S o m Amy likely in ensified by bo h human d i en
clima e change and na u al a iabili y
Con ac Au ho s
- Mi eia Gines a, Ox o d Sus ainable Law P og amme - Uni e si y o Ox o d, UK
📨
mi eia.gines a@smi hschool.ox.ac.uk
🗣
English, Spanish, Ca alan
- Da ide Fa anda, IPSL-CNRS, F ance
📨
da ide. a [email p o ec ed]
🗣
English,F ench,
I alian
P ess Summa y
■ S o ms simila o Amy a e up o 2 hP deepe , up o 4 km/h (10%) windie o e
he A lan ic coas s o I eland and F ance, and up o 3 mm/day (up o 10%) we e
in he p esen han hey would ha e been in he pas .
■ S o m Amy was d i en by e y a e me eo ological condi ions.
■ Na u al a iabili y alone canno explain he inc ease in ain and wind linked o
S o m Amy
E en Desc ip ion
S o m Amy was a apidly deepening No h A lan ic ex a opical cyclone and he fi s
named s o m o he 2025/26 UK–I eland season. The emnan s o Hu icanes Humbe o
and Imelda in e ac ed wi h he No h A lan ic je s eam, accele a ing i s flow and
con ibu ing o he apid in ensifica ion o he low-p essu e sys em ha became S o m
Amy.
The sys em acked eas wa ds om he mid-A lan ic and eached he UK and I eland
be ween F iday 3 Oc obe and Sa u day 4 Oc obe 2025, b inging i s s onges impac s
du ing ha pe iod. Ad ance wa nings we e issued o 2–4 Oc obe . Al hough he e was
some unce ain y abou he exac magni ude o he s o m, mos majo o ecas models
— including hose om he UK Me O fice, ECMWF, GFS, and ICON — consis en ly
p edic ed a pe iod o e y s ong winds and hea y ain all ac oss he UK and No he n
I eland. A peak gus o 96 mph (155 km/h) was eco ded on he Island o Ti ee a he
heigh o he s o m. The s o m showed a cen al p essu e o 947.9hPa a Bal asound,
She land, exceeding he p e ious Oc obe eco d o 950.9hPa in 1988.
The s ong winds and hea y ain caused widesp ead a el dis up ion, allen ees,
powe ou ages, and significan coas al impac s ac oss Sco land, no he n England, and
much o I eland. In I eland, he mos se e e condi ions occu ed in he wes and
no hwes , whe e au ho i ies issued o ange and ed wind and flood wa nings. One
s o m- ela ed a ali y was confi med in Coun y Donegal. S o m Amy also b ough
se e e wea he o pa s o con inen al Eu ope, including No way, Sweden, Belgium, and
F ance. On Sa u day a e noon, powe ul gus s associa ed wi h he s o m esul ed in
wo a ali ies in no he n F ance, whe e au ho i ies issued an o ange-le el wea he
ale .
The Su ace P essu e Anomalies e eal a la ge nega i e (cyclonic) anomaly o e Sco land
wi h alues up o -8hPA. In he con ex o A lan ic ex a opical s o ms, his se up can
be associa ed wi h ex ensi e p ecipi a ion and in ense wind due o he deepening
p essu e g adien . Tempe a u e Anomalies show empe a u es up o 4 °C wa me han
a e age o e he UK, a ec ed by he wa m sec o o he cyclone. P ecipi a ion Da a
show in ense daily p ecipi a ion (up o 50 mm/day) o e I eland, UK, F ance and No h
Sea. Windspeed da a show la ge a eas o UK and F ance a ec ed by sus ained winds
be ween 40 km/h and 60 km/h, and highe han 60 km/h o e he No he n sea.
Clima e and Da a Backg ound o he Analysis
The IPCC AR6 Chap e 11 s a es ha clima e change a ec s s o miness in Eu ope, wi h
nega i e epe cussions ha a e exace ba ed by ising sea le els and hea y p ecipi a ion.
F om a global wa ming le el o 2 °C o highe a sligh ly inc eased equency and
ampli ude o ex a opical cyclones, s ong winds and ex a opical s o ms is p ojec ed
o no he n Eu ope wi h medium confidence (IPCC AR6 WGI Chap e 12). The ex eme
p ecipi a ion and plu ial flooding ha e shown a s ong end in No he n Eu ope in he
pas and he e is a s ong consensus among models om global o kilome e scale
con ec ion pe mi ing models ha a e going o inc ease om 2 °C o highe (IPCC AR6
WGI Chap e 12). S udies, such as Gines a e al. (2024), confi m ha he in ensi y o
ecen s o ms, including bo h wind speed and p ecipi a ion, is likely o inc ease in he
mos impac ed egions o Eu ope.
