npj | clima e and a mosphe ic science A icle
Published in pa ne ship wi h CECCR a King Abdulaziz Uni e si y
h ps://doi.o g/10.1038/s41612-025-01003-1
Fu u e clima e esponse o obse ed
s ong El Niño analogues
Check o upda es
Paloma T ascasa-Cas o1,2 , Yohan Rup ich-Robe 1& Amanda C. Maycock2
Thee ec o u u eclima echange on he bo ealwin e esponse o s ong ElNiñoisin es iga edusing
pacemake simula ions wi h he EC-Ea h3-CC model cons ained owa ds obse ed opical Pacific
sea su ace empe a u e anomalies. Unde he Sha ed Socioeconomic Pa hway 2-4.5, he su ace
empe a u e esponse o s ong El Niño in ensifies in No h Ame ica, no he n A ica, Aus alia and he
No h A lan ic compa ed o p esen day. Howe e , u u e s ong El Niño has a weake clima e impac in
sou he n Ame ica and A ica. Tempe a u e ex emes unde s ong El Niño in ensi y in he u u e in
some egions, wi h mo e cool days in eas e n No h Ame ica, while wa m days in no he n Sou h
Ame ica dec ease. Assuming ha he cha ac e is ics o s ong El Niño e en s will no change in he
u u e, we dis inguish be ween changes in El Niño eleconnec ions and backg ound clima e changes,
and ound ha he la e domina es he absolu e clima e esponse o s ong El Niño e en s.
El Niño-Sou he n Oscilla ion (ENSO) is he dominan mode o in e annual
clima e a iabili y ac oss he global opics. E e y 2-7 yea s, anomalously
wa m o cool sea su ace empe a u e (SST) anomalies de elop in he
equa o ial PacificOceanand igge aglobalcascadeo emo ee ec s
1.The
wa m phase o ENSO, known as El Niño, leads o d y and ho condi ions
o e Aus alia2, no he n Sou h Ame ica3, sou he n A ica4,5and he Indian
Summe monsoon egion6,aswellaswe andcoole condi ionso e some
pa s o No h Ame ica7and Eas e n A ica8.
The s onges El Niño e en s since he la e 19 h cen u y occu ed in
1982/83, 1997/98, and 2015/16, when SST anomalies in he Niño3.4 egion
(SST a e aged o e 5°S-5°N, 170 W°-120°W) eached 2.2oC, 2.4oCand
2.6oC, espec i ely (es ima ed as 3-mon h unning means wi h he
ERSST. 5 da ase 9). These s ong El Niño e en s led o widesp ead de as-
a ing impac s, including se e e d ough s in Sou heas Asia, No he n and
Sou he n A ica10, heAmazon
11, Mexico and Philippines, and a mang o e
dieback in No he n Aus alia12; widesp ead fi es in Indonesia13,14;s o msin
he US, flooding in Cuba, Pe u, Boli ia and Ecuado 15;la gechangesin he
global ca bon cycle16,17; and ood insecu i y10,18.
Du ing s ong El Niño e en s, SST and p ecipi a ion anomalies ex end
u he eas in he equa o ial Pacific compa ed o mode a e e en s19–21
pa icula ly in Decembe -Janua y-Feb ua y (DJF). The eleconnec ions o
El Niño o he Indo-Pacificsec o
22,Aus alian
23,No he n
24 and Sou h
Ame ican25 and No h A lan ic-Eu opean26 egions depend on he ampli-
ude o he e en , wi h he impac s o s ong El Niño o en mo e se e e han
weak o mode a e e en s.
The global impac s o s ong El Niño e en s may change in he
u u e due o inc eases in an h opogenic adia i e o cing and, as a
di ec consequence, con inued global mean su ace wa ming. Such
changes could a ise om wo e ec s: 1) changes in he equency and/
o cha ac e is ics o s ong El Niño e en s; 2) changes in he p ocesses
ha lead o eleconnec ions. Bo h e ec s depend on he in e ac ions
be ween ENSO and he mean clima e s a e. Rega ding mechanism 1, in
he opical Pacific mos clima e models p ojec a mean weakening o
he Walke ci cula ion and a educ ion o he zonal SST g adien in he
equa o ial Pacific by he end o he 21s cen u y27. This change in mean
s a e leads o an inc ease o ENSO a iabili y unde inc eased
an h opogenic o cing in mos models om he Six h phase o he
Coupled Model In e compa ison P ojec (CMIP6)28,29, including an
inc ease in he likelihood o s ong El Niño e en s20,30 pa ially d i en
by a s ong inc ease in empe a u e o e he cen al and eas e n
equa o ial Pacific. Howe e , he e is unce ain y associa ed wi h his
esponse, in pa due o he gene ally poo ep esen a ion o ecen
opical Pacific SST ends in his o ical model simula ions31.In ac ,
despi e mos CMIP6 models simula ing an “El Niño-like”SST s a e in
he u u e, he models which mo e closely cap u e he obse ed he
opical Pacific SST end p ojec a con inua ion o he “La Niña-like
mean s a e”, which e en ually migh weake he ampli ude o El Niño
e en s in he u u e32.
Rega ding mechanism 2, he Six h Assessmen Repo o he
In e go e nmen al Panel on Clima e Change highligh s a obus end
owa ds an inc ease in ENSO-d i en p ecipi a ion a iabili y in a
wa me clima e33. Following he Clausius-Clapey on ela ionship,
absolu e humidi y inc eases by app oxima ely 7% pe Kel in, hence he
changes in opical humidi y d i en by ENSO a e expec ed o inc ease
in he u u e34.Yinge al.
35 ound ha an an h opogenically- o ced
signa u e in ENSO- ela ed p ecipi a ion a iabili y will eme ge a ound
1Ea h Sciences Depa men , Ba celona Supe compu ing Cen e , Ba celona, Spain. 2School o Ea h and En i onmen , Uni e si y o Leeds, Leeds, UK.
e-mail: [email p o ec ed]
npj Clima e and A mosphe ic Science | (2025) 8:116 1
1234567890():,;
1234567890():,;
2040, decades be o e a obus change in ENSO SST can be de ec ed in
models27. This means ha we migh expec changes in ENSO ele-
connec ions o eme ge be o e any changes in ENSO cha ac e is ics
may be de ec able om in e nal clima e a iabili y.
While se e al modelling s udies ha e examined ENSO ele-
connec ions in a u u e clima e, hey ha e gene ally used eely-
e ol ing coupled clima e models and he e o e do no pe mi a
sepa a ion o mechanisms 1 and 229,36–39.McG ego e al.
38 ound
significan changes in ENSO impac s o e a ound one hi d o global
land a ea in CMIP6 models, bu i was unclea wha ac o s caused
hese changes. A small numbe o s udies ha e isola ed mechanism 2
by p esc ibing iden ical SST anomalies on op o p esen and u u e
backg ound clima es40,41. These s udies ound an amplifica ion o
ENSO eleconnec ions in bo eal win e in he No h Pacificand
No h A lan ic-Eu opean sec o s unde u u e clima e condi ions.
