El Nino-Southern Oscillation Prediction Research Articles

A Hybrid Adaptive Covariance Inflation Method for EnKF-Based ENSO Prediction

Abstract The ensemble-based data assimilation method is usually used for the initialization of El Niño–Southern Oscillation (ENSO) prediction. Because of sampling errors caused by a finite ensemble, imperfect physical parameterizations, and other factors, the multiplicative covariance inflation method is commonly employed in the standard ensemble Kalman filter (EnKF) to increase the prior variance and alleviate filter divergence. Given computational resource constraints, utilizing larger ensemble sizes to minimize sampling errors in high-dimensional oceanic or atmospheric models poses a challenge. The authors propose a new hybrid adaptive covariance inflation scheme in small ensembles and apply this method to an intermediate coupled model (ICM) used at the Institute of Oceanology, Chinese Academy of Sciences (IOCAS), named the IOCAS ICM for ENSO prediction. Hybrid refers to performing both prior and posterior inflation. Results show that the hybrid t-X adaptive inflation scheme performs best within the ICM framework, which can reduce the analysis errors by 46% for the daily SST anomaly compared to the standard EnKF algorithm using the fixed multiplicative covariance inflation factor. The t-X adaptive algorithm enhances the standard EnKF’s forecasting ability by optimizing the initial forecast field and improving internal model errors. This method notably improves the prediction skill for the Niño-1 + 2 SST anomaly, particularly in phase transitions. Regarding SST anomaly prediction, the advantages of the hybrid t-X adaptive method over the standard EnKF scheme mainly occur in the equatorial eastern Pacific and south boundaries of the ICM.

The Unique Role of the Intraseasonal Zonal Wind in March Over the Equatorial Western Pacific Contributes to Shaping the Subsequent ENSO Development

AbstractThe predictability of the El Niño‐Southern Oscillation (ENSO) is significantly limited by the spring predictability barrier (SPB). Our observational analysis suggests that considering high‐frequency wind components during the spring season may mitigate the SPB impact on ENSO predictability. The intraseasonal zonal wind in March over the equatorial western Pacific initially perturbs the sea surface temperature (SST). It contributes nearly 40% of the east‐west SST gradient after 2–3 months, ranking first among other calendar months. This significant contribution causes March to become the earliest month to effectively indicate the following ENSO direction due to the active eastward‐propagating Madden‐Julian oscillation (MJO) activity. Through the obvious variation of the eastward‐propagating MJO speed, it also shows the possible close relationship between the mean SST state and the interdecadal variability in intraseasonal zonal wind. Additionally, the current strong variability of intraseasonal zonal wind suggests the important role of atmospheric information in recent ENSO development.

Process-Based Evaluation of Intraseasonal Oceanic Kelvin Waves in the Pacific Ocean in CMIP6 Models

Abstract Oceanic intraseasonal Kelvin waves (KWs) help modulate upper-ocean thermal characteristics, providing feedbacks to important coupled air–sea phenomena in the tropics. The recent availability of daily thermocline depth fields from several phase 6 of the Coupled Model Intercomparison Project (CMIP6) models makes it possible to evaluate the performance of KWs and identify potential sources of bias. Most models fail to simulate a realistic spatial distribution of KW variability. Models simulate a large variability of KWs in the western or eastern Pacific rather than in the central Pacific as observed. The modeled KWs propagate slowly (about 1.5 m s−1) compared to observations (about 2.5 m s−1). This slow propagation is also identified in wavenumber–frequency spectra for KWs and meridional KW structures, which is more consistent with a second baroclinic mode structure in models compared to the first baroclinic mode structure in observations. An analysis of the relative contributions of the vertical wavenumber and background ocean stability to KW phase speeds indicates that the high vertical wavenumber bias in models contributes most to the slow propagation, in which the higher-than-observed vertical wavenumbers imply the biased incorporation of higher baroclinic modes in the model KW structure. This finding is further supported by the results of vertical mode decomposition that incorporates background density profiles. These results indicate that a realistic representation of the KW vertical structure is essential to produce realistic KW propagations in models. Significance Statement Oceanic intraseasonal Kelvin waves (KWs) play a significant role in regulating the heat content and temperature of the ocean, which provides feedbacks to coupled ocean–atmosphere phenomena in the tropics such as El Niño–Southern Oscillation (ENSO). Consequently, identifying the biases in KWs and the potential sources of those biases in state-of-the-art models is essential to improve simulations of ENSO and its diversity and advance forecasts of weather extremes around the globe induced by ENSO. We find that KWs in the models propagate slower than observations, mainly due to biases in their vertical structure, with secondary effects due to biases in model ocean stability. These issues may be potential sources for current imperfect ENSO model simulations and predictions.

