Assessment of the accuracy in UV index modelling using the UVIOS2 system during the UVC-III campaign

Geosci. Model De ., 18, 7451–7473, 2025
h ps://doi.o g/10.5194/gmd-18-7451-2025
© Au ho (s) 2025. This wo k is dis ibu ed unde
he C ea i e Commons A ibu ion 4.0 License.
Model e alua ion pape
Assessmen o he accu acy in UV index modelling using he
UVIOS2 sys em du ing he UVC-III campaign
Ilias Foun oulakis1, Ky iaki Papach is opoulou2, S elios Kazadzis2, G ego Hülsen2, Julian G öbne 2,
Ioannis-Panagio is Rap is3, Dimi a Kouklaki4,5, Ak i i Masoom2, Na alia Kou eme i2, Cha alampos Kon oes4, and
Ch is os S. Ze e os1,6,7,8
1Resea ch Cen e o A mosphe ic Physics and Clima ology, Academy o A hens, 10680 A hens, G eece
2Physikalisch-Me eo ologisches Obse a o ium Da os, Wo ld Radia ion Cen e (PMOD/WRC), Da os Do , Swi ze land
3Ins i u e o As onomy, As ophysics, Space Applica ions and Remo e Sensing,
Na ional Obse a o y o A hens, A hens, G eece
4Ins i u e o En i onmen al Resea ch and Sus ainable De elopmen ,
Na ional Obse a o y o A hens, GR15236 A hens, G eece
5Labo a o y o Clima ology and A mosphe ic En i onmen , Depa men o Geology and Geoen i onmen ,
Na ional and Kapodis ian Uni e si y o A hens, 15784 A hens, G eece
6Biomedical Resea ch Founda ion o he Academy o A hens, 11527 A hens, G eece
7Na a ino En i onmen al Obse a o y (N.E.O.), 24001 Messenia, G eece
8Ma iolopoulos-Kanaginis Founda ion, 10675 A hens, G eece
Co espondence: Ilias Foun oulakis (i. oun oulakis@academyo a hens.g )
Recei ed: 21 Sep embe 2024 – Discussion s a ed: 4 Decembe 2024
Re ised: 14 Augus 2025 – Accep ed: 8 Sep embe 2025 – Published: 21 Oc obe 2025
Abs ac . The hi d campaign o he calib a ion and in e -
compa ison o sola UV adiome e s (UVC III) ook place
a Da os, Swi ze land in June–Augus 2022. Mo e han 70
adiome e s pa icipa ed in he campaign and measu ed side-
by-side wi h he po able e e ence spec o adiome e QA-
SUME. The UVIOS2 sys em is a lexible UVI modelling ool
ha can be exploi ed o di e en applica ions depending on
he inpu s. Thus, di e en combina ions o sa elli e, eanaly-
sis, and/o g ound-based inpu s we e used o es he UVIOS2
pe o mance when i is used as a ool o UVI nowcas ing
o o clima ological s udies. While UVIOS2 p o ided qui e
accu a e es ima es o he a e age ( o he pe iod o he cam-
paign) UVI le els, la ge de ia ions we e ound o indi id-
ual es ima es. The a e age ag eemen be ween he UVI om
he UVIOS2 and QASUME was be e han 1% o all he
di e en se s o inpu s ha we e used o he s udy. The ange
o he a iabili y was o he o de o 40% o ins an aneous
measu emen s (15min), mainly due o he model’s inabil-
i y o cap u e he ins an aneous e ec s o cloudiness, espe-
cially unde b oken cloud condi ions. Unde clea -sky condi-
ions he model was ound o pe o m much be e , wi h he
di e ences be ween he model es ima es and he QASUME
measu emen s being smalle han 12% o 95 % o he s ud-
ied cases. E en a he p is ine en i onmen o Da os, single
sca e ing albedo (SSA) was ound o con ibu e signi ican ly
o he modelling unce ain ies unde cloudless condi ions.
Fo Ae osol Op ical Dep h (AOD) o he o de o 0.2–0.4
a 550nm, he ole o he SSA was ound o be compa able
o he ole o AOD in he modelling o he UVI.
1 In oduc ion
Exposu e o sola ul a iole (UV) adia ion is i al o many
li ing o ganisms including humans (e.g., Caldwell e al.,
1998; E ickson III e al., 2015; Häde , 1991; Häde e al.,
1998; Juzeniene e al., 2011; Lucas e al., 2019) bu can
be ha m ul when i exceeds ce ain limi s (Di ey, 1991).