Howe e , D oua d and Cassou40 used ENSO SST anomalies de i ed
om a model, which may no be ully ealis ic, and Zhou e al.41 used
an a mosphe ic model meaning any ole o a mosphe e-ocean cou-
pling in modi ying eleconnec ions was excluded. Nei he s udy
assessed he global e ec s o u u eclima echangesonElNiño
impac s and did no specifically add ess he impac s o s ong El Niño
e en s, which as no ed abo e a e o en he mos impac ul.
Ou s udy ollows a simila app oach o D oua d and Cassou40 and
Zhou e al.41 by explo ing mechanism 2 and assuming ha ENSO cha -
ac e is ics do no a y wi h clima e change. We specifically ocus on s ong
El Niño and use obse ed SST anomalies combined wi h a pacemake
modelling app oach o enable a mosphe e-ocean coupled in e ac ions
ou side o he opical Pacific (see Me hods). We simula e h ee obse ed
s ong El Niño e en s (1982/83, 1997/98 and 2015/16) unde p esen day
and p ojec ed u u e clima e s a es co esponding o he yea 2090 unde he
SSP2-4.5 scena io. We ocus on esponses in he bo eal win e season (DJF)
and find he e en s beha e simila ly in ou pacemake simula ions, so in
mos o he analyses we combine hem in o a mul i-e en mean (MEM).
Resul s
Changes in he opical Pacific induced by s ong El Niño and
clima e change
Figu e 1a shows he MEM El Niño SST anomalies unde p esen day. The
anomalies closely esemble he pacemake a ge SST pa e n, wi h he
canonical eas e n Pacific s uc u e o s ong El Niño. In he opical Pacific,
s ong El Niño e en s d i e an eas wa d shi o he p ecipi a ion maximum,
leading o p ecipi a ion anomalies o abo e 10 mm day−1o e he cen al
and eas e n equa o ial Pacific(Fig.1b), simila o he obse ed p ecipi a ion
pa e n du ing El Niño (Fig. S1). Those posi i e anomalies con as wi h
d ie condi ions o he equa o a ound 10°N and in he Sou h Pacific
Con e gence Zone.
The e a e e y small di e ences in he El Niño SST anomalies in he
u u e (Fig. 1c), indica ing he pacemake me hod wo ks as in ended (see
“Me hods”and Fig. S2). The e a e, howe e , significan di e ences in El
Niño-d i en p ecipi a ion in he u u e (Fig. 1d), wi h highe p ecipi a ion
a es in he eas e n (mo e we ing) and wes e n (less d ying) equa o ial
Pacifico up o2mmday
−1and nega i e anomalies (less we ing) o a leas
2mmday
−1o e he cen al equa o ial Pacific. We no e ha hese di e -
ences do no ully ma ch he p ojec ed change in El Niño-d i en p e-
cipi a ion anomalies in he opical Pacific sugges ed by he li e a u e,
consis ing o an eas wa d shi o p ecipi a ion du ing El Niño unde
g eenhouse38 o cing42. This migh indica e an impo an ole o El Niño-
SST changes in he u u e El Niño-p ecipi a ion ela ionship seen in coupled
clima e models, o be specific o he e ec o clima e change on he p e-
cipi a ion esponse o s ong El Niño e en s, which a e gene ally no isola ed
in o he s udies. I is also possible he pacemake se –up influences he
associa ed p ecipi a ion esponse, o example because i cons ains su ace
Fig. 1 | Changes in he opical Pacific induced by s ong El Niño and clima e
change. DJF SST (K, le ) and p ecipi a ion (mm day−1, igh ) anomalies du ing El
Niño in he p esen (a,b), di e ences be ween u u e and p esen El Niño anomalies
(c,d), he mean s a e change (e, ) and di e ence in El Niño s a es be ween u u e and
p esen day (g,h). Ha ched a eas a e no s a is ically significan a he 95%
confidence le el.
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 2
u bulen fluxes, bu simila di e ences we e ound in an expe imen wi h
weake flux es o ing (no shown).
In e ms o mean s a e changes, he opical Pacific wa ms by a ound
2.5Kin he u u e(Fig.1e), wi h he Niño4 (5°N-5°S, 160°E-150W) and
Niño3 (5°N-5°S, 150°W-90°W) egions being 4.7% and 6.6% wa me ,
espec i ely, compa ed wi h he p esen day. Consequen ly, he zonal eas -
wes equa o ial Pacific SST g adien dec eases by 19.5%, simila o he
esponse in o he CMIP6 models43. Coinciden wi h he mean SST changes,
he e is a weakening o he Walke ci cula ion, e idenced by a dec ease in
p ecipi a ion o e he wa m pool and an inc ease in p ecipi a ion o e he
Cen al and Eas e n Pacific(Fig.1 ).
To unde s and how s ong El Niño condi ions would appea in he
u u e compa ed o p esen day, we nex conside he combined mean s a e
plus El Niño anomalies (Fig. 1gand1h). Since he El Niño SSTs a e la gely
unchanged in he u u e, he o al nea -su ace empe a u e signal mainly
eflec s he change in opical Pacific mean s a e, whe eas o he o al
p ecipi a ion anomaly, he subs an ial nega i e anomalies be ween 170°E-
150°W seen as a di e ence in he esponse o El Niño in he u u e, expand
he loca ion o d y condi ions u he owa ds he cen al Pacific. Such a
change migh be expec ed o al e he eleconnec ions om El Niño, which is
explo ed in he nex sec ion.
Teleconnec ions om s ong El Niño e en s a e modula ed by
clima e change
In he p esen day, he modelled s ong El Niño d i es significan su ace
empe a u e impac s ac oss he globe in DJF (Fig. 2a). No he n Sou h
Ame ica expe iences wa m anomalies eaching up o 3 K o e he Amazon,
as well as la ge pa s o he A ican con inen , wi h he excep ion o nea
Fig. 2 | Fu u e global changes in s ong El Niño eleconnec ions in DJF. Su ace
empe a u e (K, le ) and p ecipi a ion (mm day−1, igh ) anomalies du ing El Niño
in he p esen (a,b), di e ences in u u e and p esen El Niño anomalies (c,d), mean
s a e changes (e, ) and changes in absolu e alues be ween u u e and p esen (g,h).
Ha ched a eas a e no s a is ically significan a he 95% confidence le el.
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 3
Lake Vic o ia. Wes e n Aus alia expe iences ho e summe s du ing s ong
El Niño, and sou h Asia also sees wa m anomalies. The e a e some egions,
hough, whe e El Niño leads o cold anomalies in DJF, such as he sou he n
and wes e n US and sou h-eas Sou h Ame ica (SESA). No e ha he EC-
Ea h3-CC model cap u es he ampli ude and spa ial loca ion o he win e
su ace esponse o El Niño o e no he n Sou h Ame ica, No h Ame ica
and Sou he n A ica simila o obse a ions (Fig. S1).