Revisiting the Role of Atmospheric Initial Signals in Predicting ENSO

Abstract El Niño–Southern Oscillation (ENSO) provides an important source of global seasonal–interannual predictability, while its prediction encounters bottlenecks. Besides the slow-varying air–sea feedbacks, high-frequency atmospheric signals (HFASs) act on ENSO evolution. This study revisits the role of atmospheric initial signals in ENSO prediction using an atmosphere–ocean coupled model. Two sets of sensitivity hindcasts are conducted. One utilizes a pure sea surface temperature (SST) nudging for initialization so that no observed atmospheric internal signal is assimilated. The other applies a combination of the SST nudging and spectral nudging of JRA-55 reanalysis, which not only assimilates the observed atmospheric states and hence realizes skillful predictions of HFAS at the initial stage but also improves the oceanic initial conditions (ICs), especially around the thermocline. Unexpectedly, the better atmosphere–ocean ICs neither improve ENSO prediction nor overcome the spring prediction barrier. Further analysis of Bjerknes stability index suggests that underestimated negative feedbacks and insufficient responses of ocean currents and thermocline slope to atmospheric internal winds may account for the failure. Nevertheless, assimilating the atmospheric information partly improves the prediction of El Niño onset, especially for two recent extreme cases (i.e., 1997/98 and 2015/16). This improvement is associated with better representations of initial westerly wind bursts (WWBs) that excite subsequent downwelling equatorial Kelvin waves. However, due to unpredictable WWBs and underestimated wind response to the SST warming owing to the cold tongue biases in boreal summer, the zonal advective feedbacks are underestimated and thus further development of El Niño cannot be well predicted. This study suggests the importance of initial atmospheric signals despite the limitation, prompting further efforts to improve model physics.

ENSO predictive analytics based on layered rendering

El Niño-Southern Oscillation (ENSO) is a periodic climate phenomenon in the equatorial Pacific that significantly influences global climate patterns. Accurate prediction and monitoring of ENSO events are essential for meteorological agencies and governmental institutions. This study introduces a content-guided attention module within a convolutional neural network to improve prediction accuracy. This module models inter-channel relationships and enhances information interaction by integrating channel and spatial attention weights. These advancements substantially improve prediction accuracy and help overcome the spring prediction barrier in ENSO forecasting. The research emphasizes global feature modeling and proposes a novel content-guided ENSO prediction model. It also includes an ocean data generation model utilizing global attention. Furthermore, a layered rendering technique is employed to invert ocean data, facilitating detailed analysis and contributing to the development of an ocean synthetic dataset.

A probabilistic forecast for multi-year ENSO using Bayesian convolutional neural network

Abstract A robust El Niño Southern Oscillation (ENSO) prediction is essential for monitoring the global climate, regional monsoons, and weather extremes. Despite dedicated efforts spanning decades, the precise prediction of ENSO events through numerical modeling beyond a couple of seasonal lead times remains a daunting challenge. The advent of deep learning-based approaches marks a transformative era in climate and weather prediction. However, many machine learning-based studies attempting ENSO prediction are confined to singular estimates, lacking adequate quantification of uncertainty in learned parameters and overlooking the crucial need for a nuanced understanding of ENSO prediction confidence. Here, we introduce a deep learning-based Bayesian convolutional neural network model that provides robust probabilistic predictions for ENSO with a lead time of up to 9–10 months across all seasons. The Bayesian layers within the convolutional neural network maintain the capability to predict a distribution of learned parameters. Augmented with bias correction, our model reproduces the amplitude of the Niño 3.4 index with fidelity for lead up to 9–10 months. The inherent capacity for uncertainty modeling enhances the reliability of bayesian neural networks (BNNs), making them particularly valuable in operational services. This research holds substantial socio-economic implications as it enhances our forecasting capabilities and rigorously quantifies forecast uncertainties, providing valuable insights for planning and policymaking.

Contribution of El Niño Southern Oscillation (ENSO) Diversity to Low‐Frequency Changes in ENSO Variance

AbstractEl Niño Southern Oscillation (ENSO) diversity is characterized based on the longitudinal location of maximum sea surface temperature anomalies (SSTA) and amplitude in the tropical Pacific, as Central Pacific events are typically weaker than Eastern Pacific events. SSTA pattern and intensity undergo low‐frequency modulations, affecting ENSO prediction skill and remote impacts, and resulting in low‐frequency changes in ENSO variance. Yet, how different ENSO types contribute to these decadal variance changes remains unclear. Here, we decompose the low‐frequency changes of ENSO variance into contributions from ENSO diversity categories. We propose a fuzzy clustering of monthly SSTA to allow for non‐binary event category memberships, where each event can belong to different clusters. Our approach identifies two La Niña and three El Niño categories and shows that the major shift of ENSO variance in the mid‐1970s was associated with an increasing likelihood of strong La Niña and extreme El Niño events.

