Probabilistic projections of El Niño Southern Oscillation properties accounting for model dependence and skill

The El Niño - Southern Oscillation (ENSO) is a dominant mode of global climate variability. Nevertheless, future multi-model probabilistic projections of ENSO properties have not yet been made. Main roadblocks that have been hindering making these projections are climate model dependence and difficulty in quantifying historical model performance. Dependence is broadly defined as similarity between climate model output, assumptions, or physical parameterizations. Here, we propose a unifying metric of relative model performance, based on the probability density function (PDF) of ENSO paths. This metric is applied to assess the overall skill of Climate Model Intercomparison Project phase 6 (CMIP6) climate models at capturing ENSO. We then perform future multi-model probabilistic projections of changes in ENSO properties (from years 1850-1949 to 2040-2099) under the shared socioeconomic pathway scenario SSP585, accounting for model skill and dependence. We find that future ENSO will likely be more seasonally locked (89% chance), and have a longer period (67% chance). Yet, the jury is still out on future ENSO amplification. Our method reduces uncertainty by up to 37% compared to a simple approach ignoring model dependence and skill.

Coral Sr/Ca records provide realistic representation of eastern Indian Ocean cooling during extreme positive Indian Ocean Dipole events

Extreme positive Indian Ocean Dipole (pIOD) events are amplified by non-linear ocean–atmosphere interactions and are characterized by pronounced cooling in the eastern equatorial Indian Ocean. These non-linear feedbacks are not adequately represented in historical products of sea surface temperatures that underestimate the magnitude of extreme pIOD events. Here, we present a sea surface temperature (SST) reconstruction based on monthly coral Sr/Ca ratios measured in two coral cores from Enggano Island (Indonesia), that lies in the eastern pole of the IOD. The coral SST reconstruction extends from 1930 to 2008 and captures the magnitude of cooling during extreme pIOD events as shown in recent satellite and reanalysis data of SST that include ocean dynamics. The corals indicate that the 1961 pIOD event was at least as severe as the 1997 event, while the 1963 pIOD was more comparable to the 2006 event. The magnitude 1967 pIOD is difficult to assess at present due to poor replication between coral cores, and may be comparable to either 1997 or 2006. Cooling during the 1972 pIOD was short-lived and followed by pronounced warming, as seen in the moderate pIOD event of 1982. A combination of coral SST reconstructions and an extension of new reanalysis products of SST to historical time scales could help to better assess the severity and impact of past pIOD events such as the ones seen in the 1960s.

Spatiotemporal neural network with attention mechanism for El Niño forecasts

To learn spatiotemporal representations and anomaly predictions from geophysical data, we propose STANet, a spatiotemporal neural network with a trainable attention mechanism, and apply it to El Niño predictions for long-lead forecasts. The STANet makes two critical architectural improvements: it learns spatial features globally by expanding the network's receptive field and encodes long-term sequential features with visual attention using a stateful long-short term memory network. The STANet conducts multitask learning of Nino3.4 index prediction and calendar month classification for predicted indices. In a comparison of the proposed STANet performance with the state-of-the-art model, the accuracy of the 12-month forecast lead correlation coefficient was improved by 5.8% and 13% for Nino3.4 index prediction and corresponding temporal classification, respectively. Furthermore, the spatially attentive regions for the strong El Niño events displayed spatial relationships consistent with the revealed precursor for El Niño occurrence, indicating that the proposed STANet provides good understanding of the spatiotemporal behavior of global sea surface temperature and oceanic heat content for El Niño evolution.

A flexible data-driven cyclostationary model for the probability density of El Niño-Southern Oscillation

Model simulations of El Niño-Southern Oscillation (ENSO) are usually evaluated by comparing them to observations using a multitude of metrics. However, this approach cannot provide an objective summary metric of model performance. Here, we propose that such an objective model evaluation should involve comparing the full joint probability density functions (pdf's) of ENSO. For simplicity, ENSO state is defined here as sea surface temperature anomalies over the Niño 3 region and equatorial Pacific thermocline depth anomalies. We argue that all ENSO metrics are a function of the joint pdf, the latter fully specifying the underlying stochastic process. Unfortunately, there is a lack of methods to recover the joint ENSO pdf from climate models or observations. Here, we develop a data-driven stochastic model for ENSO that allows for an analytic solution of the non-Markov non-Gaussian cyclostationary ENSO pdf. We show that the model can explain relevant ENSO features found in the observations and can serve as an ENSO simulator. We demonstrate that the model can reasonably approximate ENSO in most GCMs and is useful at exploring the internal ENSO variability. The general approach is not limited to ENSO and could be applied to other cyclostationary processes.

