise to nonlinearities that are consistent with the observed ENSO statistics. In particular, we are interested in understanding whether the nonlinearities reside in the system dynamics or in the fast atmospheric forcing. Our results indicate that one important dynamical nonlinearity often introduced in the ROM cannot justify a non-Gaussian system behavior, while the nonlinearity in the atmospheric forcing can instead produce a statistics similar to the observed. The implications of the non-Gaussian character of ENSO statistics for the frequency of extreme El Niño events is then examined.

Idealized Experiments for Optimizing Model Parameters Using a 4D-Variational Method in an Intermediate Coupled Model of ENSO

Large biases exist in real-time ENSO prediction, which can be attributed to uncertainties in initial conditions and model parameters. Previously, a 4D variational (4D-Var) data assimilation system was developed for an intermediate coupled model (ICM) and used to improve ENSO modeling through optimized initial conditions. In this paper, this system is further applied to optimize model parameters. In the ICM used, one important process for ENSO is related to the anomalous temperature of subsurface water entrained into the mixed layer (Te), which is empirically and explicitly related to sea level (SL) variation. The strength of the thermocline effect on SST (referred to simply as “the thermocline effect”) is represented by an introduced parameter, αTe. A numerical procedure is developed to optimize this model parameter through the 4D-Var assimilation of SST data in a twin experiment context with an idealized setting. Experiments having their initial condition optimized only, and having their initial condition plus this additional model parameter optimized, are compared. It is shown that ENSO evolution can be more effectively recovered by including the additional optimization of this parameter in ENSO modeling. The demonstrated feasibility of optimizing model parameters and initial conditions together through the 4D-Var method provides a modeling platform for ENSO studies. Further applications of the 4D-Var data assimilation system implemented in the ICM are also discussed.

A theoretical model of strong and moderate El Niño regimes

The existence of two regimes for El Nino (EN) events, moderate and strong, has been previously shown in the GFDL CM2.1 climate model and also suggested in observations. The two regimes have been proposed to originate from the nonlinearity in the Bjerknes feedback, associated with a threshold in sea surface temperature ( $$T_c$$ ) that needs to be exceeded for deep atmospheric convection to occur in the eastern Pacific. However, although the recent 2015–16 EN event provides a new data point consistent with the sparse strong EN regime, it is not enough to statistically reject the null hypothesis of a unimodal distribution based on observations alone. Nevertheless, we consider the possibility suggestive enough to explore it with a simple theoretical model based on the nonlinear Bjerknes feedback. In this study, we implemented this nonlinear mechanism in the recharge-discharge (RD) ENSO model and show that it is sufficient to produce the two EN regimes, i.e. a bimodal distribution in peak surface temperature (T) during EN events. The only modification introduced to the original RD model is that the net damping is suppressed when T exceeds $$T_c$$ , resulting in a weak nonlinearity in the system. Due to the damping, the model is globally stable and it requires stochastic forcing to maintain the variability. The sustained low-frequency component of the stochastic forcing plays a key role for the onset of strong EN events (i.e. for $$T>T_c$$ ), at least as important as the precursor positive heat content anomaly (h). High-frequency forcing helps some EN events to exceed $$T_c$$ , increasing the number of strong events, but the rectification effect is small and the overall number of EN events is little affected by this forcing. Using the Fokker–Planck equation, we show how the bimodal probability distribution of EN events arises from the nonlinear Bjerknes feedback and also propose that the increase in the net feedback with increasing T is a necessary condition for bimodality in the RD model. We also show that the damping strength determines both the adjustment time-scale and equilibrium value of the ensemble spread associated with the stochastic forcing.

The ENSO Transition Probabilities

ENSO is investigated here by considering it as a transition from different states. Transition probability matrices can be defined to describe the evolution of ENSO in this way. Sea surface temperature anomalies are classified into four categories, or states, and the probability to move from one state to another has been calculated for both observations and a simulation from a GCM. This could be useful for understanding and diagnosing general circulation models elucidating the mechanisms that govern ENSO in models. Furthermore, these matrices have been used to define a predictability index of ENSO based on the entropy concept introduced by Shannon. The index correctly identifies the emergence of the spring predictability barrier and the seasonal variations of the transition probabilities. The transition probability matrices could also be used to formulate a basic prediction model for ENSO that was tested here on a case study.

An Agro Climate Study - Atmospheric Teleconnection

The year climate signal of the equatorial Pacific Ocean sea surface pattern in association with the Southern Oscillation has been responsible in altering weather systems in impacting the aberrated climate scenario across the globe. A comparison of Normalized Difference Vegetation Index(NDVI) pattern over the five selected regions in India with the derived moisture adequacy from the water balance model and multivariate ENSO index unfolded the phenological feature of greenness up and down with a lag. It is assumed that a 60% of moisture adequacy is essential for the sustenance of crop or vegetation growth and development and can be traced in metric of NDVI. A statistical model is suggested for All India rice yield and is of useful in agrometeorological advisories.

