El Niño

El Niño is a recurring climate phenomenon characterized by unusually warm surface waters in the central and eastern equatorial Pacific Ocean, accompanied by changes in winds, atmospheric pressure, and tropical rainfall. It is the warm phase of the El Niño–Southern Oscillation (ENSO), a coupled system linking the ocean and the atmosphere. Its cool counterpart is La Niña, while neutral conditions occur between episodes. El Niño involves interactions across the tropical Pacific rather than simply a warm current along the South American coast. (noaa.gov)

Through atmospheric teleconnections, El Niño influences climate far beyond the Pacific, altering the likelihood of drought, heavy rainfall, and other seasonal anomalies. Changes in ocean conditions also affect marine productivity, fisheries, and food webs. These effects vary between events and regions: El Niño shifts probabilities rather than determining every local weather outcome. Its importance has made it a major subject of ocean–atmosphere research, sustained observation, and seasonal forecasting. (pmel.noaa.gov)

Definition, terminology, and relationship to ENSO

El Niño is the coupled ocean–atmosphere warm phase of the El Niño–Southern Oscillation (ENSO), characterized by above-average sea-surface temperatures in the central and eastern equatorial Pacific together with associated changes in tropical winds, rainfall, and atmospheric pressure. It is therefore more than a patch of warm water: changes in the ocean and atmosphere interact. During El Niño, the normally westward-blowing equatorial trade winds weaken, and the region of enhanced tropical rainfall shifts eastward into the central or eastern Pacific. Ocean temperatures influence atmospheric circulation, while the altered winds affect ocean temperatures and currents, helping sustain the coupled state. (cpc.ncep.noaa.gov)

“Warm” refers to a temperature anomaly, not to a fixed absolute water temperature. An anomaly is the difference between an observed temperature and a reference average for the same location and time of year. A positive anomaly consequently means warmer than the local seasonal expectation; it does not mean that the water is warmer than every other part of the Pacific. Nor does ordinary seasonal warming by itself establish El Niño. Monitoring uses temperature departures averaged over specified regions and periods, alongside evidence that the atmosphere is responding consistently with the oceanic warming. A temperature index describes the oceanic component, but the presence or absence of atmospheric coupling is essential to distinguishing ENSO conditions. (aoml.noaa.gov)

The Southern Oscillation is the atmospheric component of ENSO: a large-scale fluctuation in surface air pressure between the western and eastern tropical Pacific. During El Niño, pressure is generally higher than average over Indonesia and the western tropical Pacific and lower than average farther east; during La Niña, these departures are broadly reversed. The commonly used Southern Oscillation Index measures departures from the usual pressure difference between Tahiti and Darwin, Australia, and generally takes negative values during El Niño and positive values during La Niña. The combined name El Niño–Southern Oscillation identifies the linked oceanic and atmospheric variations, rather than two independent events that merely occur together. (cpc.ncep.noaa.gov)

La Niña is ENSO’s cold phase, with below-average sea-surface temperatures in the central and eastern equatorial Pacific, stronger equatorial easterly winds, and a contrasting tropical rainfall pattern. ENSO-neutral denotes conditions in which neither El Niño nor La Niña is established. Ocean temperatures and atmospheric circulation are often near their seasonal averages, but neutral does not require every measured variable to equal its average. It can also describe circumstances in which ocean temperatures resemble a warm or cold phase without the corresponding atmospheric response, or vice versa. El Niño, La Niña, and neutral thus describe states within one coupled climate system; “ENSO” is not simply another name for El Niño. (content-drupal.climate.gov)

The Spanish name El Niño means “the boy” and, in this usage, refers to the Christ child. Its origin is associated with fishermen in northern Peru who recognized warm coastal waters around Christmas. Accounts recorded by Peruvian geographers in the late nineteenth century described a warm, southward-flowing “El Niño current”; the term initially concerned seasonal coastal warming, rather than the modern basin-scale phenomenon. During the early twentieth century, scientific usage increasingly applied the name to unusually pronounced warming episodes and their accompanying climatic disturbances. By about 1960, researchers recognized that warming off Peru could form part of a much larger equatorial Pacific disturbance. The subsequent association with the Southern Oscillation, developed through work including Jacob Bjerknes’s coupled explanation in the late 1960s, gave the term its broader climatic meaning. The Christmas association explains the name, not a requirement that every event begin at Christmas. (aoml.noaa.gov)

The tropical Pacific background state

The western equatorial Pacific contains a broad warm pool, with mean sea-surface temperatures above 28°C near Indonesia and the surrounding seas. Farther east, a relatively cool “cold tongue” extends westward along the equator from the vicinity of South America, sometimes reaching close to the international date line. This temperature contrast persists despite strong tropical solar heating across the basin. Prevailing easterly trade winds—winds blowing from east to west—drive surface water westward and help maintain the contrast. The trades are generally more persistent in the central and eastern equatorial Pacific than over the western warm pool, where mean winds are weaker and deep atmospheric convection is frequent. (pmel.noaa.gov)

Wind stress drives the westward South Equatorial Current and accumulates warm upper-ocean water in the western basin. Consequently, the thermocline, the layer in which temperature decreases rapidly with depth, lies deeper in the west and slopes upward toward the east. Oceanographers commonly use the depth of the 20°C isotherm as an indicator of this structure. A mean equatorial section assembled from observations during the Tropical Ocean–Global Atmosphere programme showed water warmer than 28°C extending through roughly the upper 100 metres in the western Pacific. The associated sea-surface slope establishes an eastward oceanic pressure-gradient force. Beneath the predominantly westward surface flow, the Equatorial Undercurrent carries water eastward within the thermocline, supplying part of the water subsequently brought toward the surface in the eastern basin. (pmel.noaa.gov)

Equatorial upwelling results from divergence of wind-driven surface transport. Under easterly winds, Earth’s rotation produces transport away from the equator: northward on its northern side and southward on its southern side. Water rises from below to replace the departing surface water. Because the eastern thermocline is relatively shallow, this upward movement and associated turbulent mixing readily introduce cooler subsurface water into the surface layer. The cold tongue therefore reflects a balance among upwelling, mixing, horizontal transport, and surface heating, rather than simply a current carrying cold coastal water westward. William S. Kessler’s 2006 review reported estimated equatorial upwelling speeds of a few metres per day, while emphasizing uncertainty in their spatial distribution. (pmel.noaa.gov)