Unified deep learning model for El Niño/Southern Oscillation forecasts by incorporating seasonality in climate data

Although deep learning has achieved a milestone in forecasting the El Niño-Southern Oscillation (ENSO), the current models are insufficient to simulate diverse characteristics of the ENSO, which depends on the calendar season. Consequently, a model was generated for specific seasons which indicates these models did not consider physical constraints between different target seasons and forecast lead times, thereby leading to arbitrary fluctuations in the predicted time series. To overcome this problem and account for ENSO seasonality, we developed an all-season convolutional neural network (A_CNN) model. The correlation skill of the ENSO index was particularly improved for forecasts of the boreal spring, which is the most challenging season to predict. Moreover, activation map values indicated a clear time evolution with increasing forecast lead time. The study findings reveal the comprehensive role of various climate precursors of ENSO events that act differently over time, thus indicating the potential of the A_CNN model as a diagnostic tool.

A novel approach for discovering stochastic models behind data applied to El Niño-Southern Oscillation

Stochastic differential equations (SDEs) are ubiquitous across disciplines, and uncovering SDEs driving observed time series data is a key scientific challenge. Most previous work on this topic has relied on restrictive assumptions, undermining the generality of these approaches. We present a novel technique to uncover driving probabilistic models that is based on kernel density estimation. The approach relies on few assumptions, does not restrict underlying functional forms, and can be used even on non-Markov systems. When applied to El Niño–Southern Oscillation (ENSO), the fitted empirical model simulations can almost perfectly capture key time series properties of ENSO. This confirms that ENSO could be represented as a two-variable stochastic dynamical system. Our experiments provide insights into ENSO dynamics and suggest that state-dependent noise does not play a major role in ENSO skewness. Our method is general and can be used across disciplines for inverse and forward modeling, to shed light on structure of system dynamics and noise, to evaluate system predictability, and to generate synthetic datasets with realistic properties.

Fokker-Planck dynamics of the El Niño-Southern Oscillation

The asymmetric nature of the El Niño-Southern Oscillation (ENSO) is explored by using a probabilistic model (PROM) for ENSO. Based on a Fokker–Planck Equation (FPE), PROM describes the dynamics of a nonlinear stochastic ENSO recharge oscillator model for eastern equatorial Pacific temperature anomalies and equatorial Pacific basin-averaged thermocline depth changes. Eigen analyses of PROM provide new insights into the stationary and oscillatory solutions of the stochastic dynamical system. The first probabilistic eigenmode represents a stationary mode, which exhibits the asymmetric features of ENSO, in case deterministic nonlinearities or multiplicative noises are included. The second mode is linked to the oscillatory nature of ENSO and represents a cyclic asymmetric probability distribution, which emerges from the key dynamical processes. Other eigenmodes are associated with the temporal evolution of higher order statistical moments of the ENSO system. The model solutions demonstrate that the deterministic nonlinearity plays a stronger role in establishing the observed asymmetry of ENSO as compared to the multiplicative stochastic part.

Influence of El Niño events on sea surface salinity over the central equatorial Indian Ocean

The El Niño event is a major large-scale air-sea interaction phenomenon over the tropical Pacific region. Previous studies classified El Niño into three types - canonical El Niño, El Niño Modoki I, and El Niño Modoki II. This research reveals that different types of El Niño present dramatic effects on the interannual variability of sea surface salinity over the central equatorial Indian Ocean in the boreal autumn. The decreasing (increasing) sea surface salinity during the canonical El Niño and the EI Niño Modoki I (the EI Niño Modoki II) events is identified. The salinity budget analysis is performed to identify the dominant factors responsible for the variability of sea surface salinity over the central Indian Ocean. The results indicate that the wind-driven anomalous zonal advection plays an important role in sea surface salinity variability during the El Niño events associated with the forcing from the anomalous Walker circulation over the equatorial Indian Ocean.