Asymmetric Efficacies Between Warm and Cold Pacific Meridional Modes in Inducing ENSO

AbstractThis study investigates boreal spring events of Pacific Meridional Mode (PMM) from 1950 to 2022, revealing that cold PMM is more effective in triggering subsequent La Niña compared to warm PMM's induction of following El Niño. This asymmetry stems from the varying origins and sub‐efficacies of PMM groups. The cold PMM is primarily initiated by pre‐existing La Niña, while the warm PMM is comparably activated by pre‐existing El Niño and internal atmospheric dynamics. PMMs initiated by pre‐existing El Niño or La Niña play a crucial role in determining the efficacies of PMMs in triggering subsequent El Niño‐Southern Oscillation (ENSO). The strong discharge of pre‐existing El Niño hampers warm PMM's induction of subsequent El Niño, whereas weak recharge from pre‐existing La Niña enhances the efficacy of cold PMM in inducing subsequent La Niña. Comprehending not only the PMM phase but also its origin is crucial for ENSO research and prediction.

ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific

AbstractRecent studies have demonstrated great values of deep‐learning (DL) methods for improving El Niño‐Southern Oscillation (ENSO) predictions. However, the black‐box nature of DL makes it challenging to physically interpret mechanisms responsible for successful ENSO predictions. Here, we demonstrate an interpretable method by performing perturbation experiments to predictors and quantifying input‐output relationships in predictions by using a transformer‐based model; ENSO‐related thermal precursors serving as initial conditions during multi‐month time intervals (TIs) are identified in the equatorial‐northern Pacific, acting to precondition input predictors to provide for long‐lead ENSO predictability. Results reveal the existence of upper‐ocean temperature anomaly pathways and consistent phase propagations of thermal precursors around the tropical Pacific. It is illustrated that three‐dimensional thermal fields and their basinwide evolution during long TIs act to enhance long‐lead prediction skills of ENSO. These physically explainable results indicate that neural networks can adequately represent predictable precursors in the input predictors for successful ENSO predictions.

Strong El Niño Events Lead to Robust Multi‐Year ENSO Predictability

AbstractThe El Niño‐Southern Oscillation (ENSO) phenomenon—the dominant source of climate variability on seasonal to multi‐year timescales—is predictable a few seasons in advance. Forecast skill at longer multi‐year timescales has been found in a few models and forecast systems, but the robustness of this predictability across models has not been firmly established owing to the cost of running dynamical model predictions at longer lead times. In this study, we use a massive collection of multi‐model hindcasts performed using model analogs to show that multi‐year ENSO predictability is robust across models and arises predominantly due to skillful prediction of multi‐year La Nina events following strong El Niño events.

Adaptive Graph Spatial-Temporal Attention Networks for long lead ENSO prediction

Adaptive Graph Spatial-Temporal Attention Networks for long lead ENSO prediction

On the Divergent Evolution of ENSO After the Coastal El Niños in 2017 and 2023

AbstractCoastal El Niño is an extreme situation of El Niño‐Southern Oscillation (ENSO) with sea surface temperature warming confined in the far‐eastern equatorial Pacific. Some coastal El Niños evolve into a basin scale El Niño, and some don't, implying a diversity in ENSO evolutions after a coastal El Niño event. In this study, the coastal El Niños in 2017 and 2023 are selected to examine their subsequent evolution. Both coastal El Niños developed after a La Niña, with the former followed by a La Niña and the latter by a basin‐scale El Niño. The cold (warm) subsurface temperatures in 2017 (2023) were key factors leading to the divergent ENSO evolution. Convection over the western tropical Pacific and the atmospheric circulation anomalies across the equatorial Pacific also contributed to the differences. Model predictions suggest that differences in ENSO evolution after a coastal El Niño are associated with differences in ENSO predictability.