Alongside su ace empe a u e changes, s ong El Niño e en s d i e
d y anomalies o up o 2 mm day−1o e he Amazon basin, al hough based
on he model e alua ion hese anomalies may be o e es ima ed (Fig. S1),
and a ound 1 mm day−1o e Sou he n A ica. The e is a dipole p ecipi a-
ion anomaly o e he No h A lan ic, likely d i en by a sou hwa d shi o
he No h A lan ic je s eam and s o m ack (Fig. 2b), cha ac e is ic o he
link be ween El Niño and he nega i e phase o he win e No h A lan ic
Oscilla ion (NAO)44. O e he Congo basin and he Ho n o A ica, s ong El
Niño leads o weak bu significan posi i e p ecipi a ion anomalies. P e-
cipi a ion also inc eases by 2 mm day−1o e SESA du ing s ong El Niño,
leading o we e han usual summe s in he no h o A gen ina, U uguay
and Pa aguay. EC-Ea h3-CC e ec i ely cap u es he sign and spa ial
loca ion o posi i e p ecipi a ion anomalies du ing El Niño, bu sligh ly
unde es ima es i s ampli ude compa ed o obse a ions (Fig. S1). In e ms o
he abili y o he model o ep esen he complexi y o he amazon clima e, i
has shown o co ec ly simula e he annual cycle o empe a u e and p e-
cipi a ion, and ou pe o ms mos CMIP6 models in he ep esen a ion o
he sou h Ame ican monsoon dynamics45.
In he u u e, he DJF su ace empe a u e esponse o s ong El Niño
inc eases in se e al egions (Fig. 2c). S ong El Niño d i es e en ho e
condi ions in mos o A ica (excep in i s sou he n pa ), and ho e summe s
in he sou h-wes o he Amazon and Congo basins and eas e n Aus alia,
being he u u e esponse o El Niño in he la e s onge han in he p esen .
O e no h-eas No h Ame ica, he cold anomalies d i en by s ong El Niño
a e also enhanced in he u u e. On he o he hand, mean DJF empe a u e
esponsesd i enbys ongElNiñoin he u u eweakenino he egions
po en ially due o shi s in a mosphe ic ci cula ion. Fo example, cold
anomalies o e no h-eas A gen ina, Pa aguay and sou h B azil ex end mo e
owa ds he no h, ollowing he no hwa d shi o p ecipi a ion shown in
Fig. 2d. O e sou he n A ica, he empe a u e and p ecipi a ion esponses o
s ong El Niño weaken in ou simula ion, leading o milde summe s.
In Aus alia, he influence o u u e clima e change on he esponse o
s ong El Niño shows dipole anomalies consis ing o highe p ecipi a ion
a es o e wes e n Aus alia and d ie condi ions o e he eas in DJF, which
is opposi e o he p ojec ed p ecipi a ion end in aus al summe in he
model (Fig. 2 ). We no e ha o he clima e models end o show an inc ease
in Aus alian summe ain all in u u e p ojec ions33,whichwouldcon-
s uc i ely in e e e wi h he al e ed s ong El Niño signal in wes e n Aus-
alia. In a limi ed numbe o egions, he p ecipi a ion esponse o s ong El
Niño e en s weakens in he u u e compa ed o p esen day. Fo ins ance, in
he p esen day s ong El Niño d i es d y anomalies o e he Amazon basin
(Fig. 2b),bu hischangesin he u u e,whe e he me eo ological d ough
induced by s ong El Niño becomes significan ly weake (Fig. 2d).
Figu e 2e, show he mean s a e su ace empe a u e and p ecipi a ion
changes be ween he u u e and p esen day. By he la e 21s cen u y, EC-
Ea h3-CC simula es an o e all s a is ically significan inc ease in em-
pe a u e ac oss he globe, especially in he No he n Hemisphe e (in
ag eemen wi h Lee e al.33), whe e land a eas expe ience a wa ming o abo e
4 K in DJF. In he Sou he n Hemisphe e, egions like wes e n Aus alia,
sou he n A ica and he Amazon basin a e wa ming as e han o he s. In
e ms o p ecipi a ion, besides he p ojec ed eas wa d shi in opical Pacific
p ecipi a ion esul ing om a u u e weakening o he Walke ci cula ion,
he e’s an o e all inc ease in p ecipi a ion ac oss he opics, in ag eemen
wi h CMIP6 p ojec ions33. Regions like Aus alia, No h Ame ica and Eas
Asia a e expec ed o expe ience a sligh inc ease p ecipi a ion in DJF, in
ag eemen wi h mul imodel p ojec ions, al hough he change in p ecipi a-
ion is no as e iden as in empe a u e due o he highe a iabili y o he
o me and he la ge model unce ain y ega ding dynamical changes as a
esponse o clima e change. The o al anomalies du ing s ong El Niño in he
u u e compa ed o p esen day (Fig. 2g, h) show cons uc i e in e e ence
and exace ba ed El Niño-d i en wa m anomalies o e eas e n Aus alia,
No he n Sou h Ame ica and Cen al A ica du ing he aus al summe
mon hs, which would lead o inc eased hea - ela ed isks46,47.Themean
wa ming unde clima e change masks any enhancemen o he cooling
e ec o s ong El Niño, such as o e No h Ame ica o Eas Asia. Ne e -
heless, o e Sou h Ame ica, Fig. 2h highligh s he con ibu ion o changes
in s ong El Niño impac s o he absolu e esponse o e he Amazon basin,
which d i es an inc ease in p ecipi a ion in DJF, opposing he mean d ying
unde global wa ming.
T oposphe ic ci cula ion esponse o s ong El Niño and i s
changes unde u u e condi ions
To unde s and u u e changes in eleconnec ions o s ong El Niño, we
explo e oposphe ic geopo en ial heigh a 500 hPa (z500) and zonal wind
a 200 hPa (u200) anomalies in DJF. The ex a opical eleconnec ions o El
Niño depend on he backg ound flow, specificallyon he uppe oposphe ic
zonal wind48, which is s onge in he No he n hemisphe e in DJF due o a
s onge sub opical je . In he p esen day, s ong El Niño gene a es posi i e
z500 anomalies o up o 60 m o he coas o Japan (Fig. 3a), a deepening o
he Aleu ian low and ba o opic Rossby wa e ains ha p ojec on o he
Pacific Sou h Ame ican (PSA1) and Pacific No h Ame ican (PNA) pa -
e ns. As opposed o mode a e and weak El Niño e en s, s ong episodes
shi he PNA u he eas 24, making hese e en s unique and incompa able
wi h El Niño e en s o lesse ampli ude.
Ou side o he Pacific sec o , s ong El Niño d i es nega i e geopo-
en ial anomalies o e he sou h-eas USA, which ex end eas wa d o e he
sub opical A lan ic Ocean. Those nega i e anomalies con as wi h posi i e
anomalies spanning om he no h-eas o Canada o he No wegian Sea.
This co esponds o a nega i e phase o he NAO and a sou hwa d shi o
he No h A lan ic wes e ly je s eam, in ag eemen wi h he sou hwa d
shi o p ecipi a ion anomalies shown in Fig. 2b.
