Wes e n Medi e anean d ough s os e ed by
A c ic sea-ice loss
Rami o I. Sau al1,2,3,4, F ancisco J. Doblas-Reyes1,5, James A. Sc een6, Jenni e L. Ca o6,
S ephanie Hay6, and Hao Yu6
1 Ba celona Supe compu ing Cen e (BSC), Ba celona, Spain
2 CONICET-Uni e sidad de Buenos Ai es. Cen o de In es igaciones del Ma y la A mós e a (CIMA),
Buenos Ai es, A gen ina
3 Uni e sidad de Buenos Ai es, Facul ad de Ciencias Exac as y Na u ales, Depa amen o de Ciencias
de la A mós e a y los Océanos, Buenos Ai es, A gen ina
4 CNRS-IRD-CONICET-UBA. Ins i u o F anco-A gen ino pa a el Es udio del Clima y sus Impac os
(IRL 3351 IFAECI), Buenos Ai es, A gen ina
5 Ins i ució Ca alana de Rece ca i Es udis A ança s (ICREA), Ba celona, Spain
6 Depa men o Ma hema ics and S a is ics, Uni e si y o Exe e , Exe e , Uni ed Kingdom
Co esponding au ho : Rami o I. Sau al ( ami o.sau a[email p o ec ed])
Manusc ip (non-LaTeX)
1
Ea ly Online Release: This p elimina y e sion has been accep ed o publica ion in Jou nal o Clima e,
may be ully ci ed, and has been assigned DOI 10.1175/JCLI-D-25-0066.1. The inal ypese copyedi ed
a icle will eplace he EOR a he abo e DOI when i is published.
© 2025 Ame ican Me eo ological Socie y. This is an Au ho Accep ed Manusc ip dis ibu ed unde he e ms o
he de aul AMS euse license. Fo in o ma ion ega ding euse and gene al copy igh in o ma ion, consul he
AMS Copy igh Policy (www.ame soc.o g/PUBSReuseLicenses).
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
“This is he accep ed e sion o he ollowing a icle: Sau al, R. I., F. J. Doblas-Reyes, J. A. Sc een, J.
L. Ca o, S. Hay, and H. Yu, 2025: Wes e n Medi e anean d ough s os e ed by A c ic sea-ice loss. J.
Clima e, h ps://doi.o g/10.1175/JCLI-D-25-0066.1, in p ess
CC BY 4.0 license
Addendum o Resea ch Funded by cOAli ion S O ganiza ions
2
ABSTRACT
Cu -o lows (COLs), de ined as isola ed mid- oposphe ic low p essu e sys ems, a e
esponsible o a la ge ac ion o he annual mean and ex eme p ecipi a ion o e he
Medi e anean Sea egion. In his s udy we quan i y he impac s om A c ic sea-ice loss on
he equency and dis ibu ion o COLs. We use model ou pu s om he Pola Ampli ica ion
Model In e compa ison P ojec (PAMIP) o ced only by he p ojec ed educ ion in he
amoun o A c ic sea ice in a globe 2ºC wa me han in he p eindus ial pe iod. We ind ha
sea-ice loss can, h ough al e a ions o he uppe -le el je al eady documen ed in p e ious
s udies, signi ican ly a ec he equency o COLs o e sou he n Eu ope: in pa icula , a
sha p educ ion is simula ed o e he no heas A lan ic and he Ibe ian Peninsula as a
consequence o mo e an icyclonic condi ions p e ailing o e ha egion. This educ ion in
he numbe o COLs is accompanied by signi ican ly less p ecipi a ion o e he wes e n
Medi e anean, which could po en ially lead o wa e a ailabili y a ec a ion he e.
Keywo ds: Sea ice; A c ic Ampli ica ion; cu -o lows; ex eme p ecipi a ion; d ough s
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
3
1. In oduc ion
A c ic sea-ice loss and i s associa ed impac s on he a mosphe ic ci cula ion ha e been a opic
o ac i e esea ch in ecen yea s (e.g. Ge ais e al., 2024; No z and S oe e, 2016; Sc een and
Simmonds, 2010; Sc een e al., 2018; Smi h e al., 2022; Ye e al., 2024; Zappa e al., 2018). I
is well es ablished ha he pola egions o he No he n Hemisphe e ha e wa med
subs an ially mo e han he global a e age du ing he pas decades (Chemke e al., 2021;
P e idi e al., 2021; Ran anen e al., 2022; Se eze e al., 2009; Wang e al., 2016) pa ly due
o sea-ice loss (Goosse e al., 2018; Sc een and Simmonds, 2010) and changes in he egional
ci cula ion (Zhang e al., 2023). The educ ion in sea-ice co e and he associa ed wa ming o
he su ace ocean ha e igge ed posi i e eedback mechanisms which ha e ac ed o in ensi y
he egional lowe - oposphe ic wa ming (e.g., Boeke e al., 2021; Dai and Jenkins, 2023; He
e al., 2019), e en hough i has been ecen ly shown ha his excessi e wa ming could occu
e en wi hou conside ing sea ice (Russo o and Biasu i, 2020; England and Feldl, 2024).
P e ious s udies ha e concluded ha u u e sea-ice loss o e he A c ic could lead o a numbe
o impac s on he clima e o he midla i udes. These include a po en ial shi in he loca ion o
he oposphe ic uppe -le el je (Ba nes and Sc een, 2015; P e idi e al., 2021; Sc een and
Blackpo , 2019; Sc een e al., 2022; Ye e al., 2023), changes in he occu ence o ex eme
wea he e en s (Cohen e al., 2014) and al e a ions o daily wea he pa e ns (Ge ais e al.,
2024), among o he s. Fu u e sea-ice mel is also expec ed o lead o a signi ican educ ion o
cold ex emes a ec ing midla i udes (Lo e al., 2023) as well as o la ge changes in he
equency, in ensi y and associa ed wind speeds o ex a opical cyclones (Hay e al., 2023).