Coastal upwelling operates through a related but geographically distinct mechanism. Along the coast of Peru, winds directed approximately alongshore toward the equator drive surface transport offshore; colder subsurface water rises to replace it. Here the coastline provides the boundary from which surface water diverges, whereas equatorial upwelling involves divergence between the two hemispheres without requiring a coastal boundary. Although coastal and equatorial cool waters form a visually connected band, their cooling is substantially maintained by local wind-driven processes. (oceanservice.noaa.gov)

The atmospheric counterpart is the Walker circulation, an east–west overturning circulation. Warm western waters support moist rising air, deep convective clouds, and substantial precipitation, with relatively low surface pressure. Air flows eastward aloft and descends over the cooler east-central Pacific, where surface pressure is relatively high and deep convection is less prevalent. Near-surface easterlies complete the circulation. NOAA’s description of the mean December–February pattern places the strongest ascent over Indonesia and strongest descent over the east-central equatorial Pacific. This is a large-scale mean structure, not an unchanging cell: its position and intensity vary seasonally and with ENSO. (cpc.ncep.noaa.gov)

Ocean–atmosphere mechanisms

El Niño growth is sustained by the Bjerknes positive feedback, described by meteorologist Jacob Bjerknes in 1966. An initial weakening of the easterly trade winds changes equatorial currents and reduces the supply of cold water to the surface, allowing the central and eastern Pacific to warm. The resulting reduction in the east–west sea-surface temperature contrast shifts deep atmospheric convection eastward and weakens the atmospheric pressure gradient supporting the trades. Further wind weakening then amplifies the oceanic warming. This coupled loop connects changes in the ocean to changes in the atmosphere; neither wind anomalies nor surface warming alone describe the complete mechanism. (pmel.noaa.gov)

Two physically distinct processes reduce cooling by upwelling. First, weaker easterly wind stress reduces the divergence of near-surface currents away from the equator, decreasing the upwelling velocity: less cold water enters the surface layer. Second, a deeper thermocline—the layer where temperature decreases rapidly with depth—changes the temperature of the water supplied from below. Even where upward motion continues, the water reaching the surface can be warmer because the cold reservoir lies farther beneath it. This thermocline feedback is particularly important in the eastern equatorial Pacific, where the thermocline is normally shallow. A reduction in upwelling speed and a warming of its source waters can act together, but they are not interchangeable explanations. (pmel.noaa.gov)

Horizontal transport also contributes through zonal advective feedback. When the trades relax, the normally westward surface current weakens and may reverse, allowing warm western-Pacific water to move eastward. Eastward current anomalies transport warmer water into the central Pacific, while weaker westward flow reduces the delivery of cooler eastern water. The expanding warm region alters convection and winds, reinforcing the current changes. In model experiments examining the March 1997 westerly wind event, Matthieu Lengaigne and colleagues found an eastward warm-pool displacement of approximately 2,000 km within one month. Their 2003 study also showed how the resulting temperature pattern could weaken the trades and favor additional westerly wind events; these quantitative results describe model experiments rather than a universal displacement rate. (pmel.noaa.gov)

Equatorial waves communicate wind forcing across the basin. Downwelling Kelvin waves propagate eastward along the equator, raising sea level, deepening the thermocline, and producing eastward current anomalies. Observations analyzed by William Kessler and colleagues identified propagation speeds near $2.4\ \mathrm{m\,s^{-1}}$ and central-Pacific thermocline displacements of roughly 20 m in either direction. Rossby waves propagate westward more slowly and redistribute thermocline depth and upper-ocean heat; wind relaxation can generate upwelling Rossby signals alongside eastward downwelling Kelvin signals. Boundary reflection can convert part of one wave type into another, while Kelvin-wave energy reaching the Americas can also travel poleward along the coast. These are propagating adjustments of currents and subsurface structure, not simply parcels of warm water crossing the ocean at the wave speed. (pmel.noaa.gov)

Westerly wind bursts are short-lived episodes of eastward winds or strong westerly wind anomalies in the western and central equatorial Pacific. They can accelerate eastward currents and launch downwelling Kelvin waves, contributing to event initiation or growth. Some accompany intraseasonal atmospheric disturbances, including the Madden–Julian Oscillation. Their effects depend on the ocean’s existing thermal structure and the location and timing of the forcing; the developing warm-pool pattern can itself influence subsequent bursts. They therefore represent possible contributors rather than a universal trigger: ENSO research distinguishes amplification by transient atmospheric variability from the slower coupled feedbacks that sustain an event. (pmel.noaa.gov)

Development, seasonality, and termination

The El Niño–Southern Oscillation (ENSO) varies irregularly on timescales commonly described as approximately two to seven years; this range is not a timetable for successive El Niño events. Individual warm episodes generally last about 9–12 months, although some persist longer. Development commonly begins during March–June, strengthening continues through boreal summer and autumn, and maximum intensity usually occurs during November–February. Weakening often follows during March–June of the next year. This tendency to mature near the end of the calendar year is called seasonal phase locking: it describes a statistical preference, not a requirement that every event begin or end in the same month. (climate.gov)

A major source of the system’s persistence is the slowly changing reservoir of warm water beneath the tropical Pacific surface. In the recharge oscillator framework developed by Fei-Fei Jin in 1997, wind-driven circulation builds up equatorial upper-ocean heat content before a warm event, then redistributes warm water away from equatorial latitudes during its development. A commonly used measure is warm-water volume above the $20^\circ\mathrm{C}$ isotherm within $5^\circ\mathrm{N}$–$5^\circ\mathrm{S}$, from $120^\circ\mathrm{E}$ to $80^\circ\mathrm{W}$. This measures a subsurface reservoir rather than surface temperature alone. Christopher Meinen and Michael McPhaden’s 2000 analysis of observations from 1980–1999 found that the recharge mode preceded changes in eastern-Pacific surface temperature by about seven months on average. That observed lead was specific to the analysed record, not a fixed countdown to an event. (pmel.noaa.gov)

During El Niño, surface warming and depletion of this subsurface reservoir can occur simultaneously. The east–west tilt of the thermocline—the layer where temperature decreases rapidly with depth—changes relatively quickly with the winds, while its basin-wide average depth adjusts more slowly as warm water is discharged toward higher latitudes. Consequently, a deep thermocline in the eastern Pacific can initially sustain warm surface conditions even as the equatorial reservoir is shrinking. Eventually, the reduced reservoir favours a shallower thermocline and colder water available to upwelling, counteracting further warming. Recharge alone does not guarantee onset: substantial equatorial heat-content increases in early 2000 and 2001 did not immediately produce El Niño events. (pmel.noaa.gov)