Unde u u e condi ions, s ong El Niño d i es a weake PNA
esponse, wi h he simula ions showing a ela i ely s onge Aleu ian low in
bo eal win e wi h posi i e z500 anomalies o abo e 40 m (Fig. 3b). The
eas wa d shi o z500 anomalies in he No h Pacific and No h Ame ica, in
ag eemen wi h Johnson e al.49 and Geng e al.50 coincide wi h below no mal
empe a u es shown in Fig. 2c. Simila condi ions a e shown o e cen al
and no he n Eu ope, whe e s ong El Niño e en s d i e ela i ely low
p essu e anomalies associa ed wi h colde win e s. The e is a high p essu e
anomalyin heCa ibbean egiondu ings ongElNiñoe en sin he u u e,
which ansla es in o he d ying end shown in Fig. 2c. In he Sou he n
hemisphe e, s ong El Niño igge s an anomalous wa e ain spanning he
ex a opics and subpola egions and d i ing a dipole o posi i e and
nega i e z500 anomalies in he wes and eas e n coas s o Sou h Ame ica,
espec i ely. The p ojec ed changes in El Niño-d i en z500 o e Sou h
Ame ica, do no esemble hose shown in Johnson e al.49 possibly due o he
weake clima e change signal in ou simula ions (compa ed o SSP5-8.5
used in hei s udy) and he ac ha changes in El Niño anomalies
modula ed by clima e change a e only expec ed o eme ge a he end o he
cen u y. Also, as i is he case o he eleconnec ion owa ds no h Ame ica,
he ac ha ampli ude o he egional esponse o El Niño is no linea ly
ela ed o he s eng h o he e en makes i ha d o compa e wi h s udies
based on El Niño composi es o all s eng hs as in Johnson e al.49 and
McG ego e al.38.
In e mso means a echanges,Fig.3c shows a deepening o he
Aleu ian low by 20 m in he u u e clima ology, compa ed o p esen -day
condi ions. In he Sou h Pacific Ocean, he e is a s eng hening o he zonal
low p essu e anomalies ac oss SESA, and posi i e z500 anomalies in he
Sou he n Ocean o he coas o Sou h Ame ica. I is impo an o no e ha
he ampli ude o u u e changes in El Niño-d i en z500 anomalies (Fig. 3b)
is wice as la ge as he mean s a e changes (Fig. 3c), highligh ing he sig-
nificance o changes in ENSO eleconnec ions in he u u e ela i e o
changes in he backg ound flow. The absolu e changes in z500 and zonal
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 4
wind du ing s ong El Niño e en s in he u u e a e domina ed by changes in
El Niño eleconnec ions a he han by mean s a e changes, as sugges ed by
Be e ley e al.51.
Regional ex eme empe a u e esponses o s ong u u e El
Niño e en s
Building on he e idence in he p e ious sec ions showing egional sig-
na u es o he modula ion o s ong El Niño eleconnec ions unde u u e
clima e change, we now in es iga e egional changes in me eo ological
ex emes unde s ong El Niño, in No h Ame ica, sou he n Sou h Ame ica
and sou he n A ica. Special a en ion is paid o shi s in he dis ibu ion o
daily maximum empe a u e om he p esen o he u u e conside ing
bo h cool and wa m days (see Me hods).
We fi s conside cool days in No h Ame ica using a cool day index
(see Me hods) o unde s and changes o he dis ibu ion o cold anomalies
d i en by El Niño in he u u e. No e his index is pe cen ile-based and
he e o e e ec i ely emo es he clima ological shi in cool days d i en by
an h opogenic o cing.
In he p esen day, s ong El Niño d i es an inc ease o cool days in
No h Ame ica o up o 7% pe win e , especially along he Rocky Moun ains
and cen al-eas e n US (Fig. 4a), and a 6% educ ion in cool days pe win e
o e eas e n Canada. The e is a ela i einc ease in cool days pe win e du ing
s ong El Niño e en s o e mos o he USA in he u u e compa ed o p esen
day (Fig. 4b), meaning an in ensifica ion o cold anomalies in he eas and he
appea ance o cool anomalies o e he no he n US and Alaska. The mean
clima e change signal shi s he local clima ological 10 h pe cen ile o max-
imumdaily empe a u e owa dshighe aluesin hewholeNo hAme ica
(See Fig. S3c), especially o e he No heas o he Uni ed S a es, Canada and
Alaska, whe e he h eshold o cool days inc eases by up o 8 °C. The changes
in ela i e occu ence o cold days du ing s ong El Niño could ha e socie al
implica ions, since a ocus on adap a ion o a wa me clima e may educe he
ocusonp epa edness o colddays. A he same ime, pe cen ile-based
indica o s migh no be as good a p oxy o clima e impac s on heal h in he
u u e as hey a e in he p esen , i.e. cold days in he u u e will be less ha m ul
o human heal h. This di e ence o e ides he in ensifica ion o he cold El
Niño anomalies expec ed a he end o he cen u y, since he o e all em-
pe a u e will be highe due o he dominance o he global wa ming signal o
he o al empe a u e anomaly.
No he n Sou h Ame ica is one o he egions whe e s ong El
Niño e en s ha e he la ges impac s on daily empe a u e ex emes.
F om Decembe o Feb ua y, s ong El Niño leads o an inc ease o
wa m days o up o 40%, especially a ound he Amazon basin in B azil,
Venezuela and he Pacific coas o Colombia and Ecuado . In he
u u e, he s iking impac o s ong El Niño e en s in he egion ge s
mu ed in he Amazon basin, whe e he pe cen age o wa m days
dec eases up o 15%. In sou hwes B azil and Boli ia, howe e , he e is
asignifican s eng hening o he El Niño signal, wi h up o 15% mo e
wa m days om Decembe o Feb ua y. Changes in ex eme em-
pe a u e impac s d i en by El Niño in he u u e will play a key ole in
he equency and in ensi y o o es fi esin heAmazon egion, ueled
by a nega i e p ecipi a ion and soil mois u e end52.
Acco ding o Fig. 6, du ing s ong El Niño, a eas o A ica sou h o he
equa o expe ience a s ong inc ease in wa m days in summe , wi h he
la ges inc eases o e Equa o ial Guinea, Republic o he Congo and Angola,
wi h up o 25% addi ional wa m days pe summe . In he u u e expe imen ,
he e is a ela i ely smalle inc ease in wa m days pe summe due o s ong
El Niño (app oxima ely 7% less in some a eas). In addi ion, he e a e
enhanced wa m anomalies owa ds he no he n egions o lakes Tanga-
nyika and Malawi, Tanzania and he Democ a ic Republic o he Congo,
whe e he numbe o anomalous wa m days pe summe du ing s ong El
Niño is ound o inc ease by up o 10%.
Figu e S3g-i shows he shi o he 90 h Tmax pe cen ile owa ds highe
alues in sou he n A ica, especially o e he wes and sou hwes o he
egion, whe e he empe a u e h eshold o a day o be conside ed wa m
will ise om 30 °C o 35 °C. The smalle inc ease in wa m days in he u u e
ela i e o p esen -day condi ions is caused by a aise o he Tmax h eshold
o wa m days. Ne e heless, in he u u e s ong El Niño s ill leads o an
inc ease in wa m days in sou he n A ica, u he exace ba ing he unde -
lying signal o clima e change consis ing o a shi owa ds highe em-
pe a u es in he egion in aus al summe . In he coas al a eas o
Mozambique and Tanzania he e is a significan inc ease o up o 10% in he
numbe o El Niño-d i en wa m days in summe , which cons uc i ely
in e e e wi h clima e change leading o an exace ba ion o wa m summe
days in he egion. As shown in Figu e S3i, he la ges shi in he clima o-
logical wa m day h eshold is ound in Namibia, Bo swana and
Mozambique.