Summary of Commonly Used ENSO Indices

This paper introduces 12 commonly used El Niño-Southern Oscillation (ENSO) indices and compares their creation based on different datasets, regions, representing anomalies, functions, advantages, and limits. It summarizes two application cases for each of the three different ENSO indices: Niño 3.4, Ocean Niño Index (ONI), and the Multivariate ENSO index (MEI). The application examples for Niño 3.4 are ENSO prediction and drought pattern analysis; for ONI, they are responses of ice shelves and rainfalls to ENSO; and for MEI, they are building two modeling approaches for ENSO monitoring and hydrological events prediction. The summary of these ENSO indices demonstrates their importance in climate studies in providing simplified anomalies for building models and making scientists communicate with each other about their findings in ENSO-related climate variabilities in agriculture, fisheries, water resource management, disaster preparedness, etc.

Efficient inference and learning of a generative model for ENSO predictions from large multi-model datasets

Historical simulations of global sea-surface temperature (SST) from the fifth phase of the Coupled Model Intercomparison Project (CMIP5) are analyzed. A state-of-the-art deep learning approach is applied to provide a unified access to the diversity of simulations in the large multi-model dataset in order to go beyond the current technological paradigm of ensemble averaging. Based on the concept of a variational auto-encoder (VAE), a generative model of global SST is proposed in combination with an inference model that aims to solve the problem of determining a joint distribution over the data generating factors. With a focus on the El Niño Southern Oscillation (ENSO), the performance of the VAE-based approach in simulating various central features of observed ENSO dynamics is demonstrated. A combination of the VAE with a forecasting model is proposed to make predictions about the distribution of global SST and the corresponding future path of the Niño index from the learned latent factors. The proposed ENSO emulator is compared with historical observations and proves particularly skillful at reproducing various aspects of observed ENSO asymmetry between the two phases of warm El Niño and cold La Niña. A relationship between ENSO asymmetry and ENSO predictability is identified in the ENSO emulator, which improves the prediction of the simulated Niño index in a number of CMIP5 models.

Skillful prediction of length of day one year ahead in multiple decadal prediction systems

Despite a small amplitude, Length of Day (LOD) change, which varies from one year to another due to changes in Atmospheric Angular Momentum (AAM), determines the accuracy of Global Positioning System (GPS) time calculation. In this study, we examine the prediction skill of LOD and AAM in nine decadal prediction systems archived for the Decadal Climate Prediction Project. A persistence and rebound in LOD prediction skill at one year or longer lead time is found in most models. A poleward propagation of AAM anomaly via wave-mean flow interaction is also qualitatively well reproduced. This long-lead prediction of LOD and AAM is attributed to reliable predictions of the El Niño–Southern Oscillation (ENSO) and the Quasi-Biennial Oscillation (QBO), the former being more systematically related than the latter. This result indicates that the improved ENSO prediction and atmospheric wave-mean flow interaction may lead to better prediction of LOD, AAM and related extratropical climate in the decadal prediction systems.

Quantifying the Relative Contributions of the Global Oceans to ENSO Predictability With Deep Learning

AbstractWe propose a unified statistical method based on deep learning and analysis to quantify the relative contributions of the global oceans to El Niño–Southern Oscillation (ENSO) predictability. By incorporating subsurface signals in the Indian Ocean and Atlantic, the forecast lead can be skillfully extended by about one season. This skill enhancement mainly originates from the tropical Indian Ocean, presumably related to signals of the Indian Ocean Dipole passing to the tropical Pacific through the Indonesian Throughflow. The sea surface temperature anomaly (SSTA) in the Indian Ocean accounts for nearly 50% of surface contributions to both El Niño and La Niña predictions at a 15‐month lead. The north tropical Atlantic SSTA has a moderate impact on La Niña at a 9‐month lead. The Pacific Meridional Mode plays a significant role in both ENSO phases at a 12‐month lead. Thus, our study suggests that trans‐basin effects for ENSO are more vigorous than previously thought.

Multi-scale spatial features and temporal attention mechanisms: advancing the accuracy of ENSO prediction

The exponential progression in oceanic observational technology has fostered the accumulation of substantial time series data pivotal for predictions in ocean meteorology. Foremost among the phenomena observed is El Niño-Southern Oscillation (ENSO), a critical determinant in the interplay of global ocean atmosphere interactions, with its severe manifestations inducing extreme meteorological conditions. Therefore, precisely predicting ENSO events carries immense gravitas. Historically, predictions hinged primarily on dynamic models and statistical approaches; however, the intricate and multifaceted spatiotemporal dynamics of ENSO events have often impeded the accuracy of these traditional methodologies. A notable lacuna in contemporary research is the insufficient exploration of long-term dependencies within oceanic data and the suboptimal integration of spatial information derived from spatiotemporal data. To address these limitations, this study introduces a forward-thinking ENSO prediction framework synergizing multiscale spatial features with temporal attention mechanisms. This innovation facilitates a more profound exploration of temporal and spatial domains, enhancing the retention of extensive-period data while optimizing the use of spatial information. Preliminary analyses executed on the global ocean data assimilation system dataset attest to the superior efficacy of the proposed method, underscoring a substantial improvement over established methods including SA-convolutional long short-term memory, particularly in facilitating long-term predictions.The source code and datasets are provided. The code is available at https://github.com/tse1998/ENSO-prediction.