A he same ime, i has been shown ha he obse ed la ge wa ming o e he A c ic egion
compa ed o he global mean (known as A c ic Ampli ica ion, AA) has al eady educed he
day- o-day su ace empe a u e a iance ac oss he No he n Hemisphe e (Sc een, 2014;
Blackpo e al., 2021). Na u ally, he e a e many o he d i e s o clima e a iabili y in
midla i udes beyond sea-ice loss, and he ela i e con ibu ion o sea-ice loss ela i e o o he
ac o s di e s be ween seasons, egions and a iables (e.g. Hay e al., 2022; Ouda e al., 2017;
Yu e al., 2024). In his s udy, howe e , we ocus exclusi ely on he ole o sea-ice loss.
The ongoing loss o sea-ice co e o e he A c ic egion may lead o sea ice- ee condi ions by
mid-cen u y (Kim e al., 2023; Shen e al., 2023), which could impac he midla i ude clima e.
In his ega d, a numbe o modelling s udies ha e shown inconsis en esul s on he possible
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
4
linkages be ween AA and midla i ude wea he and clima e (Blackpo and Sc een, 2020;
Galy ska e al., 2023; Sc een e al., 2018) in some cases due o unde sampling he ac ual in e nal
a iabili y o he clima e sys em (Peings e al., 2021; Ye e al., 2024), which mo i a es u he
analysis o he physical mechanisms sus aining such in e ac ions. This also includes dedica ed
assessmen s on how AA may al e he spa ial dis ibu ion, equency, in ensi y and pa hways
o synop ic-scale sys ems esponsible no only o e e yday wea he bu also o he occu ence
o ex emes. An example o such sys ems a e cu -o lows (COLs), which a e known o be
associa ed wi h ex eme ain all e en s o e se e al midla i ude egions.
COLs a e de ined as cold-co e quasi-ba o opic low p essu e sys ems ha o m in he mid- o
uppe -le els o he oposphe e om an ampli ied ba oclinic ough. This ough, unde
conduci e pa e ns o ho izon al empe a u e and ela i e o ici y ad ec ion, seg ega es a
o ici y maximum ha becomes de ached om he main low, leading o he o ma ion o a
COL (Palmen and New on, 1969; P ice and Vaughan, 1992). These sys ems ypically o m
o e se e al egions a ound he globe such as eas e n Aus alia (G os eld e al., 2021), sou he n
A ica (Fa e e al., 2013), wes e n Sou h Ame ica (Choquehuanca e al., 2025; Pinhei o e al.,
2017), he eas e n No h A lan ic and No h Paci ic as well as o e China (Nie o e al., 2005)
and can lead o se e e wea he condi ions including s ong winds and hea y ain all (Muñoz
and Shul z, 2021). In pa icula o e Eu ope, COLs o igina ing in he eas e n No h A lan ic
egion usually each he Medi e anean Sea, whe e hey can igge in ense ain all e en s
which accoun o a signi ican ac ion o he o al annual p ecipi a ion (Mas an onas e al.,
2021; Nie o, 2021; Po cù e al., 2007). The e o e, any e ec on he equency o COLs
eme ging ei he om na u al a iabili y o as a o ced clima e change esponse could lead o
signi ican impac s on he dis ibu ion and magni ude o ex eme ain all e en s as well as on
o al ain all amoun s, which migh be pa icula ly ele an o he Medi e anean Sea egion.
The main objec i e o his s udy is o assess whe he A c ic sea-ice loss can, h ough al e a ions
in he oposphe ic ci cula ion, modula e he dis ibu ion and equency o COLs o igina ing in
and a ec ing he Medi e anean Sea egion, and how hese could a ec p ecipi a ion
a iabili y. I is wo h men ioning ha a ecen pape by Hay e al. (2023) concluded ha A c ic
sea-ice loss can indeed shape se e al p ope ies o su ace ex a opical cyclones such as hei
equency, mean in ensi y o mean speed. Howe e , mos o he COLs a ec ing he
Medi e anean Sea do no ha e a su ace signa u e (i.e. a e no accompanied by a low p essu e
sys em a he su ace; Nie o e al., 2005) while s ill being capable o b inging se e e wea he
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
5
condi ions, which suppo s a dedica ed analysis o COLs a iabili y and changes o e ha
egion.
In his s udy, we use a se o Pola Ampli ica ion Model In e compa ison P ojec (PAMIP)
expe imen s o add ess he po en ial impac s o sea-ice mel on COLs and how hese can
modula e ex eme ain all e en s and o al p ecipi a ion o e he Medi e anean Sea egion.
These esul s could be o pa icula ele ance in he con ex o he se e e d ough condi ions
ha ha e been plaguing se e al pa s o he wes e n Medi e anean Sea egion in ecen yea s.
I is wo h s essing he e, howe e , ha om he app oach used in his s udy we canno
conclude ha A c ic sea-ice loss has caused (o will cause) p ecipi a ion anomalies o e
sou he n Eu ope, bu a he i i could (see ele an discussion in Ba nes and Sc een, 2015).
We p opose in u n a plausible physical mechanism linking sea-ice loss o he a iabili y o
COLs and Medi e anean p ecipi a ion.
This manusc ip is s uc u ed as ollows: Sec ion 2 ou lines he da ase s and me hods used in
he s udy, including a desc ip ion o he algo i hm used o iden i y COLs. Resul s on he
impac s om sea-ice loss on he equency o COLs and hei linkages wi h p ecipi a ion o e
he Medi e anean Sea egion a e included in Sec ion 3. Finally, a discussion o esul s and he
concluding ema ks a e summa ized in Sec ion 4.