Deep learning for multi-year ENSO forecasts

Variations in the El Niño/Southern Oscillation (ENSO) are associated with a wide array of regional climate extremes and ecosystem impacts¹. Robust, long-lead forecasts would therefore be valuable for managing policy responses. But despite decades of effort, forecasting ENSO events at lead times of more than one year remains problematic². Here we show that a statistical forecast model employing a deep-learning approach produces skilful ENSO forecasts for lead times of up to one and a half years. To circumvent the limited amount of observation data, we use transfer learning to train a convolutional neural network (CNN) first on historical simulations³ and subsequently on reanalysis from 1871 to 1973. During the validation period from 1984 to 2017, the all-season correlation skill of the Nino3.4 index of the CNN model is much higher than those of current state-of-the-art dynamical forecast systems. The CNN model is also better at predicting the detailed zonal distribution of sea surface temperatures, overcoming a weakness of dynamical forecast models. A heat map analysis indicates that the CNN model predicts ENSO events using physically reasonable precursors. The CNN model is thus a powerful tool for both the prediction of ENSO events and for the analysis of their associated complex mechanisms.

The changing relationship between ENSO and its extratropical response patterns

El Niño-Southern Oscillation (ENSO) can influence Northern Hemisphere seasonal conditions through its interaction with the Pacific-North American pattern (PNA) and the Tropical Northern Hemisphere pattern (TNH). Possibly due to Earth’s changing climate, the variability of ENSO, as well as the zonal location of its sea-surface temperature (SST) anomaly, is changing. Along with this, the strength and location of the jet, in which these atmospheric patterns are embedded, are changing. Using a simple tracking algorithm we create a continuous time series for the zonal location of ENSO’s SST anomaly, and show that the relationship between ENSO and the PNA is linearly sensitive to this location, while its relationship with the TNH is not. ENSO’s relationship with both the TNH and PNA is shown to be strongly influenced by the position of the Pacific jet stream. The ENSO-TNH relationship is found to be linked to phase changes of the Atlantic Multidecadal Oscillation, and the resulting changes in SST and jet speed.

El Niño-Southern Oscillation complexity

El Niño events are characterized by surface warming of the tropical Pacific Ocean and weakening of equatorial trade winds that occur every few years. Such conditions are accompanied by changes in atmospheric and oceanic circulation, affecting global climate, marine and terrestrial ecosystems, fisheries and human activities. The alternation of warm El Niño and cold La Niña conditions, referred to as the El Niño-Southern Oscillation (ENSO), represents the strongest year-to-year fluctuation of the global climate system. Here we provide a synopsis of our current understanding of the spatio-temporal complexity of this important climate mode and its influence on the Earth system.

The pre-Argo ocean reanalyses may be seriously affected by the spatial coverage of moored buoys

Assimilation methods, meant to constrain divergence of model trajectory from reality using observations, do not exactly satisfy the physical laws governing the model state variables. This allows mismatches in the analysis in the vicinity of observation locations where the effect of assimilation is most prominent. These mismatches are usually mitigated either by the model dynamics in between the analysis cycles and/or by assimilation at the next analysis cycle. However, if the observations coverage is limited in space, as it was in the ocean before the Argo era, these mechanisms may be insufficient to dampen the mismatches, which we call shocks, and they may remain and grow. Here we show through controlled experiments, using real and simulated observations in two different ocean models and assimilation systems, that such shocks are generated in the ocean at the lateral boundaries of the moored buoy network. They thrive and propagate westward as Rossby waves along these boundaries. However, these shocks are essentially eliminated by the assimilation of near-homogenous global Argo distribution. These findings question the fidelity of ocean reanalysis products in the pre-Argo era. For example, a reanalysis that ignores Argo floats and assimilates only moored buoys, wrongly represents 2008 as a negative Indian Ocean Dipole year.

Robust contribution of decadal anomalies to the frequency of central-Pacific El Niño

During year-to-year El Niño events in recent decades, major sea surface warming has occurred frequently in the central Pacific. This is distinct from the eastern Pacific warming pattern during canonical El Niño events. Accordingly, the central-Pacific El Niño exerts distinct impacts on ecosystems, climate and hurricanes worldwide. The increased frequency of the new type of El Niño presents a challenge not only for the understanding of El Niño dynamics and its change but also for the prediction of El Niño and its global impacts at present and future climate. Previous studies have proposed different indices to represent the two types of El Niño for better understanding, prediction and impact assessment. Here, we find that all popularly used indices for the central-Pacific El Niño show a dominant spectral peak at a decadal period with comparatively weak variance at interannual timescales. Our results suggest that decadal anomalies have an important contribution to the occurrence of the central-Pacific El Niño over past decades. Removing the decadal component leads to a significant reduction in the frequency of the central-Pacific El Niño in observations and in Coupled Model Intercomparison Project Phase 5 simulations of preindustrial, historical and future climate.