Delayed oceanic adjustments provide additional negative feedbacks. In the delayed-oscillator mechanism, westward-propagating Rossby waves can reflect at the western Pacific boundary into eastward-propagating Kelvin waves that raise the thermocline and favour cooling. Typical basin-crossing times are approximately six months for the relevant Rossby waves and two months for Kelvin waves, so the ocean responds to earlier wind changes over several seasons. Cooling Kelvin waves can also be generated directly by renewed easterly wind anomalies; boundary reflection is therefore not the sole termination mechanism. Surface heat exchange usually damps warm anomalies as well, through changes in evaporation and solar heating, although cloud-related feedbacks vary with the atmospheric regime. (pmel.noaa.gov)

Termination can lead either to ENSO-neutral conditions or to La Niña, depending on the remaining subsurface heat deficit and subsequent atmospheric forcing. The contrasting endings of two events illustrate this contingency. The 1997–1998 El Niño was followed by rapid cooling and a prolonged cold phase extending into 2001. After the 2002–2003 event, however, an emerging cooling trend was interrupted by westerly wind bursts in May and July 2003. These generated warming oceanic adjustments, including a downwelling Kelvin wave, and near-neutral conditions prevailed by September instead of the anticipated La Niña. An event’s decay therefore does not establish an inevitable alternating sequence of warm and cold phases. (pmel.noaa.gov)

Indices and event classification

Sea-surface temperature (SST) indices describe departures from a climatological average within defined Pacific monitoring regions. Niño 1+2 covers 0°–10°S, 90°–80°W, near Peru and Ecuador; Niño 3 covers 5°N–5°S, 150°–90°W in the eastern equatorial Pacific; Niño 3.4 covers 5°N–5°S, 170°W–120°W in the east-central Pacific; and Niño 4 covers 5°N–5°S, 160°E–150°W, spanning the International Date Line farther west. Their different geographical coverage allows monitoring of warming across the basin rather than treating every warm anomaly as equivalent. An index expressed as $+1^\circ\mathrm{C}$ denotes a departure from its specified reference average, not the absolute water temperature. (cpc.ncep.noaa.gov)

The traditional Oceanic Niño Index (ONI) is a three-month running mean of Niño 3.4 SST anomalies. NOAA’s historical ONI series based on Extended Reconstructed Sea Surface Temperature version 5 (ERSSTv5) uses centered 30-year climatological base periods, updated every five years. Its historical El Niño criterion is an ONI of at least $+0.5^\circ\mathrm{C}$ for five consecutive overlapping three-month seasons; the corresponding La Niña threshold is $-0.5^\circ\mathrm{C}$ or lower. For example, December–February, January–March, February–April, March–May, and April–June constitute five overlapping seasons, not five independent three-month intervals. This persistence requirement distinguishes sustained warming from shorter fluctuations, but the oceanic index alone does not establish that atmospheric conditions are consistent with coupled ENSO. (cpc.ncep.noaa.gov)

In February 2026, NOAA’s Climate Prediction Center adopted the Relative Oceanic Niño Index (RONI) for official ENSO monitoring and prediction. RONI subtracts the mean SST anomaly across the global tropical oceans, 20°N–20°S, from the Niño 3.4 anomaly, then rescales the difference to preserve the variance of the original Niño 3.4 index. Its operational ERSSTv6 series uses the 1991–2020 reference period and three-month averaging, retaining the $\pm0.5^\circ\mathrm{C}$ thresholds and five-overlapping-season historical criterion. The subtraction reduces sensitivity to warming shared across the tropics: Niño 3.4 can be warmer than its historical average without being comparably warm relative to other tropical oceans. RONI therefore measures a temperature contrast relevant to atmospheric response rather than simply the local departure from an older climatology. Historical classifications are also reevaluated with RONI, while ONI remains available as a separate record. (cpc.ncep.noaa.gov)

Atmospheric confirmation includes weakened equatorial easterly trade winds and an eastward shift of tropical rainfall. The Southern Oscillation Index (SOI) provides a complementary pressure-based measure: it describes the standardized anomaly of the sea-level pressure difference between Tahiti and Darwin, Australia. El Niño is generally associated with negative SOI values. The Australian Bureau of Meteorology’s Troup SOI multiplies the standardized anomaly by ten and uses a 1933–1992 reference period; sustained values below −7 commonly accompany El Niño. That numerical threshold belongs to this particular SOI convention and is not interchangeable with thresholds from differently normalized pressure indices. (noaa.gov)

Agency definitions also differ in monitoring region, averaging interval, and persistence. The Japan Meteorological Agency defines El Niño when the five-month running mean Niño 3 SST departure reaches at least $+0.5^\circ\mathrm{C}$ for six consecutive months or longer. The Australian Bureau uses sustained monthly relative Niño 3 or Niño 3.4 values above $+0.8^\circ\mathrm{C}$ as typical of El Niño, assessed alongside winds, cloudiness, subsurface temperatures, and SOI. These differences can produce different event dates or classifications without implying contradictory observations. Comparisons of event strength must consequently identify the index, SST dataset, reference climatology, and averaging interval. Changing from ONI to RONI can alter an episode’s numerical magnitude or historical label without changing the weather and ocean conditions that occurred during it. (data.jma.go.jp)

Spatial diversity and coastal warming

Eastern-Pacific (EP) El Niño events have their largest positive sea-surface temperature anomalies in the eastern equatorial Pacific, with warming extending toward the South American coast. Central-Pacific (CP) events instead concentrate their strongest anomalies farther west, near the international date line, and generally have weaker warming in the far eastern Pacific. These descriptions concern departures from local seasonal averages, not the location of the warmest absolute water temperatures. The 1997–1998 and 2004–2005 events are commonly used as contrasting EP and CP examples, respectively; both nevertheless involved warming across substantial portions of the tropical Pacific. (repository.library.noaa.gov)

“Central-Pacific El Niño” is a broad geographical description, whereas El Niño Modoki originated with a more specific spatial pattern. In 2007, Karumuri Ashok and colleagues described central-Pacific warming flanked by cooler-than-average waters to both the east and west; modoki is a Japanese expression meaning “similar but different.” Their El Niño Modoki Index measures the central region’s temperature anomaly minus half the anomaly in each of two flanking regions. Consequently, it measures a three-region contrast rather than central warming alone. A classification based simply on whether central-Pacific warming exceeds eastern-Pacific warming need not identify the same events as this index. (agupubs.onlinelibrary.wiley.com)