2. Da a and me hods
a. PAMIP expe imen s
The PAMIP expe imen s (Smi h e al., 2019) used in his s udy a e named pdSST-pdSIC (“PD”)
and pdSST- u A cSIC (“FUT”) and we e designed as ep esen a i e o p esen day and u u e
(when global wa ming is 2°C abo e he p e-indus ial le el) sea-ice condi ions. The model
ou pu s consis o 100 ensemble membe s pe clima e model o 14-mon h-long a mosphe e-
only ime-slice uns in which a seasonal cycle o mon hly mean sea-su ace empe a u e (SST)
and sea-ice concen a ion (SIC) is conside ed. The i s wo mon hs o each simula ion a e
conside ed as spin up and disca ded, esul ing in a 12-mon h-long pe iod o analysis ( om he
mon h o June o he i s yea o he mon h o May o he ollowing yea ). In he case o PD,
he p esc ibed seasonal cycles o SST and SIC a e ob ained om obse a ions in he pe iod
1979-2008 om he Hadley Cen e Ice and Sea Su ace Tempe a u e da a se (HadISST;
Rayne e al., 2003), while in he case o FUT he A c ic SIC is de i ed om he ensemble
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
6
mean o CMIP5 model p ojec ions om he 30-yea pe iod in which global mean empe a u e
is 2 ºC abo e ha in he p e-indus ial pe iod. A eas in which sea ice is los ha e hei
co esponding SST alues aken om u u e p ojec ions, while elsewhe e hey a e he same as
in he PD expe imen s. As in simila s udies (e.g. Hay e al., 2023), he esponses o sea-ice
loss a e compu ed as he ensemble-mean di e ence be ween FUT and PD, wi h s a is ical
signi icance assessed in wo s eps: i s applying a wo-sided S uden ’s - es wi h a p- alue o
5% and hen imp o ing i by compu ing he alse disco e y a e p oposed by Wilks (2006) in
o de o a oid o e es ima ing ejec ion o he null hypo hesis (i.e. labeling as “signi ican ” a
di e ence ha is ac ually no ). Va iables conside ed o his s udy a e daily 500-hPa
geopo en ial heigh (h500) and daily p ecipi a ion co e ing he No he n Hemisphe e a ea
be ween 20ºN and 70ºN, as well as mon hly-mean zonal wind speed a 500 hPa (u500). These
we e ob ained om se en di e en PAMIP models (AWI-CM-1-1-MR, CESM2, CNRM-
CM6-1, EC-Ea h3, FGOALS- 3-L, IPSL-CM6A-LR and MIROC6) based on he a ailabili y
o daily h500 da a, all o which we e eg idded in o a common 1.5ºx1.5º g id p io o he
compu a ion o he me ics. The plo s and analysis in his s udy a e mos ly buil on he mul i-
model ensemble means (MMEM).
b. COL de ec ion algo i hm
We use he de ec ion algo i hm p oposed by Kasuga e al. (2021) o iden i y COLs a daily ime
s ep. This scheme allows o he iden i ica ion o COLs by using a 2D slope unc ion applied
o he h500 ield. The algo i hm looks o local minima in h500 occu ing oge he wi h local
maxima in he “a e age slope” o he same a iable, which is de ined as he spa ial a e age o
he h500 slopes om a gi en poin P owa ds he eas , wes , no h and sou h om he poin a
a gi en dis ance R. This dis ance R a ies om 200 o 2100 km in inc emen s o 100 km in
o de o co e he ypical spa ial ex en o hese wea he -scale sys ems (Ken a chos and Da ies,
1998). A gi en poin P is agged as a COL i a local minimum in h500 and a local maxima o
he a e age slope occu simul aneously. This me hodology is applied o all poin s P wi hin he
s udy domain, esul ing in daily ields o COL occu ence and loca ion. I should be men ioned
ha Kasuga e al. (2021) also p opose a numbe o es ic ions o hese condi ions in o de o
a oid classi ying as COLs sys ems ha a e ei he oo small o lie oo close o he equa o , o
example. These same es ic ions a e also applied in his s udy. Ano he poin wo h no ing is
ha he scheme de eloped by Kasuga e al. (2021) is, in spi e o i s simplici y o elying on a
single a iable a a single e ical le el, as accu a e in de ec ing COLs as o he mo e
sophis ica ed de ec ion schemes ha make use o mo e a iables/le els o hei compu a ions
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
7
(e.g. he one p oposed by Muñoz e al., 2020). The scheme om Kasuga e al. (2021) has been
used in a numbe o s udies, including a ecen assessmen o he skill o he ope a ional GFS
model o o ecas COL o ma ion and displacemen globally (Lupo e al., 2023). Fo a mo e
de ailed desc ip ion o he COL de ec ion scheme please e e o Kasuga e al. (2021) and Lupo
e al. (2023).
In his s udy, he de ec ion algo i hm was applied o daily h500 ields o e he No he n
Hemisphe e egions be ween 20ºN and 70ºN o PAMIP expe imen s. In o de o alida e he
ou comes o he de ec ion scheme, a simila p ocedu e was applied o daily ERA5 (He sbach
e al., 2020) eanalysis da a o h500. The daily ou pu s o he algo i hm we e agg ega ed
annually and hei analysis was cen e ed o e an ex ended Medi e anean Sea egion unning
om 10ºW o 30ºE and om 35ºN o 50ºN o accoun o COLs ha o igina e wes o he
Ibe ian Peninsula and a ec sou he n Eu ope (Nie o e al., 2005; Po cù e al., 2007).
c. Assessmen o COL-p ecipi a ion ela ionship
Linkages be ween COLs and p ecipi a ion o e he Medi e anean Sea egion a e explo ed by
using he me hodology p oposed by Ca o e al. (2012) and e ined by Ca o and P ahl (2013)
and P ahl and We nli (2012). This me hod was al eady applied on he explo a ion o linkages
be ween p ecipi a ion and su ace cyclones (P ahl and We nli, 2012) and on s (Ca o e al.,
2012; Ca o and P ahl, 2013), bu i is di ec ly adap able o o he wea he ea u es such as he
one conside ed in his s udy. The au oma ic p ocedu e assigns a gi en p ecipi a ion e en o a
COL i a COL lies wi hin a 5º box a ound he p ecipi a ion loca ion, as p oposed in Ca o e al.