Climate mediates hypoxic stress on fish diversity and nursery function at the land-sea interface

Coastal ecosystems provide numerous important ecological services, including maintenance of biodiversity and nursery grounds for many fish species of ecological and economic importance. However, human population growth has led to increased pollution, ocean warming, hypoxia, and habitat alteration that threaten ecosystem services. In this study, we used long-term datasets of fish abundance, water quality, and climatic factors to assess the threat of hypoxia and the regulating effects of climate on fish diversity and nursery conditions in Elkhorn Slough, a highly eutrophic estuary in central California (United States), which also serves as a biodiversity hot spot and critical nursery grounds for offshore fisheries in a broader region. We found that hypoxic conditions had strong negative effects on extent of suitable fish habitat, fish species richness, and abundance of the two most common flatfish species, English sole (Parophrys vetulus) and speckled sanddab (Citharichthys stigmaeus). The estuary serves as an important nursery ground for English sole, making this species vulnerable to anthropogenic threats. We determined that estuarine hypoxia was associated with significant declines in English sole nursery habitat, with cascading effects on recruitment to the offshore adult population and fishery, indicating that human land use activities can indirectly affect offshore fisheries. Estuarine hypoxic conditions varied spatially and temporally and were alleviated by strengthening of El Niño conditions through indirect pathways, a consistent result in most estuaries across the northeast Pacific. These results demonstrate that changes to coastal land use and climate can fundamentally alter the diversity and functioning of coastal nurseries and their adjacent ocean ecosystems.

Mixed layer depth trends in the Bay of Biscay over the period 1975-2010

Wintertime trends in mixed layer depth (MLD) were calculated in the Bay of Biscay over the period 1975-2010 using the Simple Ocean Data Assimilation (SODA) package. The reliability of the SODA database was confirmed correlating its results with those obtained from the experimental Argo database over the period 2003-2010. An iso-thermal layer depth (TLD) and an iso-pycnal layer depth (PLD) were defined using the threshold difference method with ΔT = 0.5°C and Δσθ = 0.125 kg/m3. Wintertime trends of the MLD were calculated using winter extended (December-March) anomalies and annual maxima. Trends calculated for the whole Bay of Biscay using both parameters (TLD and PLD) showed to be dependent on the area. Thus, MLD became deeper in the southeastern corner and shallower in the rest of the area. Air temperature was shown to play a key role in regulating the different spatial behavior of the MLD. Negative air temperature trends localized in the southeastern corner coincide with MLD deepening in this area, while, positive air temperature trends are associated to MLD shoaling in the rest of the bay. Additionally, the temperature trend calculated along the first 700 m of the water column is in good agreement with the different spatial behavior revealed for the MLD trend.

Variability of coastal and ocean water temperature in the upper 700 m along the Western Iberian Peninsula from 1975 to 2006

Temperature is observed to have different trends at coastal and ocean locations along the western Iberian Peninsula from 1975 to 2006, which corresponds to the last warming period in the area under study. The analysis was carried out by means of the Simple Ocean Data Assimilation (SODA). Reanalysis data are available at monthly scale with a horizontal resolution of 0.5°×0.5° and a vertical resolution of 40 levels, which allows obtaining information beneath the sea surface. Only the first 21 vertical levels (from 5.0 m to 729.35 m) were considered here, since the most important changes in heat content observed for the world ocean during the last decades, correspond to the upper 700 m. Warming was observed to be considerably higher at ocean locations than at coastal ones. Ocean warming ranged from values on the order of 0.3°C dec⁻¹ near surface to less than 0.1°C dec⁻¹ at 500 m, while coastal warming showed values close to 0.2°C dec⁻¹ near surface, decreasing rapidly below 0.1°C dec⁻¹ for depths on the order of 50 m. The heat content anomaly for the upper 700 m, showed a sharp increase from coast (0.46 Wm⁻²) to ocean (1.59 Wm⁻²). The difference between coastal and ocean values was related to the presence of coastal upwelling, which partially inhibits the warming from surface of near shore water.