The EP–CP distinction does not establish two sharply separated classes within ENSO. Events differ in the longitude, breadth, and duration of warming, and some combine characteristics of both patterns. Different classification methods emphasize regional averages, statistical spatial patterns, or the location of maximum warming, producing different assignments for borderline cases. A synthesis led by Antonietta Capotondi in 2015 described ENSO as a diverse continuum of coupled ocean–atmosphere events, rather than a phenomenon adequately represented by two fixed templates. (repository.library.noaa.gov)

Coastal El Niño refers to anomalous warming near Peru and Ecuador, which can accompany basin-scale El Niño but can also develop without comparable central-Pacific warming. In January 2017, coastal temperatures rose abruptly above 26°C at several northern Peruvian locations while the central equatorial Pacific was transitioning from La Niña toward neutral conditions. Peru’s ENFEN monitoring commission formally classified December 2016–May 2017 as a moderate coastal El Niño. Research reported local coastal anomalies reaching approximately +5°C. Weakened southerly winds reduced evaporative cooling and altered ocean circulation and upwelling; coastal warming and enhanced atmospheric convection then helped sustain a regional positive feedback. This was not simply the coastal expression of a simultaneous basin-wide warm event. (enfen.imarpe.gob.pe)

Spatial structure and timing therefore provide information that a single strength index cannot retain. An average over Niño 3.4 cannot show whether warming reaches the South American coast or is concentrated near the date line, nor whether the accompanying atmospheric convection shifts with it. Seasonal background temperatures also matter: coastal warming documented in July 2008 did not produce strong precipitation because absolute temperatures remained insufficiently high, whereas warming during the austral summer of 2017 supported intense convection and rain. Event descriptions consequently distinguish anomaly magnitude from absolute temperature, specify the season and geographical extent, and examine atmospheric coupling alongside the oceanic pattern. The associated regional climate responses are not interchangeable between events with similar index values. (frontiersin.org)

Atmospheric teleconnections and regional climate

During El Niño, deep tropical convection shifts eastward from the western Pacific toward the central and sometimes eastern equatorial Pacific. Condensation in these thunderstorms releases latent heat, relocating a major source of atmospheric heating. The resulting upper-level outflow excites nearly stationary Rossby wave trains that propagate poleward and eastward, producing alternating pressure anomalies far from the tropical Pacific. These atmospheric teleconnections alter prevailing winds and temperature advection. Tropical heating also strengthens and extends the subtropical ridges eastward, helping shift the Pacific jet streams equatorward and extend them toward the Americas. Because these currents steer midlatitude weather systems, their displacement changes storm tracks; extratropical responses are generally strongest during each hemisphere’s winter. (pmel.noaa.gov)

Along western South America, rainfall responses differ sharply between coastal, mountain, and interior regions. Coastal Ecuador and northern Peru can receive exceptional rainfall during austral summer when strong local ocean warming supports convection; this is not a uniform response throughout the Andes. Farther south, parts of central Chile tend to receive more winter precipitation as Pacific storm tracks shift. Elsewhere on the continent, northern areas commonly become drier, whereas southeastern South America, including southern Brazil, Uruguay, and northeastern Argentina, tends toward wetter conditions. Central America’s Pacific-side Dry Corridor commonly experiences reduced rainfall during its rainy season, particularly between June and December. (pmel.noaa.gov)

Suppressed convection over Indonesia favors deficient rainfall, especially during the dry season and transition toward the wet season. In Australia, the clearest drying tendency occurs across eastern and central regions during winter and spring, rather than uniformly across the continent or throughout the year. Bureau of Meteorology comparisons illustrate this variability: winter–spring 2002 was widely dry, whereas 2015 showed a more mixed spatial pattern. In South Asia, El Niño generally reduces the June–September Indian summer monsoon, but does not guarantee a deficit: the 1997 monsoon produced above-average rainfall despite the developing strong event. Southeastern India and Sri Lanka show a contrasting tendency toward enhanced October–December rainfall during the northeast monsoon season. (bom.gov.au)

Across Africa, the season and subregion are equally important. Equatorial East Africa commonly receives enhanced October–December “short rains,” while much of southern Africa tends toward deficient November–March rainfall. Parts of the Sahel and western sectors of East Africa instead show drying tendencies during July–September. In North America, the winter circulation commonly favors wetter, cooler conditions along the southern United States, including the Gulf Coast, and milder conditions across much of Canada and the northern United States. The Pacific Northwest often becomes drier. California’s rainfall response varies substantially among events, so the southern wet tendency should not be interpreted as a statewide rainfall guarantee. (fao.org)

El Niño also reorganizes tropical cyclone environments. Stronger upper-level westerlies over the Caribbean and tropical Atlantic generally increase vertical wind shear—the change in wind speed or direction with height—which disrupts the vertical alignment needed for cyclone development and intensification. Atlantic seasonal activity therefore tends to decrease, while eastern and central North Pacific activity tends to increase under more favorable circulation and shear conditions. In the western North Pacific, an important response is geographical: cyclone formation commonly shifts southeastward, changing tracks and the distance storms can travel over warm water. South Pacific cyclone activity likewise tends to extend farther east as the convergence zone shifts. (climate.gov)

These relationships describe changes in probability, not deterministic outcomes. Their strength depends on the location of tropical warming, the season, other oceanic conditions, and internal atmospheric variability. A seasonal rainfall deficit can coexist with a damaging rainstorm, and reduced Atlantic hurricane activity does not exclude a destructive landfall. Attributing an individual storm or drought requires a separate analysis of its circulation, moisture supply, and other contributing influences; occurrence during El Niño alone does not establish causation. (aoml.noaa.gov)

Ecological and biogeochemical effects

El Niño reduces the delivery of subsurface nutrients to sunlit waters in much of the central and eastern equatorial Pacific and along the coast of Peru. Weaker upwelling and changes in the depth of nutrient-rich waters restrict the supply supporting phytoplankton, the photosynthetic component of plankton. Their reduced growth lowers primary production and the energy available to marine food webs. During the 1997–1998 event, Francisco Chavez and colleagues measured nutrient-depleted surface waters and the lowest equatorial Pacific chlorophyll concentrations in the observational record available to their 1999 study. The response reversed rapidly in mid-1998: renewed nutrient supply produced an exceptionally extensive phytoplankton bloom, illustrating that biological recovery can accompany the physical transition out of El Niño. (aoml.noaa.gov)