The effects of remote SST forcings on ENSO dynamics, variability and diversity

Air-sea interactions with remote regions in the tropical Indian and Atlantic, and extra-tropical oceans can influence ENSO features in the tropical Pacific. In this study these effects are explored by using an AGCM coupled with a Slab Ocean and a simple recharge oscillator ENSO model through switched on/off air-sea interaction in respective ocean area. It is shown that the decoupling in different remote regions has different impacts on ENSO dynamics, variability and diversity. The most interesting result is that the air-sea interactions with remote tropical oceans provide a delayed negative feedback to ENSO similar to that of the tropical Pacific Ocean internal wave dynamics. This is caused by the ENSO teleconnections: they lead to a delayed remote warming and cooling, which in turn feedbacks to ENSO effectively giving a delayed negative feedback. The model simulations suggest that this remote delayed feedback may contribute about 40% to the total delayed negative feedback of ENSO. Thus a central element of ENSO dynamics is partly due to interactions with other tropical ocean basins by atmospheric teleconnections. Furthermore, all remote regions effectively provide stochastic forcings for the ENSO variability and therefore increase the ENSO variability. The influence from the remote regions also causes different patterns of sea surface temperature (SST) variability in the tropical Pacific, contributing to the diversity of the ENSO mode. In particular the extra-tropical Pacific regions force SST variability that is different from the equatorial ENSO mode of variability. The influence that the remote regions have on the ENSO dynamics and variability is significantly altered by the interaction between the equatorial recharge oscillator dynamics and the simple thermodynamic slab ocean processes.

Drivers of coupled model ENSO error dynamics and the spring predictability barrier

Despite recent improvements in ENSO simulations, ENSO predictions ultimately remain limited by error growth and model inadequacies. Determining the accompanying dynamical processes that drive the growth of certain types of errors may help the community better recognize which error sources provide an intrinsic limit to predictability. This study applies a dynamical analysis to previously developed CCSM4 error ensemble experiments that have been used to model noise-driven error growth. Analysis reveals that ENSO-independent error growth is instigated via a coupled instability mechanism. Daily error fields indicate that persistent stochastic zonal wind stress perturbations $$(\tau_{x}^{\prime } )$$ near the equatorial dateline activate the coupled instability, first driving local SST and anomalous zonal current changes that then induce upwelling anomalies and a clear thermocline response. In particular, March presents a window of opportunity for stochastic $$\tau_{x}^{\prime }$$ to impose a lasting influence on the evolution of eastern Pacific SST through December, suggesting that stochastic $$\tau_{x}^{\prime }$$ is an important contributor to the spring predictability barrier. Stochastic winds occurring in other months only temporarily affect eastern Pacific SST for 2–3 months. Comparison of a control simulation with an ENSO cycle and the ENSO-independent error ensemble experiments reveals that once the instability is initiated, the subsequent error growth is modulated via an ENSO-like mechanism, namely the seasonal strength of the Bjerknes feedback. Furthermore, unlike ENSO events that exhibit growth through the fall, the growth of ENSO-independent SST errors terminates once the seasonal strength of the Bjerknes feedback weakens in fall. Results imply that the heat content supplied by the subsurface precursor preceding the onset of an ENSO event is paramount to maintaining the growth of the instability (or event) through fall.

The climatology and interannual variability of the South Asia high and its relationship with ENSO in CMIP5 models

The present study examines climatology and interannual variability of South Asian high (SAH) and its connection with the ENSO based on 38 coupled models from the Coupled Model Intercomparison Project phase 5 (CMIP5). Results show that multi-model ensemble (MME) can reasonably capture the climatological spatial pattern of the SAH, although its intensity is slightly underestimated. The CCSM4, CESM1-BGC and CESM1-FASTCHEM can well simulate the climatological location and intensity of the SAH. The interannual variability of the SAH is investigated by calculating ratio of the standard deviation of the ten parameters in models with those in observations. The results indicate that the MME can reasonably capture magnitudes of the interannual variability of the area index, intensity index, and longitude of the SAH center. Quasi-4-year period of the SAH intensity index can be well simulated by CMCC-CESM, CMCC-CMS and GFDL-ESM2G, and quasi-5-year period of north–south movement index can be captured by CanCM4, CESM1-CAM5, CESM1-FASTCHEM, CNRM-CM5-2, GFDL-ESM2G and HadCM3. Furthermore, MME can reasonably reproduce seasonal evolution of intensity and location of the SAH except for its east–west movement. The ENSO-SAH relationship is further evaluated. It is found that about two-thirds of the CMIP5 models can capture the observed ENSO-SAH relationship, although the relationship is distinctly exaggerated by several models. The success of these models is attributed to the reasonable simulation of both the “charge” process over the tropical Indian Ocean induced by the ENSO-related anomalous sea surface temperature (SST) over the tropical eastern Pacific (TEP) and longitude extension of the western boundary of the ENSO-related anomalous SST over the TEP.

Future Changes in Atmosphere Teleconnection over East Asia and North Pacific associated with ENSO in CMIP5 Models

Future Changes in Atmosphere Teleconnection over East Asia and North Pacific associated with ENSO in CMIP5 Models