(2012) o su ace on s. The ou pu s o he me hodology a e he p opo ion o p ecipi a ion
e en s (i.e. days wi h non-ze o p ecipi a ion amoun ) occu ing wi h a nea by COL a each g id
poin (F eq) and he p opo ion o p ecipi a ion amoun s alling in associa ion wi h a nea by
COL ela i e o he annual o al amoun (P ec). The p ocedu e is applied in his s udy o all
p ecipi a ion e en s as well as o hose associa ed wi h ex eme e en s, i.e. daily p ecipi a ion
alues exceeding he co esponding 90 h and 99 h pe cen iles (P90 and P99, espec i ely) a
each g id poin in o de o quan i y also he con ibu ion o COLs o he mos ex eme
p ecipi a ion e en s. Following his me hodology, we de i e he con ibu ion o COLs o
annual p ecipi a ion (bo h in e ms o occu ence o p ecipi a ion and o i s magni ude) and o
he occu ence o ex eme p ecipi a ion e en s (i.e., days exceeding P90 and P99 o daily
p ecipi a ion) in PD and FUT PAMIP expe imen s. Fo u he de ails on he me hodology o
de ec ing COL-p ecipi a ion associa ions, please e e o Ca o and P ahl (2013) and P ahl and
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
8
We nli (2012). Signi icance o he di e ences in P90 and P99 in FUT ela i e o PD PAMIP
expe imen s was assessed by assuming as null hypo hesis ha he wo g oups (FUT and PD)
a e iden ical (i.e. hey belong o he same popula ion and can he e o e be combined in o a
single sample). This pooled da ase is shu led and spli in o wo g oups o he same sizes as
he o iginal ones. Di e ences in pe cen iles ob ained om hese g oups a e compu ed se e al
imes (in ou case we pe o med 1000 i e a ions). We hen compu ed he ac ion o he sample
ha ac ually exceeds he gi en h eshold (P90 o P99), and when ha ac ion was below he
signi icance le el used he e (p-le el=0.05), we conside ed he di e ence as being signi ican .
This was epea ed o all g id poin s and he wo pe cen iles unde analysis, P90 and P99.
3. Resul s
a. COLs in PAMIP expe imen s and hei con ibu ion o p ecipi a ion
The MMEM annual-mean dis ibu ion o COLs in PD PAMIP expe imen s as well as hei
con ibu ion o p ecipi a ion e en s and amoun s a e displayed in Fig. 1. The spa ial ield (Fig.
1a) depic s h ee well-de ined equency maxima: one ex ending om he no heas A lan ic
in o he Ibe ian Peninsula, a second one co e ing mos o he eas e n Medi e anean Sea, and
a hi d one cen e ed o e highe la i udes be ween G eenland and Iceland. Fo e e ence, he
annual-mean ield de i ed om ERA5 da a in he same pe iod as in he PD expe imen s (1979-
2008; Fig. S1 in he Supplemen al Ma e ial) is o e all simila , e en hough he magni udes in
PAMIP a e sligh ly o e es ima ed o e he egions wi h highes COL equency, and in
pa icula o e he eas e n Medi e anean and sou he n G eenland. S ill, hese esul s highligh
he abili y o PAMIP models o cap u e he a eas wi h highes COL ac i i y o e he No h
A lan ic egion and Eu ope, e en mo e so conside ing ha e y li le in e -model sp ead is
ound o e hose a eas (no shown).
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
9
Figu e 1 MMEM annual-mean (a) equency o COLs (in numbe o e en s pe yea ) and hei
con ibu ion o (b) sou he n Eu ope annual p ecipi a ion e en s (days wi h p >0.0 mm) and (c) annual
p ecipi a ion amoun s in PD expe imen s.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
16
linea i (slope) wi h i s con idence in e al (alpha=0.05) in he uppe -le co ne . Co ela ions
signi ican ly di e en om ze o a e indica ed wi h an as e isk.
In o de o complemen ou analysis, in Fig. 6 we show he MMEM a ia ions in annual and
ex eme p ecipi a ion o e sou he n Eu ope a e sea-ice loss. The annual-mean ield (Fig. 6a)
depic s a signi ican educ ion in o al annual p ecipi a ion o e mos o cen al and eas e n
Spain, while a sligh inc ease is simula ed o e no hwes e n Spain and a b oade a ea wi h
posi i e a ia ions is ound o e cen al Eu ope, co e ing much o eas e n F ance and sou he n
Ge many. A he same ime, a ia ions in he magni udes o P90 and P99 (Fig. 6b,c) show
quali a i ely simila pa e ns, bu o P90 mos o he changes a e small and non-signi ican .
On he o he hand, he a ec a ion o P99 a ising om sea-ice loss is la ge and signi ican o e
basically he same a ea o he wes e n Medi e anean as ound o he annual mean, once again
s essing he la ge impac om COLs o he mo e ex eme p ecipi a ion e en s o e ha
egion.
4. Discussion and conclusions
Cu -o lows (COLs) a e a d i e o ex eme p ecipi a ion o e he wes e n Medi e anean Sea
egion. S ill, i is unclea whe he and how hese could be a ec ed by he ongoing loss o A c ic
sea-ice, which is expec ed o con inue in he u u e. In his s udy we ha e used a se o dedica ed
Pola Ampli ica ion Model In e compa ison P ojec (PAMIP) expe imen s o iden i y he
simula ed esponse o COLs equency and dis ibu ion o e sou he n Eu ope o educed A c ic
sea ice, and how hese could a ec mean and ex eme p ecipi a ion. The pa e n o sea-ice
educ ion used o o ce he expe imen s is ela ed o a global wa ming o 2°C abo e he p e-
indus ial le el.