Marine organisms respond through redistribution as well as changes in abundance. Peruvian anchovy can become scarce in warmer northern waters and more prevalent in cooler waters farther south, while reduced phytoplankton availability diminishes its food supply. Warm-water species may become more common in areas previously dominated by cold-water assemblages. Tuna responses differ from those of coastal forage fish: Patrick Lehodey and colleagues’ 1997 study linked shifts in skipjack distribution to large east–west movements of the western Pacific warm pool; tuna fishing grounds generally move eastward during El Niño. Such displacement changes where fish are available to fisheries, rather than necessarily indicating an equivalent change in total stock size. Reduced or redistributed prey also affects marine predators. During 2015–2016, many California Current seabirds showed reduced breeding participation and success, slower chick growth, and lower fledging weights. (aoml.noaa.gov)

In coral reefs, prolonged anomalous warmth can disrupt the association between corals and their photosynthetic symbiotic algae. Loss of these algae causes coral bleaching, depriving corals of a major food source. Bleaching is not itself death: corals can recover when stress subsides, but persistent exposure increases mortality. The 2015–2016 El Niño intensified and spread the global bleaching episode that extended from 2014 to 2017, against a background of long-term ocean warming. NOAA documented continuous central Pacific heat stress from April 2015 to May 2016; by May 2016, surveys at Jarvis Island reported approximately 98% of coral cover dead. These losses altered reef habitat, not merely coral coloration. (aoml.noaa.gov)

Terrestrial responses depend on regional rainfall and temperature anomalies. In tropical forests, drought can restrict photosynthesis, while elevated temperatures increase respiratory carbon losses; dry conditions also facilitate fire. These mechanisms differed across continents during 2015–2016. Satellite-based analysis led by Junjie Liu identified reduced plant uptake as the principal contribution in tropical South America, increased respiration in tropical Africa, and increased fire emissions in tropical Asia, particularly Indonesia. Consequently, apparently similar increases in net carbon release need not represent the same ecological process: reduced growth, accelerated decomposition and respiration, and combustion have different effects on vegetation and stored carbon. (jpl.nasa.gov)

Oceanic and terrestrial effects on the carbon cycle can have opposite signs. The equatorial Pacific normally releases carbon dioxide because upwelling brings dissolved-inorganic-carbon-rich water to the surface. During El Niño, reduced delivery of this water suppresses outgassing, despite the accompanying decline in photosynthetic carbon uptake. Chavez and colleagues estimated that the equatorial Pacific retained about 0.7 billion tonnes of carbon during 1997–1998 that would otherwise have entered the atmosphere. On land, however, reduced uptake and increased respiration and fire can outweigh this oceanic reduction. Liu’s analysis estimated that tropical South America, Africa, and Asia together contributed approximately 2.5 billion tonnes more carbon to the atmosphere in 2015 than in the comparison year 2011. This is a difference in net land–atmosphere exchange, not a measure of fire emissions alone. (pubmed.ncbi.nlm.nih.gov)

Societal effects and use of climate information

El Niño-related climatic hazards become societal losses through the interaction of exposure and vulnerability: rainfall deficits have different consequences for rainfed farms and irrigated fields, while floods cause greater damage where settlements, roads, and utilities occupy flood-prone terrain. Agricultural effects depend on whether heat or water stress coincides with planting, flowering, or grain filling, and on preceding harvests, food stocks, and purchasing power. During the 2015–2016 event, drought delayed planting and impaired crop development across southern Africa. A February 2016 joint assessment by FAO, the European Commission’s Joint Research Centre, FEWS NET, and the World Food Programme reported that regional cereal production had already fallen by 23% in the preceding season. Depleted stocks and elevated prices therefore compounded the new climatic shock; the assessment cited approximately 14 million food-insecure people in the region, excluding South Africa, even before the unfolding crisis. (fao.org)

Fishing, water supply, electricity, and transport transmit impacts beyond agriculture. Off Peru, altered nutrient-rich upwelling and changes in fish distribution can reduce catches and disrupt fishmeal exports, affecting processing employment and international animal-feed markets. During 1997–1998, fish moved southward or into deeper water, while flooding damaged roads, housing, electricity services, schools, and hospitals in Peru. Drought creates a different chain of disruption: diminished river discharge and reservoir storage constrain water availability and hydroelectricity, with consequences for industry and public services. At Kariba, shared by Zambia and Zimbabwe, a World Bank assessment reported reservoir storage at about 11% on January 27, 2016, compared with more than 50% a year earlier; estimated potential annual electricity generation had fallen by more than half, to 4,060 GWh. Reservoir conditions reflect both water inflows and operating decisions, so the electricity shortfall is not simply a direct measure of El Niño strength. (elibrary.imf.org)

Economic outcomes vary among sectors and need not have the same sign within a country. Additional rainfall can benefit crops where moisture is limiting but damage harvests and infrastructure where flooding occurs; trade and commodity prices can transmit effects to countries outside the directly affected regions. Paul Cashin, Kamiar Mohaddes, and Mehdi Raissi’s 2015 IMF study used a multicountry econometric model and observations from 1979–2013 to estimate responses of output, inflation, and commodity prices. It found negative growth responses in several economies, including Australia, India, and Indonesia, but positive responses in others, including Argentina, Canada, and the United States. These are model-based responses to an El Niño shock, not audited losses or benefits from a particular episode. Repair costs, lost production, and estimated changes in gross domestic product measure different quantities and cannot be combined indiscriminately. (imf.org)

Documented public-health associations involve several pathways rather than a single disease response. Food shortages can increase malnutrition; water scarcity and flood damage can impair sanitation; altered rainfall and temperature can change mosquito habitats; and damaged facilities can interrupt health services. WHO’s January 2016 assessment described malaria epidemics in Ethiopia and dengue outbreaks in Indonesia and Pacific islands during 1997–1998, alongside water-borne disease outbreaks in several affected countries. These geographically specific associations do not establish that El Niño alone caused an outbreak. Water and sanitation access, displacement, nutritional status, and health-service capacity also influence disease transmission and outcomes, complicating causal attribution. (cdn.who.int)

Institutions use seasonal outlooks to organize early warning and resource allocation, rather than to predict individual disasters. WMO combines forecasts from global long-range prediction centres with expert assessment; national meteorological and hydrological services interpret the implications locally. FAO’s 2023 anticipatory-action plan combined seasonal forecasts with agricultural calendars, historical impacts, and existing vulnerability to prioritize assistance. Reported measures included drought-tolerant seeds, livestock protection, prepositioned supplies, and repairs to approximately 4,600 metres of breached riverbank along Somalia’s Shabelle River, intended to protect almost 40,000 hectares of cropland. Forecast usefulness remains conditional: a high probability of El Niño persistence is not the same as a high probability of flooding at a particular location, and action requires funding, delivery capacity, and sufficient lead time before planting or peak impacts. (wmo.int)

