Explainable El Niño predictability from climate mode interactions.

The El Niño-Southern Oscillation (ENSO) provides most of the global seasonal climate forecast skill1-3, yet, quantifying the sources of skilful predictions is a long-standing challenge4-7. Different sources of predictability affect ENSO evolution, leading to distinct global effects. Artificial intelligence forecasts offer promising advancements but linking their skill to specific physical processes is not yet possible8-10, limiting our understanding of the dynamics underpinning the advancements. Here we show that an extended nonlinear recharge oscillator (XRO) model shows skilful ENSO forecasts at lead times up to 16-18 months, better than global climate models and comparable to the most skilful artificial intelligence forecasts. The XRO parsimoniously incorporates the core ENSO dynamics and ENSO's seasonally modulated interactions with other modes of variability in the global oceans. The intrinsic enhancement of ENSO's long-range forecast skill is traceable to the initial conditions of other climate modes by means of their memory and interactions with ENSO and is quantifiable in terms of these modes' contributions to ENSO amplitude. Reforecasts using the XRO trained on climate model output show that reduced biases in both model ENSO dynamics and in climate mode interactions can lead to more skilful ENSO forecasts. The XRO framework's holistic treatment of ENSO's global multi-timescale interactions highlights promising targets for improving ENSO simulations and forecasts.

Author Correction: El Niño-Southern Oscillation complexity.

In this Review, the middle initial of author Kim M. Cobb was omitted. The original Review has been corrected online.

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.

Amplified Arctic warming by phytoplankton under greenhouse warming.

Phytoplankton have attracted increasing attention in climate science due to their impacts on climate systems. A new generation of climate models can now provide estimates of future climate change, considering the biological feedbacks through the development of the coupled physical-ecosystem model. Here we present the geophysical impact of phytoplankton, which is often overlooked in future climate projections. A suite of future warming experiments using a fully coupled ocean-atmosphere model that interacts with a marine ecosystem model reveals that the future phytoplankton change influenced by greenhouse warming can amplify Arctic surface warming considerably. The warming-induced sea ice melting and the corresponding increase in shortwave radiation penetrating into the ocean both result in a longer phytoplankton growing season in the Arctic. In turn, the increase in Arctic phytoplankton warms the ocean surface layer through direct biological heating, triggering additional positive feedbacks in the Arctic, and consequently intensifying the Arctic warming further. Our results establish the presence of marine phytoplankton as an important potential driver of the future Arctic climate changes.

El Niño in a changing climate.

El Niño events, characterized by anomalous warming in the eastern equatorial Pacific Ocean, have global climatic teleconnections and are the most dominant feature of cyclic climate variability on subdecadal timescales. Understanding changes in the frequency or characteristics of El Niño events in a changing climate is therefore of broad scientific and socioeconomic interest. Recent studies show that the canonical El Niño has become less frequent and that a different kind of El Niño has become more common during the late twentieth century, in which warm sea surface temperatures (SSTs) in the central Pacific are flanked on the east and west by cooler SSTs. This type of El Niño, termed the central Pacific El Niño (CP-El Niño; also termed the dateline El Niño, El Niño Modoki or warm pool El Niño), differs from the canonical eastern Pacific El Niño (EP-El Niño) in both the location of maximum SST anomalies and tropical-midlatitude teleconnections. Here we show changes in the ratio of CP-El Niño to EP-El Niño under projected global warming scenarios from the Coupled Model Intercomparison Project phase 3 multi-model data set. Using calculations based on historical El Niño indices, we find that projections of anthropogenic climate change are associated with an increased frequency of the CP-El Niño compared to the EP-El Niño. When restricted to the six climate models with the best representation of the twentieth-century ratio of CP-El Niño to EP-El Niño, the occurrence ratio of CP-El Niño/EP-El Niño is projected to increase as much as five times under global warming. The change is related to a flattening of the thermocline in the equatorial Pacific.

Deep ocean warming-induced El Niño changes.

The deep ocean, a vast thermal reservoir, absorbs excess heat under greenhouse warming, which ultimately regulates the Earth's surface climate. Even if CO2 emissions are successfully reduced, the stored heat will gradually be released, resulting in a particular pattern of ocean warming. Here, we show that deep ocean warming will lead to El Niño-like ocean warming and resultant increased precipitation in the tropical eastern Pacific with southward shift of the intertropical convergence zone. Consequently, the El Niño-Southern Oscillation shifts eastward, intensifying Eastern Pacific El Niño events. In particular, the deep ocean warming could increase convective extreme El Niño events by 40 to 80% relative to the current climate. Our findings suggest that anthropogenic greenhouse warming will have a prolonged impact on El Niño variability through delayed deep ocean warming, even if CO2 stabilization is achieved.