Fig. 3 | Fu u e changes in mid- oposphe ic ci cula ion due o s ong El Niño and
clima e change. Eddy geopo en ial heigh a 500 hPa, (de ia ions om he zonal
mean, m, filled con ou s) and zonal wind a 200 hPa anomalies (m s−1, black con-
ou s, whe e solid lines deno e posi i e anomalies and dashed lines show nega i e
anomalies) du ing El Niño in he p esen (a), di e ences in u u e and p esen El
Niño anomalies (b), mean s a e changes (c) and changes in absolu e alues be ween
u u e and p esen (d). Ha ched a eas a e no s a is ically significan a he 95%
confidence le el.
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 5
Discussion
This s udy has used idealised pacemake simula ions wi h he EC-Ea h3-
CC model ollowing he SSP2-4.5 scena io o quan i y changes in he global
impac s o s ong El Niño e en s in a wa me clima e. This app oach isola es
he ole o changing backg ound clima e condi ions on he a mosphe ic
esponse o El Niño, which has been a limi a ion o some p e ious s udies
(e.g., McG ego e al.38).
The expe imen s show p ecipi a ion anomalies du ing s ong El
Niño inc ease in he eas e n equa o ial Pacific and wa m pool unde
u u e clima e change, and dec ease o e he cen al Pacific(Fig.1d).
The changes in he s ong El Niño p ecipi a ion esponse in he
eas e n Pacific a e compa able o he changes om mean clima e
change. Alongside he impac o clima e change on he local Pacific
esponse o s ong El Niño, he la ge-scale global a mosphe ic ele-
connec ions a e also modified. The Rossby wa e ains emana ing
om he wes e n opical Pacific in o he no he n and sou he n
hemisphe es show di e ences in ampli ude and phase ha modi y
egional clima e haza ds associa ed wi h s ong El Niño.
The change in equency o cool days pe win e unde s ong El Niño
is la ge in he u u eo e eas e n No h Ame ica. I is impo an o no e ha
due o he o e all backg ound wa ming om clima e change, he pe cen ile-
based h eshold adop ed o cool days inc eases om −5 °C o 0 °C be ween
p esen and u u e meaning he absolu e empe a u es a e less se e e bu s ill
below eezing. The numbe o cool days pe win e sligh ly dec eases in he
u u e clima ology. Owing o he al e ed eleconnec ion associa ed wi h
he PNA pa e n, he e is a la ge equency o cool days d i en by El Niño in
he u u e (17.1%) han in he p esen (9.5%), as sugges ed by Liebe e al. 5.
In ag eemen wi h Meehl e al.53 we find an eas wa d shi o cold ex emes in
u u e El Niño e en s and a dec ease in cool days o e he wes coas o he
USA. This shi migh be linked o he s eng hening o he eleconnec ion o
El Niño o he No h A lan ic sec o 54.
Finally, we show ha despi e mos o he egions wi h significan
changes in s ong El Niño impac s, a limi ed numbe o a eas, like no he n
Sou h Ame ica and sou he n A ica, expe ience a weakening o he signal
d i en by El Niño unde u u e condi ions. In he Amazon basin, a egion
whe e wa m days oughly double in equency when he e is a s ong El
Niño e en , ou expe imen s sugges s a weakening o his eleconnec ion
ansla ed in o a dec ease in wa m days o e mos o he egion. Figu e 5
shows a weakening o he El Niño-d i en wa m days in mos o he egion
excep sou heas B azil and Boli ia, which migh be he esul o a sa u a ion
o he impac s o El Niño unde e en s o high ampli ude such as he ones
conside ed in his s udy. The weakening o El Niño-d i en empe a u e
ex emes in no he n Sou h Ame ica in he u u e coincides wi h he
expec ed changes in mean empe a u e (Fig. 2c).
Figu e 6sugges s ha El Niño-d i en wa m days in summe will be
lowe in he u u e, especially o e sou he n Namibia, Bo swana and cen al
Sou h A ica. While he h eshold o compu e wa m days as simula ed by
EC-Ea h3-CC inc eases om 28.8 °C in he p esen o 31.2 °C in he u u e,
in ag eemen wi h u u e CORDEX-A ica clima e p ojec ions o e he
egion55, he El Niño-d i en wa m days do no inc ease in equency in a
wa me clima e. In absolu e e ms, he backg ound wa ming induced by
clima e change will inc ease he equency o hea wa es o e Sou he n
A ica, independen o whe he he e is a s ong El Niño e en happening
o no .
In e es ingly, he p ojec ed changes in El Niño eleconnec ions
esemble hose iden ified in p e ious s udies, sugges ing ha simula ed
changes in u u e impac s o El Niño in coupled clima e models a e
domina ed by changes in he mean s a e, a he han by changes in ENSO
SSTs. In ag eemen wi h McG ego e al.38,wefind a s eng hening and
sou hwa d shi o he No h A lan ic je s eam in bo eal win e du ing El
Niño e en s in he u u e, highligh ing he ein o cemen o he El Niño-
nega i e NAO ela ionship. This is also in ag eemen wi h p e ious li e a-
u e (i.e. D oua d and Cassou40, Johnson e al.49), al hough his esponse is
no cap u ed in CMIP5 models37. Ou esul s also align wi h he o e all
CMIP5 and CMIP6 ends in he no hwa d shi o p ecipi a ion anomalies
no h o SESA and he d ying end o he sou h o B azil, al hough ou
pacemake e eals a weakening o he d y anomalies d i en by s ong El
Niño e en s o e heAmazon basin. Fu u e wo k will assess whe he he e is
any sa u a ion o d y anomalies o e his a ea wi h El Niño e en s o la ge
ampli ude such as he ones we a e ocusing on.
In e ms o u u e changes in su ace empe a u e anomalies d i en by
s ong El Niño, we find simila i ies be ween ou esul s and he o e all end
encompassing El Niño e en s o any ampli ude. Some examples a e a ho -
ge s-ho e end o e wes e n B azil, and No he n A ica, and while
CMIP5 p ojec ions do no show a significan change o e Aus alia37,ou
esul s ag ee wi h o e all CMIP6 p ojec ions38 and la ge ensemble s udies36.
Some addi ional ea u es a ise when ocusing on s ong El Niño e en s
ha a e no seen in he o e all p ojec ions o ENSO impac s based on
composi e analysis, ha indica e ei he a nonlinea clima e esponse o
s ong El Niño e en s o a disag eemen o he EC-Ea h3-CC model wi h
he CMIP6 mul i-mean esul s (i.e. McG ego e al.38). Ou esul s e eal
po en ial nonlinea i ies in he u u e clima e esponse o El Niño e en s (e.g.
he Amazon basin), whe e inc easing SST anomalies in he Niño3.4 egion
do no u he ampli y he global su ace impac s associa ed wi h a canonical
Fig. 4 | El Niño-d i en cool days in No h Ame ica. Anomalous cool days pe
win e (DJF) du ing s ong El Niño in No h Ame ica in (a) p esen condi ions and
(b) u u e - p esen condi ions. Only g id cells co esponding o land su ace a eas
a e used in his plo . A eas whe e s ong El Niño e en s do no d i e s a is ically
significan anomalies in he p esen -day simula ions appea as ha ched in subpanel
(a). Ha ching in subpanel (b) show non significan changes in he impac s o s ong
El Niño e en s be ween he u u e and p esen -day simula ions.