Exposu e o he human skin o UV adia ion is he main
mechanism ha d i es he o ma ion o i amin D, which,
in u n, con ibu es o he s eng hening o he immune sys-
em (e.g., Lucas e al., 2019; Webb e al., 2022). Mode a e
Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union.
7452 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
exposu e o UV adia ion has many mo e bene i s o hu-
man heal h ha a e no ela ed o he o ma ion o i amin D,
such as he con ibu ion o he main enance o a good men al
heal h and he cu a ion o a ious skin diseases (Juzeniene
and Moan, 2012). Ne e heless, o e exposu e o UV adia-
ion is he main en i onmen al isk ac o o non-melanoma
skin cance , and among he main en i onmen al isk ac o s
o melanoma skin cance and ca a ac (WHO, 1994). De e -
mina ion o op imal sun exposu e beha io s is no a simple
ask and, addi ionally o he su ace sola UV adia ion a ail-
abili y, i also depends on he physiology o each indi idual
pe son (e.g., A ms ong and Cus , 2017; Ho mann and Me -
e , 2005; Lucas e al., 2019; McKenzie and Lucas, 2018;
Webb e al., 2018; Webb and Engelsen, 2006).
A commonly used quan i y o human heal h pu poses is
he UV index (UVI) (Schmalwiese e al., 2017; Vanicek e
al., 2000), which is a me ic o he e iciency o UV adia-
ion o cause e y hema o he human skin. Gene ally, smalle
exposu e imes and mo e p ecau ion measu es a e ecom-
mended wi h inc easing UVI. UVIs smalle han 2 a e con-
side ed low, UVIs o 8–10 a e conside ed e y high, and
UVIs exceeding 10 a e conside ed ex eme. In he 1980s and
he 1990s, public awa eness was caused due o he se e e
ozone deple ion o e high and mid la i udes which, i con in-
ued, would esul in ex eme UVI le els o e densely popu-
la ed egions o ou plane ( an Dijk e al., 2013; Newman
and McKenzie, 2011). Al hough he adop ion and he suc-
cess ul implemen a ion o he Mon eal P o ocol p e en ed
u he deple ion o s a osphe ic ozone and he consequen
dange ous UV le els (McKenzie e al., 2019; Mo gens e n e
al., 2008), he u u e e olu ion o he le els o su ace sola
UV adia ion is s ill unce ain, mainly due o he unce ain-
ies in he impac o clima e changes on su ace sola UV
adia ion (Be nha d e al., 2023; Ze e os e al., 2023).
Since he 1980s, na ional and in e na ional ne wo ks o
he moni o ing o he UVI ha e been es ablished o ensu e
accu a e and imely in o ma ion o he public (Blum hale ,
2018; Schmalwiese e al., 2017). Main enance o a s a ion
ha p o ides eliable UV measu emen s demands p ope ly
ained pe sonnel o un he s a ion and applica ion o s ic
calib a ion and main enance p o ocols. Fu he mo e, he e
a e p e equisi es o he ins alla ion o such s a ions (e.g.,
powe supply, sa e y). Thus, i is impossible o achie e UVI
moni o ing wi h global co e age om he g ound. P og ess
in sa elli e moni o ing du ing he las decades allowed he e-
ie al o he UVI on a global scale. Cu en ly, he UVI has
been es ima ed wi h high spa ial and empo al co e age using
a ious echniques and a ious sa elli e p oduc s (e.g., see
Table 1 in Ze e os e al., 2023). One o he mos widely used
clima ological UVI da ase s is p o ided by he T oposphe ic
Emission Moni o ing In e ne Se ice (TEMIS). TEMIS p o-
ides clea -sky UV doses since 1960 and all-sky UV doses
since 2004, ha ha e been calcula ed using measu emen s
om a ious sa elli e senso s (TEMIS, 2025; Zempila e al.,
2017). Widely used clima ological da ase s o he UVI wi h
global co e age ha e been also e ie ed using measu emen s
om he To al Ozone Mapping Spec ome e (TOMS) (He -
man e al., 1999), he Ozone Moni o ing Ins umen (OMI)
(Tanskanen e al., 2006), and he TROPOsphe ic Moni o -
ing Ins umen (TROPOMI) (Lind o s e al., 2018). As a e-
sul o he apid p og ess in Ea h obse a ion moni o ing,
he a o emen ioned clima ological sa elli e-based UV p od-
uc s ha e been p o en o be eliable o e wide egions o
he plane (e.g., Lakkala e al., 2020; Zempila e al., 2016,
2017), al hough biases o he o de o 10%–20 % ha e been
epo ed o e complex and pollu ed en i onmen s, while un-
ce ain ies can be e en la ge o e highly e lec i e e ains
a high la i udes (e.g., Lakkala e al., 2020). The accu acy
o sa elli e-based es ima es is limi ed due o he ini e wid h
o he sa elli e pixel (Kazadzis e al., 2009) and he weak-
ness o sa elli e senso s o accu a ely p obe he lowe opo-
sphe e (Bais e al., 2019). In pa icula , assump ions a e made
in he sa elli e algo i hms o desc ibe he complex in e ac-
ions be ween adia ion, ae osols and clouds, which inc ease
he unce ain y in he e ie als. Unce ain ies in he assumed
ae osol p ope ies (A ola e al., 2021; Pa isi e al., 2021), in-
accu a e dis inc ion o he e ec o highly e lec ing e ains
and cloudiness (Be nha d e al., 2015; Lakkala e al., 2020),
and unce ain ies in he desc ip ion o cloud co e , especially
o e high-al i ude si es (B ogniez e al., 2016; Schenzinge e
al., 2023) a e among he main unce ain y sou ces.