Decreased Indian Ocean Dipole variability under prolonged greenhouse warming.

The Indian Ocean Dipole (IOD) is a major climate variability mode that substantially influences weather extremes and climate patterns worldwide. However, the response of IOD variability to anthropogenic global warming remains highly uncertain. The latest IPCC Sixth Assessment Report concluded that human influences on IOD variability are not robustly detected in observations and twenty-first century climate-model projections. Here, using millennial-length climate simulations, we disentangle forced response and internal variability in IOD change and show that greenhouse warming robustly suppresses IOD variability. On a century time scale, internal variability overwhelms the forced change in IOD, leading to a widespread response in IOD variability. This masking effect is mainly caused by a remote influence of the El Niño-Southern Oscillation. However, on a millennial time scale, nearly all climate models show a long-term weakening trend in IOD variability by greenhouse warming. Our results provide compelling evidence for a human influence on the IOD.

An abrupt shift in gross primary productivity over Eastern China-Mongolia and its inter-model diversity in land surface models.

The terrestrial ecosystem in East Asia mainly consists of semi-arid regions that are sensitive to climate change. Therefore, gross primary productivity (GPP) in East Asia could be highly variable and vulnerable to climate change, which can significantly affect the local carbon budget. Here, we examine the spatial and temporal characteristics of GPP variability in East Asia and its relationship with climate factors over the last three decades. We detect an abrupt decrease in GPP over Eastern China-Mongolia region around the year 2000. This is attributed to an abrupt decrease in precipitation associated with the phase shift of the Pacific decadal oscillation (PDO). We also evaluate the reproducibility of offline land surface models to simulate these abrupt changes. Of the twelve models, eight were able to simulate this abrupt response, while the others failed due to the combination of an exaggerated CO2 fertilization effect and an underrated climate impact. For accurate prediction, it is necessary to improve the sensitivity of the GPP to changes in CO2 concentrations and the climate system.

Low frequency changes in CO2 concentration in East Asia related to Pacific decadal oscillation and Atlantic multi-decadal oscillation for mid-summer and early fall.

The climatological seasonal maximum and minimum CO2 concentrations in East Asia for 1987-2020 have been recorded at April and August, respectively. We found that the CO2 concentration in East Asia during July, August, and September (JAS) is lower than normal before the late 1990s and after the early 2010s (Low_CO2 period), and higher than normal from the late 1990s to the early 2010s (High_CO2 period). The low-frequency variability of CO2 concentration in East Asia during JAS correlates with both Pacific Decadal Oscillation (PDO) and Atlantic Multi-decadal Oscillation (AMO)-related sea surface temperatures (SSTs). We analyzed atmospheric and oceanic conditions during JAS between the two periods, finding that precipitation in East Asia decreased during JAS in High_CO2 period than that in Low_CO2 period, possibly due to PDO and AMO-related SST forcing, which decreases vegetation's photosynthetic activity. This may lead to a higher CO2 concentration than normal in East Asia in High_CO2 period through reduced uptake of CO2 from the atmosphere. This implies that terrestrial vegetation activity influenced by remote SST forcings should be monitored to better understand regional carbon cycles in East Asia.

Understanding elevated CO2 concentrations in East Asia relative to the global mean during boreal spring on the slow and interannual timescales.

It is important to examine the physical processes that regulate current CO2 concentrations in East Asia to understand the global carbon cycle. To do this, we begin by defining the difference between East Asian and global CO2 concentrations (East Asian CO2 concentration minus global CO2 concentration), which is referred to as East Asian local CO2 concentration (i.e., EA_LCO2). Then, we examine the physical processes associated with the variability of EA_LCO2 during boreal spring (March-April-May) on the slow and interannual timescales. Our results indicate that there are two key factors leading to elevated CO2 concentrations in East Asia relative to the global mean during boreal spring; one is higher emissions in East Asia, which mostly explains the increasing in EA_LCO2 on the slow timescales. The other is a cool sea surface temperature (SST) in the eastern tropical Pacific (La-Nina-like SST), which is associated with an interannual higher CO2 concentration in East Asia than the global mean. Enhanced convective activity in the western tropical Pacific, which is associated with a La-Nina-like SST forcing, induces low-pressure circulation in the western North Pacific with northerly winds, leading to suppressed precipitation and cool surface temperature in East Asia. Subsequently, those suppress vegetation growth as well as gross primary product, resulting in relatively high CO2 concentrations in East Asia compared to the global mean.

Increase in convective extreme El Niño events in a CO2 removal scenario.