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 6
El Niño e en . This nonlinea i ies migh be accompanied by shi s in ci -
cula ion induced by clima e change, wi h Sou h Ame ica being he mos
illus a i e example. Acco ding o ou esul s, d y anomalies o e he
Amazonbasind i enby e ys ongElNiñowillge mu edin he u u e,
and we and cool anomalies o e he SESA a e expec ed o shi no hwa ds
in DJF. Fu he wo k is needed o assess he na u e o hese shi s and
nonlinea beha iou s.
We find an in ensifica ion o cold anomalies d i en by s ong El Niño
o e no heas No h Ame ica and a significan weakening o wa m
anomalies o e no he n Sou h Ame ica and sou he n A ica. By compa ing
changes in he eleconnec ions o s ong El Niño e en s wi h mean s a e
changes due o clima e change, we emphasise ha al hough s ong El Niño
migh d i e cool anomalies in he u u e in some egions, he unde lying
signa u e o clima e change la gely domina es he global sign o empe a u e
anomalies in he u u e, masking po en ial cooling e ec s o na u al clima e
a iabili y. The ac ha changes in wa m days d i en by s ong El Niño in
he u u eo e no he nSou hAme icaandSou he nA icaa eno e y
p onounced is caused by he use o ully mo ing h esholds (based on
clima ological alues o each pe iod) o compu e hese ex eme hea indices.
In he examples shown in his s udy, his indica es ha he modynamic
changes modula ed by clima e change a e he dominan sou ce o he shi in
he dis ibu ion o maximum daily empe a u e alues, in ag eemen wi h
Vogel e al.56. Fu u e wo k is needed o add ess po en ial nonlinea i ies in he
impac sd i enby e ys ongElNiñoe en sin he u u e.
As seen in mos CMIP6 models, EC-Ea h3-CC shows a long- e m
u u e end esembling an El Niño-like wa ming in he opical eas Pacific,
dec easing he zonal SST g adien by almos 20% owa ds he end o he 21s
cen u y. This ea u e is consis en wi h a weakening Fo he Walke ci cu-
la ion and is simila o he end p ojec ed in mos CMIP557 and CMIP6
models27, bu does no ma ch obse ed he Walke ci cula ion s eng hening
and SST ends o e ecen decades58,59.
Human-caused clima e change is wa ming he plane a unp ece-
den ed a es. Na u al clima e a iabili y can ei he ampli y o mu e he
signa u e o clima e change in he u u e. S ong El Niño e en s a e p ojec ed
o inc ease in equency20, and gi en he se e i y and each o hei impac s,
i is c ucial o unde s and how he impac s o hese ex eme e en s will
change in he u u e o acili a e adap a ion and mi iga ion policies in he
a eas a ec ed. Ou s udy sugges s ha he u u e clima e condi ions du ing
s ong El Niño e en s will be mo e impac ed by he unde lying signa u e o
clima e change han by changes in ENSO i sel . Fu u e s udies a e needed o
imp o e ou unde s anding o he mechanisms o such amplifica ion and
weakening o he u u e impac s o s ong El Niño e en s. Addi ionally, a
Fig. 6 | As in Fig. 5, bu o Sou he n A ica.
Fig. 5 | El Niño-d i en wa m days in No he n Sou h Ame ica. Anomalous wa m days pe summe (DJF) du ing s ong El Niño in No he n Sou h Ame ica in (a) p esen
condi ions and (b) he di e ence be ween u u e and p esen condi ions. The ha ching con en ion is he same as in Fig. 4.
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 7
mul i-model app oach would enable ou esul s o be e ified o a la ge se
o p ojec ions, as well as an explo a ion o o he seasons beyond he cano-
nical bo eal win e season.
Me hods
Pacemake simula ions a e pe o med wi h he EC-Ea h3-CC model60 o
unde s and changes in he impac s o s ong El Niño e en s be ween p esen
and u u e clima es.
Model
EC-Ea h3-CC is an Ea h Sys em model wi h an app oxima e a mosphe e
and ocean esolu ion o 1°. The a mosphe ic model is he Eu opean Cen e
o Medium- ange Wea he Fo ecas s (ECMWF) In eg a ed Fo ecas ing
Sys em (IFS) e sion 36 4, which is based on he ECMWF seasonal p e-
dic ion sys em 4, and has a esol ed s a osphe e. NEMO3.6 and LIM3 a e
he ocean and sea ice componen s; biogeochemical p ocesses in he ocean
and o e land a e simula ed by he PISCES and LPJ-GUESS models,
espec i ely.
The EC-Ea h3-CC model gene ally ep oduces obse ed ENSO el-
econnec ions. Figu e S1 compa es he global ENSO p ecipi a ion and su -
ace empe a u e eleconnec ions in he Decembe -Feb ua y (DJF) season
in EC-Ea h3-CC wi h obse a ions om he GPCP 2.361 and Be keley
Ea h62 da ase s. Fo he model-da a compa ison we use ou pu om a ime-
dependen opical pacemake simula ion wi h EC-Ea h3-CC co e ing he
pe iod 1900-2018, whe e his o ical SSTs a e o ced wi h obse ed SST
anomalies in he opical Pacific ocean. This ensu es he ENSO e en s
simula ed in he model b oadly ollow hose obse ed du ing he his o ical
pe iod. The model cap u es he sign and ampli ude o ENSO- ela ed p e-
cipi a ion anomalies o e he opical Pacific, bu ends o o e es ima e he
p ecipi a ion dipole esponse o e he Indian Ocean (See Fig. S1). O e he
equa o ial A lan ic, he model displays a weak zonal p ecipi a ion dipole
consis ing o d y condi ions o e he Amazon and wes e n equa o ial
A lan ic and we anomalies o e he eas e n A lan ic and he Gul o Guinea,
whe eas obse a ions show an o e all nega i e p ecipi a ion anomaly ha
peaks o he eas B azilian coas . Rega dless o hese di e ences, he o e all
pa e ns o global esponse in empe a u e and p ecipi a ion a e well cap-
u ed by EC-Ea h3-CC.
Expe imen al p o ocol
Six expe imen s a e pe o med in which he model SSTs in he equa o ial
eas e n Pacific a e es o ed owa ds he obse ed SST anomalies om he
1982-1983, 1997-98 and 2015-16 s ong El Niño e en s supe imposed on
ei he he p esen day o u u e model clima ological SST. Obse ed
mon hly SST anomalies a e ex ac ed om he ERSST 5 da ase 9calcula ed
ela i e o he pe iod 1981-2010. P esen day clima ological alues a e
es ima ed om he his o ical simula ions o EC-Ea h3-CC using he pe iod
2005-2014. Fo he u u e, we use he SSP2-4.5 scena io, a middle-o - he-
oad g eenhouse gas emissions pa hway whe e global mean su ace em-
pe a u e in EC-Ea h3-CC inc eases by 2.8 K in 2100 ela i e o p e-
indus ial imes. This SSP is consis en wi h he clima e ha would be
achie ed by he end o his cen u y i coun ies deli e on hei cu en
clima e a ge s. Acknowledging ha a s onge SSP scena io would p o ide
a la ge clima e change signal, we pe o med la ge ensemble expe imen s
wi h each El Niño e en being simula ed 30 imes in bo h p esen and u u e
clima es, ensu ing ha changes in El Niño in he u u e a e dis inguishable
om he clima e change signal. The pe iod 2085-2094 is used o define he
u u e clima ology. Fo each 10-yea clima ological pe iod, we a e age o e
10 ini ial condi ion ensemble membe s, meaning ha he clima ological
s a es a e calcula ed o e 100 yea s.