Me eo ological se ices p o ide UVI o ecas ing ha is
usually based on me eo ological o ecas ing in conjunc ion
wi h adia i e ans e models (e.g., Feis e e al., 2011; Long
e al., 1996; Roshan e al., 2020). The Cope nicus A mo-
sphe ic Moni o ing Se ice (CAMS) – A mosphe e o ex-
ample, p o ides i e days clea -sky and all-sky UVI o e-
cas s on a global scale based on he syne gis ic analyses
o Ea h-obse a ion da a, wea he p edic ion and chemis y
model o ecas s, and adia i e ans e modelling (Peuch e
al., 2022; Schulz e al., 2022). UVI o ecas s a e commonly
go e ned by he unce ain ies in he o ecas ed me eo olog-
ical pa ame e s, mainly cloudiness (e.g., Schenzinge e al.,
2023). Geos a iona y sa elli es p o ide con inuous, nea ly in-
s an in o ma ion o cloudiness o e wide egions o he
plane (De ien and Le Gléau, 2005), which can be used
o p o ide mo e accu a e UVI es ima es in nea ly eal ime
(Kosmopoulos e al., 2020) o UVI clima ological p oduc s
(e.g., A ola e al., 2002; F agkos e al., 2024; Ve debou ,
2000; Zempila e al., 2017).
Moni o ing and/o o ecas ing o he UVI a moun ain-
ous si es is excep ionally challenging. Complex a mosphe ic
condi ions and complex e ains inc ease he unce ain ies in
he modelling o he UVI, while calib a ion and main enance
o senso s is no easy due o di icul ies in access, powe
supply, and ha sh wea he condi ions. Ne e heless, UVI in-
c eases wi h al i ude and can each ex eme le els, which
makes his in o ma ion aluable o he inhabi an s and he
isi o s o such loca ions. Fo example, ex eme UVI o ∼20
has been eco ded in he Boli ian Andes (P ei e e al., 2006;
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7453
Za a i e al., 2003). Ele a ed UVI le els ha e been also
eco ded a high-al i ude dese s in A gen ina (Piacen ini e
al., 2003), while UVI equen ly exceeding 15 has been mea-
su ed a Tibe (Dahlback e al., 2007). UVIs equen ly ex-
ceeding 11 ha e been also measu ed a Eu opean alpine s a-
ions (Casale e al., 2015) as well as a high al i ude loca ions
in No hwes e n A gen ina (U illas e al., 2016). Depending
on a mosphe ic and e ain condi ions, inc eases o he su -
ace sola UV adia ion le els wi h al i ude can ange om a
ew pe cen pe km (Chuba o a and Zhdano a, 2013; P ei e
e al., 2006; Riede e al., 2010; Schmucki and Philipona,
2002; Za a i e al., 2003) o 10%–20 % (e.g., Chuba o a e
al., 2016; Sola e al., 2008), o e en o mo e han 30%km−1
when su ace albedo also inc eases wi h al i ude (Be nha d
e al., 2008; P ei e e al., 2006). Du ing summe (i snow is
absen ), UVI inc eases wi h al i ude mainly due o dec eased
Rayleigh sca e ing (Allaa e al., 2004; Blum hale e al.,
1994; Sola e al., 2008). In gene al, he change in he le -
els o he sola UV i adiance wi h al i ude depends on a -
mosphe ic composi ion and has a s ong wa eleng h depen-
dence which is in oducing di icul ies in he modelling o
he UVI a moun ainous si es (e.g., D o kin and S einbe ge ,
1999; K o ko e al., 1998). A e y high-al i ude (o /and la -
i ude) si es, ice and/o snow may pe sis e en in la e sp ing
and summe esul ing in ex emely high UV exposu e (e.g.,
Schmalwiese e al., 2017; Siani e al., 2008; U illas e al.,
2016).