F om he esul s ob ained in ou s udy, we can i s conclude ha PAMIP models a e able o
ep esen he p esen -day spa ial dis ibu ion, seasonali y and equency o COLs o e Eu ope,
e en hough wi h a sligh o e es ima ion o hei magni ude in some a eas. Meanwhile, he
compu a ion o he di e ences in COL densi y unde u u e (i.e. wi h educed sea-ice co e )
ela i e o p esen condi ions showed a educ ion in COLs o e pa s o sou hwes e n Eu ope,
and in pa icula o e he Ibe ian Peninsula. This esul s om an in ensi ica ion o he wes e ly
winds a ound 50ºN which is caused by a igh ening o he me idional p essu e g adien in
esponse o highe geopo en ial heigh s o i s sou h and mo e cyclonic condi ions u he no h,
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
17
e en hough his las ea u e is no consis en ly simula ed by all he PAMIP models conside ed
in ou s udy. I should be s essed, howe e , ha hese a ia ions in he a mosphe ic ci cula ion
a ise only om u u e sea-ice loss and should by no means be conside ed as p ojec ions. In
ac , se e al s udies (e.g. Hay e al., 2022; Sc een e al., 2022) ha e shown ha changes on he
a mosphe ic ci cula ion o he No he n Hemisphe e midla i udes d i en by sea-ice loss a e
opposi e in sign o hose d i en by g eenhouse gases inc ease, leading o a “ ug-o -wa ”
be ween bo h e ec s. As such, he u u e a ec a ion o COL equency o e sou he n Eu ope
esul ing om bo h mechanisms ac ing oge he may be e y di e en o ha de i ed in he
p esen s udy. In ac , a e y ecen pape by Mish a e al. (2025) sugges s ha bo h e ec s
aken oge he would lead o o e all small changes o he equency o COLs o e he egion
conside ed in his s udy, e en hough wi h la ge di e ences be ween seasons.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
18
Figu e 6 MMEM di e ences in he magni udes o (a) annual mean p ecipi a ion (in mm mon h-1) and
(b) P90 (c) P99 (in mm day-1) in FUT ela i e o PD simula ions. Di e ences ha esul signi ican ly
di e en om ze o (p<pFDR) a e highligh ed in do s.
Rega ding he linkages be ween COLs and p ecipi a ion o e he Medi e anean Sea egion,
we ound a igh ela ionship be ween bo h a iables pa icula ly o days wi h in ense
p ecipi a ion (conside ed he e as hose exceeding pe cen iles 90 and 99 o he dis ibu ion). We
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
19
also iden i ied a signi ican d op in he annual p ecipi a ion associa ed wi h COLs o e he
wes e n Medi e anean, as well as a sha p educ ion in he magni ude o P99 o e ha egion.
I is in e es ing o men ion ha mos o he di e ences iden i ied in p ecipi a ion a e loca ed o
he eas o he a ea wi h he la ges di e ences in COL equency. This esul s om he ac
ha he la ges p ecipi a ion ac i i y in COLs is commonly loca ed on hei eas e n lank whe e
he la ges uppe -le el di e gence (and associa ed upwa d mo ion) is ound (e.g. Po cù e al.,
2007). Ou esul s also sugges ha he dynamic ela ionship exis ing be ween COLs and
p ecipi a ion o e he wes e n Medi e anean egion should no change unde sea-ice loss. In
his ega d, he simula ed educ ion in mean and ex eme p ecipi a ion is explained by he
dec ease in COLs eaching ha egion. As a side no e, i is wo h no ing ha mos o he
a ia ions in COL equency in PAMIP expe imen s is concen a ed du ing he win e season
(no shown), which is consis en wi h he ac ha AA has i s la ges inge p in on he
a mosphe ic ci cula ion du ing ha season (e.g. P e idi e al., 2021).
The main objec i e o ou s udy was o documen po en ial a ec a ions o COLs o ced
exclusi ely by A c ic sea-ice loss. This app oach p e en s any di ec compa ison wi h he
obse ed a iabili y and ends o COLs wi hin he egion o in e es , gi en ha many o he
in e nal and ex e nal o cings would no be aken in o accoun . Howe e , i is no ewo hy ha
obse ed ends in COL equency du ing he las 40 yea s (Fig. S2) show dec easing
(inc easing) equency o e he no heas A lan ic (sou h o G eenland and B i ish Isles) in a
somewha simila pa e n o ha de i ed om he PAMIP models, which hin s a a plausible
con ibu ion om sea-ice loss. This mo i a es u u e wo k o a ibu e he obse ed COL ends
ex e nal o cing, including he po en ial ole o sea-ice loss, which would be ele an o
unde s anding he d i e s o he d ough condi ions a ec ing la ge a eas o sou he n Eu ope
and how hese migh e ol e in he u u e.
Acknowledgmen s
This wo k is pa o he Pola - o-mid-la i ude linkage e ec s on cold ai ou b eaks
(Pola 2MidLa ) p ojec , unded by he Eu opean Union’s Ho izon 2021 esea ch and
inno a ion p og amme unde he Ma ie Skłodowska-Cu ie g an ag eemen No 101061202.
Da a a ailabili y s a emen
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
20
The PAMIP da a is accessible on he Ea h Sys em G id Fede a ion websi e (h ps://es g-
node.llnl.go /sea ch/cmip6/). The ERA5 da a is eely a ailable h ough he Cope nicus
Clima e Da a S o e (h ps://cds.clima e.cope nicus.eu/).
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
21
Re e ences
Ba nes, E., and J. Sc een, 2015: The impac o A c ic wa ming on he midla i ude je -s eam:
Can i ? Has i ? Will i ? WIRES Clim. Change, 6, 277-286. h ps://doi.o g/10.1002/wcc.337.