The obse ed ERSST 5 SST anomalies o he 3 s ong El Niño cases
a e in e pola ed o he model’s ocean g id (ORCA1) and added o he
p esen o u u e model clima ology o p oduce he a ge es o ing fields.
Fo each expe imen , a 30 membe ensemble is gene a ed by using 3 se s o
ini ial condi ions om each o he 10 his o ical o SSP2-4.5 simula ions a 5
yea in e als ( o example o p esen -day we use 2005, 2010 and 2014).
The SST a ge pa e n is imposed h ough su ace flux es o ing wi h a
es o ing coe ficien s ong enough o achie e a mosphe e-only condi ions
in he es o ing domain, which spans ac oss he opical Pacific omla i-
udes is 10°S - 10°N and longi udes om 130°E - 70°W, wi h a 10° bu e
zone on each side excep he eas e n ma gin which coincides wi h he
Ame ican con inen . All a iables ou side o he es o ing egion e ol e
eely in a coupled configu a ion. Each membe is in eg a ed o 2 yea s
s a ing om Janua y 1s and he SST es o ing is ac i a ed om he 1s o
Janua y o he fi s yea un il he end o Decembe o he second yea , hus
ha ing a 2-yea simula ion o each ensemble membe whe e SST anomalies
a e es o ed mon hly o ma ch he a ge obse ed pa e n. The esul is a
sui e o 30-membe ensemble simula ions o 3 s ong El Niño e en s in
p esen and u u e clima es. The analysis showed e y simila changes in
impac s o he 3 El Niño cases, so he analysis ocuses on he “mul i-e en
mean”, he ea e MEM.
We ocuson heDJFseasonwhen heElNiñoe en speakand
compa e he anomalous esponse o s ong El Niño in he u u e wi h ha o
he p esen . Anomalies om each pe iod a e calcula ed ela i e o hei
co esponding clima ological alues.
Clima e signal decomposi ion
To di e en ia e be ween changes o El Niño i sel and changes in El Niño
impac s added on op o he global wa ming signal, he figu es om his
pape ollow a simila s uc u e. We fi s in oduce he image showing
p esen -day anomalies du ing e y s ong El Niño e en s in DJF (Eq. 1),
consis ing o sub ac ing he p esen -day clima ology o heabsolu eclima e
impac s du ing El Niño:
ElNi~
noimpac s ¼ElNi~
nopMeanclima epð1Þ
We hen explo e changes in El Niño impac s in he u u e and p esen ,
emo ing he mean clima e s a e and hus accoun ing jus o changes o El
Niño eleconnec ions ega dless o he backg ound clima e change signal
(Eq. 2):
Changes 2ElNi~
noimpac s ¼ðElNi~
no Meanclima e Þ
ðElNi~
nopMeanclima epÞð2Þ
Changes in he mean clima e s a e a e also quan ified and displayed in
he plo s (Eq. 3):
Means a echanges ¼Means a e Means a epð3Þ
Adding up Eqs. 2and 3we ob ain Eq. 4, which illus a e he changes in
absolu e impac s o El Niño in he u u e, conside ing bo h changes in El
Niño eleconnec ions as well as changes in he mean s a e.
To alchanges 2ElNi~
noimpac s ¼ElNi~
no ElNi~
nopð4Þ
Clima e indices
El Niño is known o d i e wea he ex emes o e No h Ame ica15,53,
no he n Sou h Ame ica and Sou he n A ica2, among o he egions, so
hese a e used as h ee egional case s udies o examine he influence o
u u e clima e change on El Niño impac s.
In No h Ame ica, we quan i y he change in cool days pe win e ,
which is compu ed as he numbe o days when maximum daily su ace
empe a u es (Tmax) du ing he DJF season lay below he 10 h pe cen ile
based on he model clima ology63. Secondly we explo e changes in he
amoun o wa m days pe summe , which is again compu ed using max-
imum daily empe a u e and calcula ed he pe cen age change in days when
maximum empe a u e exceeds he 90 h pe cen ile. We ocus on no he n
Sou h Ame ica and Sou he n A ica o explo e changes in he equency o
wa m days pe summe .
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 8
No e ha he h esholds o compu e cool and wa m days in he
p esen and u u e a e based on he p esen (1995-2014) and u u e
(2085-2094) clima ologies, espec i ely, using ully mo ing
h esholds51. The e o e, he indices cap u e simila ela i e anomalies
in he wope iods.Whilewedono link heme eo ologicalcondi ions
o clima e impac s (e.g. on human heal h), his choice is mos ele an
o a si ua ion in which socie y unde akes some clima e adap a ion o
accommoda e mean clima e change. An al e na i e app oach would be
o use he p esen day h eshold in bo h pe iods, which would be mo e
ele an o a si ua ion in which clima e adap a ion is no unde aken.
To assess whe he changes a e s a is ically significan , we pe o med a
wo sided S uden ’s - es .
Code a ailabili y
Thecodeused ocalcula e heindica o sandp epa e hefigu es will be
a ailable upon eques . The expe imen al p o ocol can be accessed in his
Gi lab eposi o y: h ps://ea h.bsc.es/gi lab/p ascas/el-nino-pacemake -
expe imen al-design.
Recei ed: 21 No embe 2024; Accep ed: 12 Ma ch 2025;
Re e ences
1. Tasche o, A. S. e al. ENSO a mosphe ic eleconnec ions. El Niño
Sou he n Oscilla ion in a Changing Clima e, 309–335 (2020).
2. A blas e , J. M. & Alexande , L. V. The impac o he El Niño-Sou he n
Oscilla ion on maximum empe a u e ex emes. Geophys. Res. Le .,
39, (2012).
3. Cai, W. e al. Clima e impac s o he El Niño–Sou he n Oscilla ion on
Sou h Ame ica. Na . Re . Ea h En i on. 1, 215–231 (2020).
4. Ra nam, J. V., Behe a, S. K., Masumo o, Y. & Yamaga a, T. Remo e
e ec s o El Niño and Modoki e en s on he aus al summe
p ecipi a ion o sou he n A ica. J. Clim. 27, 3802–3815 (2014).
5. Liebe , R., B own, J., King, A. & F eund, M. His o ical and Fu u e
Asymme y o ENSO Teleconnec ions wi h Ex emes. J. Clim. 37,
5909–5924 (2024).
6. Mooley, D. A. & Pa hasa a hy, B. Indian summe monsoon and El
Niño. Pu e Appl. Geophysics 121, 339–352 (1983).