The con inuous ope a ion o g ound-based ne wo ks ha
p o ide highly accu a e in o ma ion is necessa y, no only
o he in o ma ion o he public, bu also o he alida-
ion and he imp o emen o sa elli e based UVI clima olog-
ical and o ecas /nowcas p oduc s (e.g., Foun oulakis e al.,
2020b). In addi ion o he s ic main enance, ope a ion, and
calib a ion p o ocols ha mus be applied by he moni o ing
s a ions ope a o s (e.g., Foun oulakis e al., 2020a; Ga ane
e al., 2006; G öbne e al., 2006; Lakkala e al., 2008),
pa icipa ion o he ins umen s o ield campaigns u he
ensu es he high quali y and he homogenei y o he mea-
su ed UVIs a di e en s a ions (Bais e al., 2001; Hülsen
e al., 2020). The unce ain y in he UVI measu ed by he
mos accu a e spec o adiome e s ha se e as wo ld e e -
ences can each 2% (G öbne and Spe eld, 2005; Hülsen e
al., 2016). B oadband il e adiome e s ha a e commonly
used in egional, na ional, o in e na ional ne wo ks o UVI
moni o ing a e a ec ed by la ge unce ain ies. In he con-
ex o he sola ul a iole il e adiome e compa ison cam-
paigns (UVC, UVC-II, and mos ecen ly UVC-III) ha we e
o ganized by he Physikalisch-Me eo ologisches Obse a o-
ium Da os, Wo ld Radia ion Cen e (PMOD/WRC) in 2006,
2017, and 2022 many b oadband adiome e s measu ed side-
by-side wi h he wo ld e e ence QASUME (e.g., Hülsen e
al., 2020; Hülsen and G öbne , 2007). Analyses o he mea-
su emen s by he 75 ins umen s ha pa icipa ed in UVC-II
esul ed in he es ima ion o a calib a ion unce ain y o 6%.
The o e all unce ain y in he measu emen s was la ge , due
o o he ac o s, mainly he impe ec angula esponse o he
adiome e s (Hülsen e al., 2020).
Fu he mo e, Da os is one o he ew moun ainous si es in
he wo ld whe e bo h, highly accu a e UVI measu emen s,
and measu emen s o he main ac o s ha de e mine he le -
els o he UVI a he su ace (and can be used as inpu s o
i s modelling) a e a ailable, which allows us o assess he e -
icacy o a s a e-o - he-a UVI model o p oduce es ima es
and econs uc UVI se ies unde such condi ions.
The i s e sion o he UVIOS (UV-Index Ope a ing Sys-
em) nowcas ing sys em has been al eady desc ibed in Kos-
mopoulos e al. (2021). The sys em has been upg aded e-
cen ly in o de o achie e as e and mo e accu a e simula-
ions. The new, imp o ed UVIOS2 adia i e ans e scheme
can be used ei he as a ool o UVI nowcas ing and o e-
cas ing o o clima ological s udies, depending on he in-
pu s. In his pape , he UVI ha has been simula ed using
he new UVIOS2 sys em wi h di e en inpu s is desc ibed
and alida ed agains e y accu a e g ound-based UVI mea-
su emen s ha we e pe o med du ing he UVCIII campaign.
The wo ld e e ence QASUME ha ope a ed du ing he cam-
paign p o ides measu emen s ha a e ideal o he alida-
ion o UVIOS2 due o hei high accu acy, which allows he
iden i ica ion o he unce ain ies in he modelling o UVI by
UVIOS2. Highly accu a e ancilla y measu emen s ha we e
a ailable a he same pe iod also allow he iden i ica ion o
he unce ain y sou ces in he UVI modelling. The main a -
ge s o he s udy can be summa ized as ollows:
–Desc ibe he upg ades in UVIOS2 ela i e o he p e i-
ous (UVIOS) sys em.