Blackpo , R., and J. Sc een, 2020: Insigni ican e ec o A c ic ampli ica ion on he ampli ude
o midla i ude a mosphe ic wa es. Sci. Ad ., 6 (8), eaay2880.
h ps://doi.o g/10.1126/sciad .aay2880.
Blackpo , R., J. Fy e, and J. Sc een, 2021: Dec easing subseasonal empe a u e a iabili y in
he no he n ex a opics a ibu ed o human in luence. Na u e Geoscience, 14, 719-723.
h ps://doi.o g/10.1038/s41561-021-00826-w.
Boeke, R. C., P. C. Taylo , and S. A. Sejas, 2021: On he na u e o he A c ic’s posi i e lapse-
a e eedback. Geophys. Res. Le ., 48, e2020GL091109.
h ps://doi.o g/10.1029/2020GL091109.
Ca o, J., C. Jakob, G. Be y, and N. Nicholls, 2012: Rela ing global p ecipi a ion o
a mosphe ic on s. Geophys. Res. Le ., 39, L10805.
h ps://doi.o g/10.1029/2012GL051736.
Ca o, J., and S. P ahl, 2013: The impo ance o on s o ex eme p ecipi a ion. J. Geophys.
Res: A mos., 118, 10791-10801. h ps://doi.o g/10.1002/jg d.50852.
Chemke, R., L. Pol ani, J. Kay, and C. O be, 2021: Quan i ying he ole o ocean coupling in
A c ic ampli ica ion and sea-ice loss o e he 21s cen u y. npj Clim. A mos. Sci., 4, 46.
h ps://doi.o g/10.1038/s41612-021-00204-8.
Choquehuanca, B., A. Godoy, and R. Sau al, 2025: Assessing he skill o high-impac wea he
o ecas s in sou he n Sou h Ame ica: a s udy on cu -o lows. Wea. Cli. Dyn., in p ess.
h ps://doi.o g/10.5194/egusphe e-2024-1063.
Cohen, J., J. Sc een, J. Fu ado, M. Ba low, D. Whi les on, D. Coumou, J. F ancis, K. De hlo ,
D. En ekhabi, J. O e land, and J. Jones, 2014: Recen A c ic ampli ica ion and ex eme mid-
la i ude wea he . Na u e Geoscience, 7, 627-637. h ps://doi.o g/10.1038/ngeo2234.
Dai, A., and M. Jenkins, 2023: Rela ionships among A c ic wa ming, sea-ice loss, s abili y,
lapse a e eedback, and A c ic ampli ica ion. Clim. Dyn., 61, 5217-5232.
h ps://doi.o g/10.1007/s00382-023-06848-x.
England, M., and N. Feldl, 2024: Robus pola ampli ica ion in ice- ee clima es elies on ocean
hea anspo and cloud adia i e e ec s. J. Clima e, 37, 2179-2197.
h ps://doi.o g/10.1175/JCLI-D-23-0151.1.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
22
Fa e, A., B. Hewi son, C. Lenna d, R. Ce ezo-Mo a, and M. Tad oss, 2013: Cu -o lows in
he Sou h A ica egion and hei con ibu ion o p ecipi a ion. Clim. Dyn., 41, 2331-2351.
h ps://doi.o g/10.1007/s00382-012-1579-6.
Galy ska, E., K. Weigel, D. Hando , R. Jaise , R. Köhle , J. Runge, and V. Ey ing, 2023:
E alua ing causal A c ic-midla i ude eleconnec ions in CMIP6. J. Geophys. Res. A mos.,
128, e2022JD037978. h ps://doi.o g/10.1029/2022JD037978.
Ge ais, M., L. Sun, and C. Dese , 2024: Impac s o p ojec ed A c ic sea ice loss on daily
wea he pa e ns o e No h Ame ica. J. Clima e, 37, 1065-1085.
h ps://doi.o g/10.1175/JCLI-D-23-0389.1.
Goosse, H., J. Kay, K. A mou , A. Bodas-Salcedo, H. Chep e , D. Docquie , A. Jonko, P.
Kushne , O. Lecom e, F. Massonne , H.-S. Pa k, F. Pi han, G. S ensson, and M.
Vancoppenolle, 2018: Quan i ying clima e eedbacks in pola egions. Na . Comm., 9, 1919.
h ps://doi.o g/10.1038/s41467-018-04173-0.
G os eld, N., S. McG ego , and A. Tasche o, 2021: An au oma ed clima ology o cool-season
cu o lows o e sou heas e n Aus alia and ela ionship wi h he emo e clima e d i e s.
Mon. Wea. Re ., 149, 4167-4181. h ps://doi.o g/10.1175/MWR-D-21-0142.1.
Hay, S., P. Kushne , R. Blackpo , K. McCuske , T. Ouda , L. Sun, M. England, C. Dese , J.
Sc een, and L. Pol ani, 2022: Sepa a ing he in luences o low-la i ude wa ming and sea ice
loss on No he n Hemisphe e clima e change. J. Clima e, 35, 2327-2349.
h ps://doi.o g/10.1175/JCLI-D-21-0180.1.
Hay, S., M. P ies ley, H. Yu, J. Ca o, and J. Sc een, 2023: The e ec o A c ic sea-ice loss in
ex a opical cyclones. Geophys. Res. Le ., 50, e2023GL102840.
h ps://doi.o g/10.1029/2023GL102840.
He, M., Y. Hu, N. Chen, D. Wang, J. Huang, and K. S amnes, 2019: High cloud co e age o e
mel ed a eas domina es he impac o clouds on he albedo eedback in he A c ic. Sci. Rep.,
9, 9529. h ps://doi.o g/10.1038/s41598-019-44155-w.
He sbach, H., Bell, B., Be is o d, P., and coau ho s, 2020: The ERA5 global eanalysis. Qua .