Enhancing ENSO predictions with self-attention ConvLSTM and temporal embeddings

El Niño-Southern Oscillation (ENSO), a cyclic climate phenomenon spanning interannual and decadal timescales, exerts substantial impacts on the global weather patterns and ecosystems. Recently, deep learning has brought considerable advances in the accurate prediction of ENSO occurrence. However, the current models are insufficient to characterize the evolutionary behavior of the ENSO, particularly lacking comprehensive modeling of local-range and longrange spatiotemporal interdependencies, and the incorporation of calendar monthly and seasonal properties. To make up this gap, we propose a Two-Stage SpatioTemporal (TSST) autoregressive model that couples the meteorological factor prediction with ENSO indicator prediction. The first stage predicts the meteorological time series by leveraging self-attention ConvLSTM network which captures both the local and the global spatial-temporal dependencies. The temporal embeddings of calendar months and seasonal information are further incorporated to preserves repeatedly-occurring-yet-hidden patterns in meteorological series. The second stage uses multiple layers to extract higher level of features from predicted meteorological factors progressively to generate ENSO indicators. The results demonstrate that our model outperforms the state-of-the-art ENSO prediction models, effectively predicting ENSO up to 24 months and mitigating the spring predictability barrier.

Evaluating the Performance of the North Pacific Victoria Mode as an ENSO Predictor Based on Multimodel Ensemble Hindcasts

Abstract The Victoria mode (VM), similar to the Pacific meridional mode (PMM), is forced by North Pacific Oscillation atmospheric variability. Both the boreal spring VM and PMM can trigger the onset of El Niño–Southern Oscillation (ENSO) events in the following winter. Previous studies have examined the precursor relationship between the PMM and ENSO based on a subset of models drawn from the North American Multimodel Ensemble (NMME) system. They suggested that the PMM can act as a precursor to El Niño events, whereas it fails to predict La Niña events. Utilizing the hindcasts of these models from NMME, this study further investigates the role of the VM as an ENSO predictor to examine the real usefulness of the VM for ENSO prediction. When compared with the PMM, the VM can predict both El Niño and La Niña events with some skill, showing that the VM seems to be a more reliable predictor of ENSO. We found that the unique role of the VM in ENSO prediction originates from the symmetric impact of the VM on ENSO events. The VM, as a basin-scale sea surface temperature (SST) pattern, combines the role of the SST over the subtropical northeastern Pacific that is similar to the PMM in initializing El Niño events with that of the SST over the western North Pacific that is different from PMM in initializing La Niña events, resulting in the symmetric effect of the VM on ENSO prediction. Thus, it is useful to consider VM variability as a reference for ENSO prediction. Significance Statement We aim to investigate whether the robust relationship between the boreal spring Victoria mode (VM) and the following winter ENSO in observations and climate models has any real predictive use, thus bridging the gap between theory and practical application. Our analyses based mainly on model hindcast datasets examine the forecasting skill of the VM for El Niño and La Niña events. We also reveal the symmetric impact of VM on ENSO through the analysis of observations, which explains the VM’s skill in predicting both El Niño and La Niña events. This study deepens our understanding of the effect of North Pacific SST variability on ENSO.

Improved prediction of extreme ENSO events using an artificial neural network with weighted loss functions

The El Niño–Southern Oscillation (ENSO) causes a wide array of abnormal climates and extreme events, including severe droughts and floods, which have a major impact on humanity. With the development of artificial neural network techniques, various attempts are being made to predict ENSO more accurately. However, there are still limitations in accurately predicting ENSO beyond 6 months, especially for abnormal years with less frequent but greater impact, such as strong El Niño or La Niña, mainly due to insufficient and imbalanced training data. Here, we propose a new weighted loss function to improve ENSO prediction for abnormal years, in which the original (vanilla) loss function is multiplied by the weight function that relatively reduces the weight of high-frequency normal events. The new method applied to recurrent neural networks shows significant improvement in ENSO predictions for all lead times from 1 month to 12 months compared to using the vanilla loss function; in particular, the longer the prediction lead time, the greater the prediction improvement. This method can be applied to a variety of other extreme weather and climate events of low frequency but high impact.