Ou analysis app oach es s on looking o wea he si ua ions simila o hose o he
e en o in e es ha ing been obse ed in he pas . Fo s o m Éowyn, we ha e
medium-high confidence in he obus ness o ou app oach gi en he a ailable clima e
da a, as he e en is simila o o he pas e en s in he da a eco d.
ClimaMe e Analysis
We analyze he e (see Me hodology o mo e de ails) how e en s simila o he
me eo ological condi ions leading o he S o m Amy ha e changed in he p esen
(1987–2023) compa ed o wha hey would ha e looked like i hey had occu ed in he
pas (1950–1986) in he egion [20°W 5°E 46°N 62°N]. The Su ace P essu e Changes
show ha s o ms simila o Amy a e up o 2 hPa deepe in he p esen clima e, wi h an
inc ease in he ho izon al p essu e g adien in he sou heas e n sec o o he s o ms.
Tempe a u e Changes show ha p esen -day e en s a e up o +1.5°C wa me in he
s o m Amy wa m sec o compa ed o simila pas s o ms. P ecipi a ion Changes show
we e condi ions, wi h up o +3 mm/day (a ound +10%) mo e ain all. Windspeed
Changes show significan inc eases o up o +5 km/h in coas al and o sho e a eas.
We also no e a sligh shi in s o m seasonali y: s o ms now end o occu mo e o en a
he s a o he season, whe eas in he pas hey we e mo e equen owa d he end.
Changes in u ban a eas e eal ha Glasgow, Bel as and Dublin expe ience windie and
we e condi ions in he p esen han in he pas .
Finally, we find ha sou ces o na u al clima e a iabili y, no ably he El Nino—Sou he n
Oscilla ion and he A lan ic Mul idecadal Oscilla ion, may ha e only pa ly influenced
he e en . This sugges s ha he changes we see in he e en compa ed o he pas may
be pa ly due o human d i en clima e change, wi h a con ibu ion om na u al
a iabili y.
Conclusion
Based on he abo e, we conclude ha s o ms simila o Amy a e 2hPa deepe , up o 3
mm/day (10%) we e , and windie by up o 4 km/h (10%) in he p esen compa ed o
he pas . We in e p e S o m Amy as an e en o e y a e me eo ological condi ions o
which na u al clima e a iabili y played a ole.
Addi ional In o ma ion : Comple e Ou pu o he Analysis
NB1: The ollowing ou pu is specifically in ended o scien is s and con ain de ails ha
a e ully unde s andable only by eading he me hodology desc ibed in Fa anda, D.,
Bou din, S., Gines a, M., K ouma, M., Noyelle, R., Pons, F., Yiou, P., and Messo i, G.: A
clima e-change a ibu ion e ospec i e o some impac ul wea he ex emes o 2021,
Wea he Clim. Dynam., 3, 1311–1340, h ps://doi.o g/10.5194/wcd-3-1311-2022, 2022.
NB2: Colo scales may a y om he ClimaMe e figu e p esen ed abo e.
The igu e shows he a e age o su ace p essu e anomaly (msl) (a), a e age 2-me e empe a u es
anomalies ( 2m) (e), cumula ed o al p ecipi a ion ( p) (i), and a e age wind-speed (wspd) in he
pe iod o he e en . A e age o he su ace p essu e analogs ound in he coun e ac ual (b) and ac ual
pe iods (c), along wi h co esponding 2-me e empe a u es ( , g), cumula ed p ecipi a ion (j, k), and
wind speed (n, o). Changes be ween p esen and pas analogues a e p esen ed o su ace p essu e
∆slp (d), 2 me e empe a u es ∆ 2m (h), o al p ecipi a ion ∆ p (i), and windspeed ∆wspd (p):
colo - illed a eas indica e signi ican anomalies wi h espec o he boo s ap p ocedu e. Violin plo s
o pas (blue) and p esen (o ange) pe iods o Quali y Q analogs (q), P edic abili y Index D ( ),
Pe sis ence Index Θ (s), and dis ibu ion o analogs in each mon h ( ). Violin plo s o pas (blue) and
p esen (o ange) pe iods o ENSO (u), AMO ( ) and PDO (w). Numbe o he Analogues occu ing
in each subpe iod (blue) and linea end (black). Values o he peak day o he ex eme e en a e
ma ked by a blue do . Ho izon al ba s in panels (q, ,s,u, ,w) co espond o he mean (black) and
median ( ed) o he dis ibu ions. (x) Numbe o analogues ound in sub pe iods when analogues a e
sea ched in he whole eanalysis pe iod.