Scientific history and notable episodes

Late nineteenth-century descriptions treated El Niño chiefly as a coastal occurrence off Peru, where unusually warm water sometimes accompanied exceptional rainfall and changes in marine life. Its reinterpretation as a basin-wide phenomenon developed through two initially separate lines of research: observations of Pacific ocean temperatures and statistical studies of atmospheric variability. In the 1920s and 1930s, Gilbert Walker and his collaborator E. W. Bliss sought predictors of the Indian monsoon using worldwide pressure, temperature, and rainfall records, often about 40 years long. Their seasonal correlation analyses identified a pressure seesaw between the Pacific and Indian Ocean regions, which Walker named the Southern Oscillation. This established a large-scale atmospheric pattern without yet supplying its ocean–atmosphere mechanism. (aoml.noaa.gov)

By about 1960, scientists recognized that warming off Peru could extend westward along the equator toward the date line. Jacob Bjerknes brought the oceanic and atmospheric evidence together in papers published in 1966 and 1969. Drawing on observations from the 1957–1958, 1963–1964, and 1965–1966 warm episodes, he related weakened equatorial trade winds to warmer surface waters and eastward-displaced rainfall. His coupled explanation made ocean temperature changes both a cause and a consequence of atmospheric circulation changes: weaker winds reduced equatorial upwelling, while the resulting warming weakened the east–west temperature contrast and associated circulation. Bjerknes named the equatorial overturning circulation the Walker circulation, honoring Walker’s earlier research. These connections provided the physical foundation for treating El Niño and the Southern Oscillation together as ENSO. (aoml.noaa.gov)

The 1982–1983 episode exposed major deficiencies in Pacific monitoring: it was not predicted, and its development was not recognized until it was already well advanced. Its scientific importance lay partly in demonstrating the need for sustained, rapidly available measurements rather than intermittent research expeditions. At NOAA’s Pacific Marine Environmental Laboratory, Stan Hayes directed development of the Autonomous Temperature Line Acquisition System, or ATLAS, a mooring that measured surface meteorological conditions and subsurface ocean temperatures and transmitted data through satellites. Prototypes were tested in early 1984, followed by a small array along 110°W later that year. This engineering response became the basis of the Tropical Atmosphere Ocean array (TAO). (pmel.noaa.gov)

The international Tropical Ocean Global Atmosphere programme (TOGA), conducted from 1985 to 1994, expanded this effort into a coordinated observing and research system. Alongside TAO, it supported drifting buoys, tide gauges, shipboard subsurface-temperature measurements, and satellite observations of winds, sea-surface temperature, and sea level. These records enabled researchers to test proposed ENSO mechanisms and develop coupled climate models for seasonal prediction. The nearly 70-mooring TAO array was completed in December 1994; its construction involved more than 400 buoy deployments on 83 cruises using 17 ships from six countries. The array continued after TOGA, turning an experimental research network into sustained observing infrastructure. (pmel.noaa.gov)

The 1997–1998 and 2015–2016 episodes illustrate why comparisons require an identified index and observational reconstruction. In NOAA’s ERSSTv6-based Oceanic Niño Index—a three-month mean of Niño 3.4 temperature anomalies calculated against centered 30-year reference periods updated every five years—the respective peaks were +2.4°C for November 1997–January 1998 and +2.6°C for November 2015–January 2016. That comparison concerns the central-eastern Pacific regional mean, not every dimension of event intensity. ECMWF’s analysis using a fixed 1981–2010 reference climate found especially large Niño 3.4 and Niño 4 anomalies in 2015–2016, whereas eastern-Pacific Niño 3 and coastal Niño 1+2 warming remained below the 1997–1998 levels. Its equatorial temperature record also showed shorter-lived peak warming in 2015–2016. Thus the two extensively observed episodes differed in geographical concentration and persistence as well as in their index values. (cpc.ncep.noaa.gov)

Observation and forecasting

El Niño monitoring combines sustained measurements at fixed locations with mobile instruments and satellite coverage. The Tropical Atmosphere Ocean (TAO) array, completed with nearly 70 moorings in December 1994, established a basin-scale network of equatorial Pacific observations; the subsequent TAO/TRITON system measures surface winds, sea-surface temperature, humidity, air temperature, and subsurface temperatures, including measurements at ten depths within the upper 500 metres. These records reveal changes below the surface as well as atmospheric conditions associated with ocean warming. Near-real-time transmission makes the observations available for monitoring and forecast initialization. Argo profiling floats provide complementary temperature and salinity profiles: standard floats typically complete a cycle every ten days, sampling during ascent from approximately 2,000 metres and transmitting their records after surfacing. (pmel.noaa.gov)

Ships supply ocean-temperature profiles and surface meteorological measurements along repeated routes, while island and coastal tide gauges measure sea-level variations at fixed sites. Satellite remote sensing extends coverage across the basin, observing sea-surface temperature, surface winds, and sea-surface height. These measurements are complementary rather than interchangeable: satellite observations describe broad spatial patterns, whereas moorings and profiling instruments directly sample subsurface structure. Sea-level observations also help constrain estimates of the ocean’s evolving thermal structure when combined with temperature and salinity measurements in an assimilation system. The integrated observing network described by Michael McPhaden and colleagues in their OceanObs99 assessment included ships, moorings, drifting buoys, tide gauges, and satellites, rather than relying on a single instrument type. (pmel.noaa.gov)

Data assimilation combines observations with a model-generated estimate of the ocean–atmosphere state, accounting for uncertainties in both. It produces spatially complete estimates from measurements that are unevenly distributed in space and time. For example, ECMWF’s OCEAN5 system assimilates temperature and salinity profiles from moorings, ships, and Argo floats, together with satellite-altimeter sea-level anomalies; sea-surface temperature observations additionally constrain upper-ocean temperature. The resulting analysis supplies initial conditions for forecasts. Reanalysis applies this approach retrospectively to historical observations, providing evolving gridded fields for research and forecast evaluation. Such fields remain model-based estimates: changes in observing coverage and instrument availability can introduce artificial variability, requiring quality control and bias correction. (ecmwf.int)