Convective extreme El Niño (CEE) events, characterized by strong convective events in the eastern Pacific, are known to have a direct link to anomalous climate conditions worldwide, and it has been reported that CEE will occur more frequently under greenhouse warming. Here, using a set of CO2 ramp-up and ramp-down ensemble experiments, we show that frequency and maximum intensity of CEE events increase further in the ramp-down period from the ramp-up period. These changes in CEE are associated with the southward shift of the intertropical convergence zone and intensified nonlinear rainfall response to sea surface temperature change in the ramp-down period. The increasing frequency of CEE has substantial impacts on regional abnormal events and contributed considerably to regional mean climate changes to the CO2 forcings.

Hysteresis of the El Niño-Southern Oscillation to CO2 forcing.

El Niño-Southern Oscillation (ENSO) is the strongest interannual climate variability with far-reaching socioeconomic consequences. Many studies have investigated ENSO-projected changes under future greenhouse warming, but its responses to plausible mitigation behaviors remain unknown. We show that ENSO sea surface temperature (SST) variability and associated global teleconnection patterns exhibit strong hysteretic responses to carbon dioxide (CO2) reduction based on the 28-member ensemble simulations of the CESM1.2 model under an idealized CO2 ramp-up and ramp-down scenario. There is a substantial increase in the ensemble-averaged eastern Pacific SST anomaly variance during the ramp-down period compared to the ramp-up period. Such ENSO hysteresis is mainly attributed to the hysteretic response of the tropical Pacific Intertropical Convergence Zone meridional position to CO2 removal and is further supported by several selected single-member Coupled Model Intercomparison Project Phase 6 (CMIP6) model simulations. The presence of ENSO hysteresis leads to its amplified and prolonged impact in a warming climate, depending on the details of future mitigation pathways.

Hemispherically asymmetric Hadley cell response to CO2 removal.

A poleward shift of the Hadley cell (HC) edge in a warming climate, which contributes to the expansion of drought-prone subtropical regions, has been widely documented. The question addressed here is whether this shift is reversible with CO2 removal. By conducting large-ensemble experiments where CO2 concentrations are systematically increased and then decreased to the present-day level, we show that the poleward-shifted HC edge in a warming climate does not return to its present-day state when CO2 concentrations are reduced. While the Southern Hemisphere HC edge remains poleward of its present-day state, the Northern Hemisphere HC edge ends up farther equatorward of its present-day state. Such hemispherically asymmetric HC edge changes are closely associated with the changes in vertical wind shear in the subtropical atmosphere, which result from the long adjustment time of the ocean response to CO2 removal. Our findings suggest that CO2 removal may not guarantee the recovery of the subtropical dryness associated with the HC changes.

Influence of the recent winter Arctic sea ice loss in short-term simulations of a regional atmospheric model.

Notable changes in the wintertime Arctic atmospheric circulation have occurred over the last few decades. Despite its importance in understanding the recent changes in the Northern Hemisphere midlatitude climate, it remains unclear whether and how these changes are affected by recent Arctic sea ice loss. In this study, a regional scale model is used to separate the direct sea ice influence from the natural variability of large-scale atmospheric circulation. Results show that, in response to sea ice loss, the increase of geopotential height in the mid-to-upper troposphere is robust across the simulations, but the magnitude of the response is highly dependent on the background state of the atmosphere. In most cases the sea ice loss-induced atmospheric warming is trapped near the surface due to the high vertical stability of winter Arctic lower troposphere, accordingly, resulting in a small response of geopotential height. However, when a low-pressure system is located over the Barents Sea, the relatively weak stability allows an upward transport of the surface warming, causing a significantly larger geopotential height increase. This strong state-dependence of atmospheric response which is also found in recent studies using global-scale model experiments, highlights the importance of accurately representing the atmospheric background state for numerical model assessments of sea ice influence.

Tropical origins of the record-breaking 2020 summer rainfall extremes in East Asia.

The East Asian countries have experienced heavy rainfalls in boreal summer 2020. Here, we investigate the dynamical processes driving the rainfall extremes in East Asia during July and August. The Indian Ocean basin warming in June can be responsible for the anticyclonic anomalies in the western North Pacific (WNP), which modulate the zonally-elongated rainfalls in East Asia during July through an atmospheric Rossby wave train. In August, the East Asian rainfall increase is also related to the anticyclonic anomalies in the subtropical WNP, although it is located further north. The north tropical Atlantic warming in June partly contributes to the subtropical WNP rainfall decrease in August through a subtropical teleconnection. Then the subtropical WNP rainfall decrease drives the local anticyclonic anomalies that cause the rainfall increase in East Asia during August. The tropical Indian Ocean anomalously warmed in June and the subtropical WNP rainfall decreased in August 2020, which played a role in modulating the WNP anticyclonic anomalies. Therefore, the record-breaking rainfall extremes in East Asia that occurred during summer 2020 can be explained by the teleconnections associated with the tropical origins among the Indian, Pacific, and Atlantic Oceans and their interbasin interactions.