J. Roy. Me . Soc., 146, 1999-2049. h ps://doi.o g/10.1002/qj.3803.
Kasuga, S., M. Honda, J. Uki a, S. Yamane, H. Kawase, and A. Yamazaki, 2021: Seamless
de ec ion o cu o lows and p eexis ing oughs. Mon. Wea. Re ., 149, 3119-3134.
h ps://doi.o g/10.1175/MWR-D-20-0255.1.
Kau z, L.-A., O. Ma ius, S. P ahl, J. Pin o, A. Ramos, P. Sousa, and T. Woollings, 2022:
A mosphe ic blocking and wea he ex emes o e he Eu o-A lan ic sec o - a e iew. Wea.
Cli. Dyn., 3, 305-336. h ps://doi.o g/10.5194/wcd-3-305-2022.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
23
Ken a chos, A., and T. Da ies, 1998: A clima ology o cu o lows a 200 hPa in he No he n
Hemisphe e, 1990-1994. In . J. Clima ol., 18, 379-390. h ps://doi.o g/10.1002/(SICI)1097-
0088(19980330)18:4<379::AID-JOC257>3.0.CO;2-F.
Kim, Y.-H., S.-K. Min, N. Gille , D. No z, and E. Malinina, 2023: Obse a ionally-cons ained
p ojec ions o an ice- ee A c ic e en unde a low emission scena io. Na . Comm., 14, 3139.
h ps://doi.o g/10.1038/s41467-023-38511-8.
Lenggenhage , S., and O. Ma ius, 2019: A mosphe ic blocks modula e he odds o hea y
p ecipi a ion e en s in Eu ope. Clim. Dyn., 53, 4155-4171. h ps://doi.o g/10.1007/s00382-
019-04779-0.
Lo, Y., D. Mi chell, P. Wa son, and J. Sc een, 2023: Changes in win e empe a u e ex emes
om u u e A c ic sea-ice loss and ocean wa ming. Geophys. Res. Le ., 50,
e2022GL102542. h ps://doi.o g/10.1029/2022GL102542.
Lupo, K., C. Schwa z, and G. Romine, 2023: Displacemen e o cha ac e is ics o 500-hPa
cu o lows in ope a ional GFS o ecas s. Wea. Fo ecas ing, 38, 1849-1871.
h ps://doi.o g/10.1175/WAF-D-22-0224.1.
Mas an onas, N., P. He e a-Lo mendez, L. Magnusson, F. Pappenbe ge , and J. Ma schulla ,
2021: Ex eme p ecipi a ion e en s in he Medi e anean: spa io empo al cha ac e is ics and
connec ion o la ge-scale a mosphe ic low pa e ns. In . J. Clima ol., 41, 2710-2728.
h ps://doi.o g/10.1002/joc.6985.
Mish a, A., D. Ma aun, R. Schiemann, K. Hodges, G. Zappa, and A. Ossó, 2025: Long-las ing
in ense cu -o lows o become mo e equen in he No he n Hemisphe e. Commun. Ea h
En i on., 6, 115. h ps://doi.o g/10.1038/s43247-025-02078-7.
Muñoz, C., and D. Schul z, 2021: Cu o lows, mois u e plumes, and hei in luence on
ex eme-p ecipi a ion days in cen al Chile. J. Appl. Me eo . Clima ol., 60, 437-454.
h ps://doi.o g/10.1175/JAMC-D-20-0135.1.
Muñoz, C., and G. Vaughan, 2020: A midla i ude clima ology and in e annual a iabili y o
200- and 500-hPa cu o lows. J. Clima e, 33, 2201-2222. h ps://doi.o g/10.1175/JCLI-D-
19-0497.1.
Nie o, R., 2021: Cu -o lows and ex eme p ecipi a ion in eas e n Spain: cu en and u u e
clima e. A mosphe e, 12, 835. h ps://doi.o g/10.3390/a mos12070835.
Nie o, R., L. Gimeno, L. de la To e, P. Ribe a, D. Gallego, R. Ga cía He e a, J. Ga cía, M.
Núñez, A. Redaño, and J. Lo en e, 2005: Clima ological ea u es o cu o low sys ems in
he No he n Hemisphe e. J. Clima e, 18, 3085-3103. h ps://doi.o g/10.1175/JCLI3386.1.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
24
No z, D., and J. S oe e, 2016: Obse ed A c ic sea-ice loss di ec ly ollows an h opogenic
CO2 emission. Science, 354, 747-750. h ps://doi.o g/10.1126/2019GL086749.
Ouda , T., E. Sánchez-Gómez, F. Chau in, J. Ca iaux, L. Te ay, and C. Cassou, 2017:
Respec i e oles o di ec GHG adia i e o cing and induced A c ic sea ice loss on he
No he n Hemisphe e a mosphe ic ci cula ion. Clim. Dyn., 49, 3693-3713.
h ps://doi.o g/10.1007/s00382-017-3541-0.
Palmen, E., and C. New on, 1969: A mosphe ic ci cula ion sys ems. Academic P ess, New
Yo k, 471-522.
Peings, Y., Z. Labe, and G. Magnusdo i , 2021: A e 100 ensemble membe s enough o cap u e
he emo e a mosphe ic esponse o +2ºC A c ic sea ice loss? J. Clima e, 34, 3751-3769.
h ps://doi.o g/10.1175/JCLI-D-20-0613.1.
P ahl, S., and H. We nli, 2012: Quan i ying he ele ance o cyclones o p ecipi a ion
ex emes. J. Clima e, 25, 6770-6780. h ps://doi.o g/10.1175/JCLI-D-00705.1.
Pinhei o, H., K. Hodges, M. Gan, and N. Fe ei a, 2017: A new pe spec i e o he
clima ological ea u es o uppe -le el cu -o lows in he Sou he n Hemisphe e. Clim. Dyn.,
48, 541-559. h ps://doi.o g/10.1007/s00382-016-3093-8.