7. Seage , R. e al. Mechanisms o ENSO- o cing o hemisphe ically
symme ic p ecipi a ion a iabili y. Q. J. R. Me eo ological Soc.: A J.
A mos. Sci., Appl. Me eo ol. Phys. Oceanog . 131, 1501–1527 (2005).
8. Godda d, L. & G aham, N. E. Impo ance o he Indian Ocean o
simula ing ain all anomalies o e eas e n and sou he n A ica. J.
Geophys. Res.: A mosphe es 104, 19099–19116 (1999).
9. Huang, B. e al. Ex ended econs uc ed sea su ace empe a u e,
e sion 5 (ERSST 5): upg ades, alida ions, and in e compa isons. J.
Clim. 30, 8179–8205 (2017).
10. Funk, C. e al. 18. An h opogenic enhancemen o mode a e- o-s ong El
Niño e en s likely con ibu ed o d ough and poo ha es s in sou he n
A ica du ing 2016. Bull. Am. Me eo ol. Soc. 99,S91–S96 (2018).
11. Panisse , J. S. e al. Con as ing pa e ns o he ex eme d ough
episodeso 2005, 2010and2015 in he AmazonBasin.In . J. Clima ol.
38, 1096–1104 (2018).
12. Duke, N. C. e al. La ge-scale dieback o mang o es in Aus alia’s Gul
o Ca pen a ia: a se e e ecosys em esponse, coinciden al wi h an
unusually ex eme wea he e en . Ma . F eshw. Res. 68, 1816–1829
(2017).
13. Page, S. E. e al. The amoun o ca bon eleased om pea and o es
fi es in Indonesia du ing 1997. Na u e 420,61–65 (2002).
14. Pa a,P. K.e al.TheO bi ing Ca bonObse a o y(OCO-2) acks2–3
pe a-g am inc ease in ca bon elease o he a mosphe e du ing he
2014–2016 El Niño. Sci. Rep. 7, 13567 (2017).
15. Glan z, M. H. Cu en s o change: impac s o El Niño and La Niña on
clima e and socie y. (Camb idge Uni e si y P ess, 2001).
16. Le Qué é, C. e al. Global ca bon budge 2018. Ea h Sys . Sci. Da a
Discuss. 2018,1–3 (2018).
17. Wang, J. e al. Con as ing e es ial ca bon cycle esponses o he
1997/98 and 2015/16 ex eme El Niño e en s. Ea h Sys . Dyn. 9,1–14
(2018).
18. An illa-Hughes, J. K., Jina, A. S. & McCo d, G. C. ENSO impac s child
unde nu i ion in he global opics. Na . Commun. 12, 5785 (2021).
19. San oso, A. e al. La e- wen ie h-cen u y eme gence o he El Niño
p opaga ion asymme y and u u e p ojec ions. Na u e 504, 126–130
(2013).
20. Cai, W. e al. Inc easing equency o ex eme El Niño e en s due o
g eenhouse wa ming. Na . Clim. Change 4, 111–116 (2014).
21. Lengaigne,M.&Vecchi,G.A.Con as ing he e mina iono mode a e
and ex eme El Niño e en s in coupled gene al ci cula ion models.
Clim. Dyn. 35, 299–313 (2010).
22. Chung, C. T., Powe , S. B., A blas e , J. M., Rashid, H. A. & Ro , G. L.
Nonlinea p ecipi a ion esponse o El Niño and global wa ming in he
Indo-Pacific. Clim. Dyn. 42, 1837–1856 (2014).
23. Tasche o, A. S. & England, M. H. El Niño modoki impac s on
Aus alian ain all. J. Clim. 22, 3167–3174 (2009).
24. Beniche, M. e al. A dis inc and ep oducible eleconnec ion pa e n
o e No h Ame ica du ing ex eme El Niño e en s. Sci. Rep. 14, 2457
(2024).
25. Salas, H. D. e al. P ecipi a ion o e no he n Sou h Ame ica and he
a -eas e n Pacific du ing ENSO: Phase synch oniza ion a in e -
annual ime scales. In . J. Clima ol. 44, 2106–2123 (2024).
26. T ascasa-Cas o, P., Maycock, A. C., Yiu, Y. Y. S. & Fle che , J. K. On
he linea i y o he s a osphe ic and Eu o-A lan ic sec o esponse o
ENSO. J. Clim. 32, 6607–6626 (2019).
27. Mahe , N. e al. The u u e o he El Niño–Sou he n Oscilla ion: Using
la geensembles oillumina e ime- a ying esponsesandin e -model
di e ences. Ea h Sys . Dyn. 14, 413–431 (2023).
28. F ed iksen, H. B., Be ne , J., Sub amanian, A. C. & Capo ondi, A.
How does El Niño–Sou he n Oscilla ion change unde global
wa ming—Afi s look a CMIP6. Geophys. Res. Le . 47,
e2020GL090640 (2020).
29. Cai, W. e al. Inc eased ENSO sea su ace empe a u e a iabili y
unde ou IPCC emission scena ios. Na . Clim. Change 12, 228–231
(2022).
30. Heede, U. K. & Fedo o , A. V. Colde eas e n equa o ial Pacific and
s onge Walke ci cula ion in he ea ly 21s cen u y: sepa a ing he
o ced esponse o global wa ming om na u al a iabili y. Geophys.
Res. Le . 50, e2022GL101020 (2023).
31. Seage , R., Hende son, N. & Cane, M. Pe sis en disc epancies
be ween obse ed and modeled ends in he opical Pacific Ocean.
J. Clim. 35, 4571–4584 (2022).
32. Lim, E. P. e al. Con inua ion o opical Pacific Ocean empe a u e
end may weaken ex eme El Niño and i s linkage o he Sou he n
Annula Mode. Sci. Rep. 9, 17044 (2019).
33. Lee, J.-Y. e al. Fu u e Global Clima e: Scena io-Based P ojec ions
and Nea -Te m In o ma ion. In Clima e Change 2021: The Physical
Science Basis. Con ibu ion o Wo king G oup I o he Six h
Assessmen Repo o he In e go e nmen al Panel on Clima e
Change [Masson-Delmo e, V., P. Zhai, A. Pi ani, S.L. Conno s, C.
Péan, S. Be ge , N. Caud,Y. Chen, L. Gold a b, M.I.Gomis, M. Huang,
K. Lei zell, E. Lonnoy, J.B.R. Ma hews, T.K. Maycock, T. Wa e field,
O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Camb idge Uni e si y P ess,
Camb idge, Uni ed Kingdom and New Yo k, NY, USA, pp. 553–672,
h ps://doi.o g/10.1017/9781009157896.006 (2021).
34. Hu, K., Huang, G., Huang, P., Kosaka, Y. & Xie, S. P. In ensifica ion o
El Niño-induced a mosphe ic anomalies unde g eenhouse wa ming.
Na . Geosci. 14, 377–382 (2021).
35. Ying, J. e al.Eme gence o clima e changein he opical Pacific. Na .
Clim. Change 12, 356–364 (2022).
h ps://doi.o g/10.1038/s41612-025-01003-1 A icle
npj Clima e and A mosphe ic Science | (2025) 8:116 9