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.

General circulation and global heat transport in a quadrupling CO2 pulse experiment.

To investigate the response of the general circulation and global transport of heat through both atmosphere and ocean to two-types of carbon dioxide removal scenario, we performed an earth system model experiment in which we imposed a pulse-type quadrupling of CO2 forcing for 50 years and a gradual peak-and-decline of four-time CO2 forcing. We found that the results from two experiments are qualitatively similar to each other. During the forcing-on period, a dominant warming in the upper troposphere over the tropics and on the surface at high latitudes led to a slowdown in the Hadley circulation, but the poleward atmospheric energy transport was enhanced due to an increase in specific humidity. This counteracted the reduction in poleward oceanic energy transport owing to the suppression of the meridional overturning circulation in both Hemispheres. After returning the original CO2 level, the hemispheric thermal contrast was reversed, causing a southward shift of the intertropical convergence zone. To reduce the hemispheric thermal contrast, the northward energy transports in the atmosphere and ocean surface were enhanced while further weakening of the global-scale Atlantic meridional overturning circulation led to southward energy transport in the deep ocean.

Southern Indian Ocean Dipole as a trigger for Central Pacific El Niño since the 2000s.

Despite decades of effort, predicting the El Niño-Southern Oscillation (ENSO) since the 2000s has become increasingly challenging. This is due to the weaker coupling between the ENSO and well-known precursors in tropical ocean basins, particularly in the Indian Ocean. Here we show that the Southern Indian Ocean Dipole (SIOD), which is characterized by an east-west-oriented sea surface temperature dipole pattern over the southern Indian Ocean, has become a key precursor of Central Pacific El Niño since the 2000s with a 14-month lead. The role of the SIOD in the subsequent year's ENSO is distinctive from the equatorial Indian Ocean Dipole mode in that it prolongs the ENSO period. The westward-shifted ENSO has sustained simultaneous SIOD events for longer periods since the 2000s, which leads to weak but persistent westerly anomalies over the western Pacific. This eventually results in the development of the Central Pacific El Niño in the subsequent year.

Asymmetrical response of summer rainfall in East Asia to CO2 forcing.

Understanding the regional hydrological response to varying CO2 concentration is critical for cost-benefit analysis of mitigation and adaptation polices in the near future. To characterize summer monsoon rainfall change in East Asia in a changing CO2 pathway, we used the Community Earth System Model (CESM) with 28 ensemble members in which the CO2 concentration increases at a rate of 1% per year until its quadrupling peak, i.e., 1468 ppm (ramp-up period), followed by a decrease of 1% per year until the present-day climate conditions, i.e., 367 ppm (ramp-down period). Although the CO2 concentration change is symmetric in time, the amount of summer rainfall anomaly in East Asia is increased 42% during a ramp-down period than that during a ramp-up period when the two periods of the same CO2 concentration are compared. This asymmetrical rainfall response is mainly due to an enhanced El Niño-like warming pattern as well as its associated increase in the sea surface temperature in the western North Pacific during a ramp-down period. These sea surface temperature patterns enhance the atmospheric teleconnections and the local meridional circulations around East Asia, resulting in more rainfall over East Asia during a ramp-down period. This result implies that the removal of CO2 does not guarantee the return of regional rainfall to the previous climate state with the same CO2 concentration.

Zonally asymmetric phytoplankton response to the Southern annular mode in the marginal sea of the Southern ocean.

Antarctic marine biological variability modulates climate systems via the biological pump. However, the knowledge of biological response in the Southern Ocean to climate variability still has been lack of understanding owing to limited ocean color data in the high latitude region. We investigated the surface chlorophyll concentration responses to the Southern annular mode (SAM) in the marginal sea of the Southern ocean using satellite observation and reanalysis data focusing on the austral summer. The positive phase of SAM is associated with enhanced and poleward-shifted westerly winds, leading to physical and biogeochemical responses over the Southern ocean. Our result indicates that chlorophyll has strong zonally asymmetric responses to SAM owing to different limiting factors of phytoplankton growth per region. For the positive SAM phase, chlorophyll tends to increase in the western Amundsen-Ross Sea but decreases in the D'Urville Sea. It is suggested that the distinct limiting factors are associated with the seasonal variability of sea ice and upwelling per region.