Po cù, F., A. Ca assi, C. Medaglia, F. P odi, and A. Mugnai, 2007: A s udy on cu -o low
e ica s uc u e and p ecipi a ion in he Medi e anean egion. Me eo ol. A mos. Phys., 96,
121-140. h ps://doi.o g/10.1007/s00703-006-0224-5.
P e idi, M., K. Smi h, and L. Pol ani, 2021: A c ic ampli ica ion o clima e change: a e iew
o unde lying mechanisms. En i on. Res. Le ., 16, 093003. h ps://doi.o g/10.1088/1748-
9326/ac1c29.
P ice, J., and G. Vaughan, 1992: S a is ical s udies o cu -o low sys ems. Ann. Geophys., 10,
96-102.
Ran anen, M., A. Ka pechko, A. Liponen, K. No dling, O. Hy ä inen, K. Ruos eenoja, T.
Vihma, and A. Laaksonen, 2022: The A c ic has wa med nea ly ou imes as e han he
globe since 1979. Comm. Ea h En ., 3, 168. h ps://doi.o g/10.1038/s43247-022-00498-3.
Rayne , N., D. Pa ke , E. Ho on, C. Folland, L. Alexande , D. Rowell, and A. Kaplan, 2003:
Global analyses o sea su ace empe a u e, sea ice, and nigh ma ine ai empe a u e since
he la e nine een h cen u y. Jou nal o Geophysical Resea ch, 108, 4407.
h ps://doi.o g/10.1029/2002JD002670.
Russo o, R., and M. Biasu i, 2020: Pola ampli ica ion as an inhe en esponse o a ci cula ing
a mosphe e: esul s om he TRACMIP aquaplane s. Geophys. Res. Le ., 47,
e2019GL086771. h ps://doi.o g/10.1029/2019GL086771.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
25
Sc een, J., 2014: A c ic ampli ica ion dec eases empe a u e a iance in no he n mid- o high-
la i udes. Na . Clim. Change, 4, 577-582. h ps://doi.o g/10.1038/nclima e2268.
Sc een, J., and I. Simmonds, 2010: The cen al ole o diminishing sea ice in ecen A c ic
empe a u e ampli ica ion. Na u e, 464, 1334-1337. h ps://doi.o g/10.1038/na u e09051.
Sc een, J., C. Dese , D. Smi h, X. Zhang, R. Blackpo , P. Kushne , T. Ouda , K. McCuske ,
and L. Sun, 2018: Consis ency and disc epancy in he a mosphe ic esponse o A c ic sea-
ice loss ac oss clima e models. Na u e Geoscience, 11, 155-163.
h ps://doi.o g/10.1038/s41561-018-0059-7.
Sc een, J., and R. Blackpo , 2019: How obus is he a mosphe ic esponse o p ojec ed A c ic
sea-ice loss ac oss clima e models? Geophys. Res. Le ., 46, 11406-11415.
h ps://doi.o g/10.1029/2019gl084936.
Sc een, J., R. Eade, D. Smi h, S. Thompson, and H. Yu, 2022: Ne equa o wa d shi o he je
s eams when he con ibu ion om sea-ice loss is cons ained by obse ed eddy eedback.
Geophys. Res. Le ., 49, e2022GL100523. h ps://doi.o g/10.1029/2022GL100523.
Se eze, M., A. Ba e , J. S oe e, D. Kindig, and M. Holland, 2009: The eme gence o
su ace-based A c ic ampli ica ion. C yosphe e, 3, 11-19. h ps://doi.o g/10.5194/ c-3-11-
2009.
Shen, Z., W. Zhou, J. Li, and J. Chan, 2023: A equen ice- ee A c ic is likely o occu be o e
he mid-21s cen u y. npj Clim. A mos. Sci., 6, 103. h ps://doi.o g/10.1038/s41612-023-
00431-1.
Smi h, D., J. Sc een, C. Dese , J. Cohen, J. Fy e, J. Ga cía-Se ano, T. Jung, V. Ka so , D.
Ma ei, R. Msadek, Y. Peings, M. Sigmund, J. Uki a, J.-H. Yoon, and X. Zhang, 2019: The
Pola Ampli ica ion Model In e compa ison P ojec (PAMIP) con ibu ion o CMIP6:
in es iga ing he causes and consequences o pola ampli ica ion. Geoscien i ic Model
De elopmen , 12, 1139-1164. h ps://doi.o g/10.5194/gmd-12-1139-2019.
Smi h, D., E. Eade, M. And ews, H. Ay es, A. Cla k, S. Ch ipko, C. Dese , N. Duns one, J.
Ga cía-Se ano, G. Gas ineau, L. G a , S. Ha diman, B. He, L. He manson, T. Jung, J.
Knigh , X. Le ine, G. Magnusdo i , E. Manzini, D. Ma ei, M. Mo i, R. Msadek, P. O ega,
Y. Peings, A. Scai e, J. Sc een, M. Seab ook, T. Semmle , M. Sigmond, J. S e ing, L. Sun,
and A. Walsh, 2022: Robus bu weak win e a mosphe ic ci cula ion esponse o u u e
A c ic sea ice loss. Na . Comm., 123, 727. h ps://doi.o g/10.1038/s41467-022-28283-y.
Sousa, P., R. T igo, D. Ba ioped o, P. Soa es, A. Ramos, and M. Libe a o, 2017: Responses
o Eu opean p ecipi a ion dis ibu ions and egimes o di e en blocking loca ions. Clim.
Dyn., 48, 1141-1160. h ps://doi.o g/10.1007/s00382-016-3132-5.
Accep ed o publica ion in Jou nal o Clima e. DOI 10.1175/JCLI-D-25-0066.1.
Unau hen ica ed | Downloaded 05/15/25 02:37 PM UTC