Statistical forecasts estimate future ENSO conditions from historical relationships among observed variables, using methods such as regression and time-series modelling. Dynamical forecasts instead initialize coupled ocean–atmosphere models and integrate equations representing physical processes forward in time. Ensembles sample multiple possible evolutions, while multi-model ensembles combine systems with different formulations; the North American Multi-Model Ensemble includes contributions from NOAA, NASA, and Environment and Climate Change Canada. Neither approach has uniform skill across seasons and lead times. A 2024 analysis by Azhar Ehsan and colleagues, discussed by NOAA, found that statistical forecasts issued before or during boreal spring lost skill more rapidly for subsequent seasons than dynamical forecasts. (climate.gov)

Forecast lead time separates initialization from the target period, which is commonly a three-month season. Probabilistic outlooks assign separate chances to El Niño, neutral conditions, and La Niña rather than presenting an ensemble mean as a certain outcome. Verification compares predictions with subsequent observations: correlation evaluates correspondence between forecast and observed temperature anomalies, while ranked probability skill scores assess categorical probabilities against a reference forecast, commonly climatological frequencies. In an evaluation covering June 2015–September 2018, Michelle L’Heureux and colleagues found that combining forecasters’ temperature predictions with statistically estimated uncertainty improved probabilistic performance relative to the comparison methods. This result illustrates why forecast calibration and verification matter alongside model development. (climate.gov)

The boreal spring predictability barrier is a seasonally concentrated reduction in ENSO forecast skill, not an absolute prohibition on prediction. Forecasts crossing spring are particularly difficult because this is frequently a period of event decay or transition, when the emerging ocean–atmosphere signal is less clearly established. Consequently, forecasts with identical lead times can have different skill depending on their initialization month. Predicting ENSO also differs from predicting its local consequences: a probability of Pacific warming is not the probability of drought, flooding, or above-average rainfall at a particular location. Regional seasonal outlooks must separately estimate those outcomes, and they do not specify the timing of individual weather events. (climate.gov)

Long-term variability and climate change

Paleoclimate reconstructions extend the history of ENSO beyond instrumental observations by measuring environmental signals preserved in corals, sediments, tree rings, and cave deposits. Coral skeletal oxygen-isotope ratios, expressed as $\delta^{18}\mathrm{O}$, reflect both water temperature and seawater isotopic composition, the latter influenced by rainfall, evaporation, and salinity. Strontium-to-calcium ratios provide a complementary temperature proxy. Annual growth bands establish relative chronology, while uranium–thorium dating places fossil corals in time; closely spaced samples can resolve monthly variations. These records require calibration and screening for preservation and biological effects. They are not interchangeable with instrumental temperature indices: a 2023 study at Palmyra Atoll found that seawater isotopic changes, rather than temperature, accounted for more than half of the coral oxygen-isotope signal during the 2014–2015 and 2015–2016 events. (ncei.noaa.gov)

Sedimentary archives can cover longer intervals but introduce different uncertainties. At Lake Pallcacocha in Ecuador, pale layers deposited during episodes of erosion and flooding have been interpreted as evidence of El Niño-related heavy rainfall. Radiocarbon dates constrain the chronology, but uncertainty in ages, sediment accumulation, and the climatic interpretation of individual layers affects estimates of event frequency. Crucially, a rainfall-sensitive archive records a local response rather than equatorial Pacific temperature directly. Changes in atmospheric circulation or catchment processes can therefore change its relationship to ENSO. Tree-ring, cave-deposit, and other geographically remote reconstructions face a related problem: the climatic connection between the archive location and the tropical Pacific need not remain constant over millennia. (whoi.edu)

Reconstructions show substantial variability over decades, centuries, and millennia rather than a simple progression toward stronger events. Kim Cobb and colleagues’ 2013 study of monthly resolved fossil corals from the Northern Line Islands sampled intervals across the preceding 7,000 years. It found highly variable ENSO activity without a systematic trend in variance; twentieth-century variance exceeded the fossil-record average but was not unprecedented. The IPCC Sixth Assessment Report, published in 2021, assessed with medium confidence that ENSO variability since 1950 exceeded that between 1400 and 1850, but expressed low confidence that it exceeded variability before 1400. The comparatively short instrumental record, together with strong multidecadal fluctuations, limits the detection of sustained trends. That assessment also assigned low confidence to attributing reconstructed changes in ENSO variability to anthropogenic forcing. (pubmed.ncbi.nlm.nih.gov)

El Niño’s temporary contribution to global surface warming is distinct from sustained climate change. ENSO alters the distribution of ocean heat and where it is transferred to the atmosphere; El Niño commonly raises global mean surface temperature, whereas La Niña temporarily suppresses it. These fluctuations occur against a warming background created by increasing greenhouse gas concentrations and changes in Earth’s energy budget. An El Niño can consequently help produce a record-warm year without being the cause of the underlying multidecadal trend. Simulations incorporating natural forcing and internal variability, including ENSO, cannot reproduce the observed long-term warming without human influence. (noaa.gov)

Future changes in ENSO-related rainfall are more robustly assessed than changes in sea-surface-temperature amplitude. The IPCC’s 2021 assessment concluded that rainfall variability associated with ENSO would very likely increase significantly by the second half of the twenty-first century under scenarios SSP2-4.5, SSP3-7.0, and SSP5-8.5, even without increased ENSO temperature variability. Greater atmospheric moisture can strengthen precipitation responses to a given circulation anomaly. By contrast, projected changes in ENSO temperature amplitude remain highly uncertain, and models show no robust amplitude change across all four assessed scenarios. Increased rainfall variability therefore does not, by itself, establish that temperature-defined El Niño events will become uniformly stronger or more frequent, or that one spatial event type will dominate. (ipcc.ch)

Research on changing distant climatic effects examines both ENSO anomalies and the warmer background state through which they act. The IPCC assessed an eastward shift of ENSO teleconnections over the North Pacific and North America as likely, while noting limited agreement among models on wider teleconnection changes. Persistent model biases—including errors in the equatorial cold tongue and the longitudinal extent and meridional width of ENSO anomalies—complicate these projections. Long simulations without changing external forcing and large ensembles with identical forcing but different initial conditions help distinguish internally generated fluctuations from forced responses. Longer, geographically replicated proxy records provide complementary tests of whether apparent changes in amplitude, frequency, or event diversity exceed natural variability. (ipcc.ch)