–Quan i y he unce ain ies, and he main unce ain y ac-
o s, in UVIOS2 simula ions du ing he UVCIII cam-
paign, when i is used as a ool o UVI nowcas ing and
clima ological analysis.
–E alua e and discuss in dep h he unce ain y ac o s in
he modelling o UVI a complex opog aphy si es such
as Da os.
–Discuss he unce ain y in o ecas ed UVI wi h espec
o he unce ain y in he measu emen s o il e adiome-
e s and discuss wha a e he p e equisi es o imp o ed
UVI modelling.
I mus be cla i ied ha he s udy e e s o a snow- ee pe-
iod a Da os, and hus he unce ain ies ela ed o he pa-
ame e iza ion o su ace albedo, which may be signi ican
o highe al i ude si es e en in he summe , a e no quan-
i ied o discussed he e. The pape is o ganized as ollows.
A desc ip ion o he used da a and me hods is p o ided in
Sec . 2. The esul s o he analysis a e discussed in Sec . 3,
and he main conclusions a e summa ized in Sec . 4.
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025
7454 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
2 Me hodology
The UVIOS2 sys em is a lexible ool ha can be exploi ed
o di e en applica ions depending on he inpu s. I can be
used ei he as a nowcas ing/ o ecas ing ool, o o pe o m
clima ological s udies. The accu acy o he simula ed UVI
depends on he comp omise be ween he achie emen o e-
alis ic compu a ional imes (i.e., he spa ial and empo al ex-
en o he simula ions) and he use o he mos accu a e
model inpu s. In he con ex o his wo k, we assessed he
accu acy o UVIOS2 when i ope a es o eal ime appli-
ca ions (i.e., de aul se up ha is used o simula e he eal-
ime UVI o e Eu ope) and when i is used o clima ological
s udies (i.e., using g ound-based measu emen s o eanaly-
sis da a as inpu s) a he moun ainous en i onmen o Da os,
Swi ze land du ing he UVC III campaign (Hülsen and G öb-
ne , 2023). Assessmen o he accu acy in UVIOS2 o ecas s
is ou o he scope o he p esen s udy.
2.1 The UVC-III campaign
The hi d In e na ional Sola UV Radiome e Calib a ion
Campaign (UVC-III) ook place a Da os, Swi ze land
(Fig. 1; 46.8°N, 9.83° E, 1610 m a.s.l.) om 13 June o 26
Augus 2022, and was o ganized by he PMOD/WRC as pa
o he WMO/GAW p og am (Hülsen and G öbne , 2023).
The QASUMEII da a (see Sec . 2.2) we e used as e e -
ence o he calib a ion o he b oadband adiome e s du ing
he campaign. QASUME and QASUMEII we e equen ly
calib a ed du ing he campaign using a po able calib a ion
sys em wi h 250W lamps. The wo spec o adiome e s e-
mained s able wi hin ±1% o he campaign pe iod and hei
measu emen s di e ed by less han 3%. Se en y- i e sola
UV b oadband il e adiome e s we e shipped o Da os and
pa icipa ed in he campaign. The UVI was de i ed om he
measu emen s o he pa icipa ing ins umen s using he cal-
ib a ion ac o s p o ided by he ope a o s and he calib a ion
ac o s ha we e calcula ed a Da os, and hen he UVI om
he adiome e s was compa ed o he UVI measu ed by QA-
SUMEII. All pa icipa ing ins umen s we e also cha ac e -
ized o hei angula and spec al esponse.
Ancilla y measu emen s o many pa ame e s ha a e alu-
able o he de e mina ion o he ac o s ha esul in disc ep-
ancies be ween he simula ions o UVIOS2 and he measu e-
men s we e pe o med du ing he whole pe iod o he cam-
paign. In pa icula :
–Ae osol op ical p ope ies we e measu ed by a CIMEL
adiome e (and many o he adiome e s ha ope a e a
he si e) ha is pa o he AERONET ne wo k (Holben
e al., 1998).
–To al Ozone Column (TOC) was measu ed by a B ewe
spec o adiome e (Ke , 2010; Ke e al., 1985).
–Global and di ec o al sola i adiance by py anome e s
and a py heliome e .
Figu e 1. Topog aphical map o Da os, Swi ze land.
–Hemisphe ical sky images om sky came as.