References

1. The “El Niño” FAQaoml.noaa.gov
2. What is the El Niño–Southern Oscillation (ENSO) in a nutshell?content-drupal.climate.gov
3. Climate Prediction Center - ENSO FAQcpc.ncep.noaa.gov
4. Climate Variability: Oceanic Niño Indexclimate.gov
5. The Tropical Ocean-Global Atmosphere observing system: A decade of progresspmel.noaa.gov
6. The upper ocean heat balance in the western equatorial Pacific warm pool during September-December 1992pmel.noaa.gov
7. The circulation of the eastern tropical Pacific: A reviewpmel.noaa.gov
8. WP03 TPOS_ENSOpmel.noaa.gov
9. What is upwelling?oceanservice.noaa.gov
10. Equatorial Walker circulationcpc.ncep.noaa.gov
11. El Niño and La Niña: Causes and Global Consequencespmel.noaa.gov
12. Forcing of intraseasonal Kelvin wavespmel.noaa.gov
13. The March 1997 Westerly Wind Event and the onset of the 1997/98 El Niño: understanding the role of the atmospheric responsenora.nerc.ac.uk
14. El Niño and La Niña: Frequently asked questionsclimate.gov
15. Observations of Warm Water Volume Changes in the Equatorial Pacific and Their Relationship to El Niño and La Niñapmel.noaa.gov
16. Observations of Warm Water Volume Changes in the Equatorial Pacific and Their Relationship to El Niño and La Niña: Summarypmel.noaa.gov
17. Archive Of Upper Ocean Heat Content Computed From BMRC Temperaturespmel.noaa.gov
18. Evolution of the 2002/03 El Niñopmel.noaa.gov
19. Climate Prediction Center — Current Monthly Atmospheric and Sea Surface Temperatures Index Valuescpc.ncep.noaa.gov
20. Climate Prediction Center — Oceanic Niño Index (ONI)cpc.ncep.noaa.gov
21. CPC adopts Relative Oceanic Niño Index (RONI) for reliable, responsive monitoring and tracking of ENSOcpc.ncep.noaa.gov
22. Climate Prediction Center — Relative Oceanic Niño Index (RONI)cpc.ncep.noaa.gov
23. Relative Oceanic Niño Index (RONI): What’s Changing and Why It Matterscpc.ncep.noaa.gov
24. Southern hemisphere monitoringbom.gov.au
25. Understanding El Niño & ENSOnoaa.gov
26. Historical ENSO eventsdata.jma.go.jp
27. Understanding ENSO Diversityrepository.library.noaa.gov
28. Researchers investigate whether distinct 'flavors' of El Nino/La Nina existpsl.noaa.gov
29. Forcings and Evolution of the 2017 Coastal El Niño Off Northern Peru and Ecuadorfrontiersin.org
30. Climate Prediction Center - ENSO - Jetstreamcpc.ncep.noaa.gov
31. El Niño | FAOfao.org
32. El Niño: what it means for Australia's climatebom.gov.au
33. ENSO and the Indian Monsoon… not as straightforward as you’d thinkcontent-drupal.climate.gov
34. The Strengthening Relationship between ENSO and Northeast Monsoon Rainfall over Sri Lanka and Southern Indiajournals.ametsoc.org
35. Weather Impacts of ENSOnoaa.gov
36. Can We Blame El Niño?climate.gov
37. Impacts of El Niño and La Niña on the hurricane seasonclimate.gov
38. The Influence of ENSO Flavors on Western North Pacific Tropical Cyclone Activityrepository.library.noaa.gov
39. 2010 Climate Events Connected to El Niño or La Niñaclimate.gov
40. What is upwelling? - NOAA Ocean Explorationoceanexplorer.noaa.gov
41. Biological and chemical response of the equatorial pacific ocean to the 1997-98 El Ninopubmed.ncbi.nlm.nih.gov
42. El Niño Southern Oscillation and tuna in the western Pacificnature.com
43. Ocean Indicators We’re Tracking for El Niño’s Arrival on the West Coastfisheries.noaa.gov
44. Threats to Coral - NOAA/AOMLaoml.noaa.gov
45. What is coral bleaching?oceanservice.noaa.gov
46. Global Coral Bleaching 2014-2017: Status and an Appeal for Observationscoralreefwatch.noaa.gov
47. NASA Pinpoints Cause of Earth's Recent Record Carbon Dioxide Spikejpl.nasa.gov
48. El Niño set to have a devastating impact on southern Africa’s harvests and food securityfao.org
49. El Niño: FAO’s updated Anticipatory Action and Response Plan urgently requires $160 million to support over 4.8 million people through March 2024fao.org
50. Fair Weather or Foul? The Macroeconomic Effects of El Niñoelibrary.imf.org
51. World Bank Document: El Niño Emergency Assistance Projectdocuments1.worldbank.org
52. Energy–Water Nexus in Southern Africadocuments.worldbank.org
53. El Niño Good Boy or Bad?imf.org
54. El Niño and Healthcdn.who.int
55. El Niño affects more than 60 million peoplewho.int
56. World Meteorological Organization declares onset of El Niño conditionswmo.int
57. History of the SOI: NOAA Physical Sciences Laboratorypsl.noaa.gov
58. Building TAO | Global Tropical Moored Buoy Arraypmel.noaa.gov
59. Climate Prediction Center - ONIcpc.ncep.noaa.gov
60. The 2015/2016 El Niño and beyondecmwf.int
61. Data Telemetrypmel.noaa.gov
62. Argo programaoml.noaa.gov
63. The El Niño–Southern Oscillation (ENSO) Observing Systempmel.noaa.gov
64. The use of radar altimeter products at ECMWFecmwf.int
65. Ocean Reanalysis/analysisecmwf.int
66. Data assimilationecmwf.int
67. How good have models been at predicting ENSO in the 21st century?climate.gov
68. Team effort: the North American Multi-Model Ensembleclimate.gov
69. ENSO forecast mash-ups: What’s the best way to combine human expertise with models?content-drupal.climate.gov
70. The Spring Predictability Barrier: we’d rather be on Spring Breakclimate.gov
71. NOAA/WDS Paleoclimatology - Bermuda - Sr/Ca and d18O Data and SST Reconstructionncei.noaa.gov
72. Highly Variable El Niño–Southern Oscillation Throughout the Holocenewhoi.edu
73. Contrasting Central Equatorial Pacific Oxygen Isotopic Signatures of the 2014/2015 and 2015/2016 El Niño Eventsrepository.library.noaa.gov
74. Highly variable El Niño-Southern Oscillation throughout the Holocenepubmed.ncbi.nlm.nih.gov
75. Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epochwhoi.edu
76. Chapter 3: Human Influence on the Climate Systemipcc.ch
77. Where does global warming go during La Niña?prod-01-asg-www-climate.woc.noaa.gov
78. Chapter 4: Future Global Climate: Scenario-Based Projections and Near-Term Informationipcc.ch
79. Chapter 8: Water Cycle Changesipcc.ch