–Cloud co e in oc as by a py geome e (Dü and Philo-
pona, 2004)
2.2 QASUME
QASUME is a anspo able spec o adiome e ha is ace-
able o he scale o spec al i adiance es ablished by he
Physikalisch-Technische Bundesans al (PTB) and se es as
a e e ence o spec al sola UV i adiance. The sys em
is main ained by he PMOD/WRC and i s measu emen
accu acy has been imp o ed signi ican ly in he las wo
decades. Since 2014, a second e e ence spec o adiome e
(QASUMEII) is also ope a ing and is used as an addi ional
e e ence s anda d (Hülsen e al., 2016). Upg ades o echni-
cal cha ac e is ics and imp o ed cha ac e iza ion me hodolo-
gies ha e educed he expanded unce ain ies in QASUMEII
measu emen s a wa eleng hs abo e 310 nm om 4.8 % in
2005 ( o QASUME) o 2.0% in 2016 (Hülsen e al., 2016).
Mo e in o ma ion abou QASUME and QASUME II can be
ound in se e al ele an s udies (G öbne e al., 2005, 2006;
G öbne and Spe eld, 2005; Hülsen e al., 2016). Bo h, QA-
SUME and QASUMEII we e measu ing in he ange 290–
420nm wi h a 15 min empo al esolu ion du ing he UVC-
III campaign. These spec a we e weigh ed wi h he e y-
hema ac ion spec um (Webb e al., 2011) and we e hen in-
eg a ed o calcula e he e y hemal doses, and subsequen ly
he UVI (by di iding he dose a es in mWm−2wi h 25).
Fo his wo k we ha e used only he UVI measu ed by QA-
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7455
Table 1. Inpu s o he LUT.
Pa ame e Range Resolu ion
Sola Zeni h Angle (SZA) (°) 1–89 2
To al Column o Ozone (TOC)
(DU)
200–600 10
Ae osol Op ical Dep h (AOD)
a 550nm
0–2 0.1
Single Sca e ing Albedo
(SSA)
0.6–1 0.1
Angs om Exponen (AE) 0–2 0.4
SUMEII, since he ag eemen be ween QASUMEII and QA-
SUME is be e han 3%.
2.3 The UVIOS2 sys em
UVIOS2 is buil upon he UVIOS sys em (Kosmopoulos e
al., 2021). The main change in he sys em con igu a ion el-
a i e o he p e ious e sion is ha he UVI is calcula ed in
wo s eps:
– he UVI is calcula ed unde cloudless skies and
– he e ec o clouds is quan i ied as a second s ep o he
calcula ion o he all-skies UVI.
This change in he sys em’s con igu a ion was accompa-
nied by wo majo modi ica ions/upg ades: (1) he use o a
mo e de ailed UV look up able (LUT) o cloudless-sky cal-
cula ions ha inc eases he accu acy ela i e o he o iginal
e sion, and (2) he use o he UV cloud modi ica ion ac-
o (CMFUV) concep used o he all skies UVI es ima es.
While he spec a o he LUT in he p e ious e sion o he
model we e simula ed using he a las plus mod an ex a e -
es ial spec um (ETS), in he cu en e sion, he mo e e-
cen QASUMEFTS (G öbne e al., 2017) ETS was used.
Fu he mo e, he ozone abso p ion c oss sec ions by Bass
and Pau (1980) we e used o pa ame e ize abso p ion by
ozone, and no he Molina and Molina (1986) ha we e used
in he p e ious e sion. The a iables ha co espond o each
o he i e di e en dimensions o he LUT a e lis ed in Ta-
ble 1, along wi h hei ange and esolu ion. When SZA ex-
ceeds 89°, hen UVI is conside ed equal o 0. When alues o
he o he inpu pa ame e s a e abo e/below he limi s shown
in Table 1, hen inpu s a e se o he uppe /lowe alues o
he used ange. Such occasions a e, howe e , e y a e o
mid-la i ude si es.
The adia i e ans e simula ions o he c ea ion o he
LUT we e pe o med using he UVSPEC model o he li-
bRad an e sion 2.0.4 package (Emde e al., 2016; Maye
and Kylling, 2005). Simula ions we e pe o med using he
Na ional In as uc u es o Resea ch and Technology (GR-
NET) High Pe o mance Compu ing Se ices and he com-
pu a ional esou ces o he ARIS GRNET in as uc u e.
Spec al simula ions pe 0.5nm we e pe o med o he
spec al egion 290–400nm, using he QASUMEFTSETS
(G öbne e al., 2017) and he sdiso sol e (Dahlback and
S amnes, 1991) which assumes pseudosphe ical a mosphe e.
Using a di e en ETS migh esul o di e ences in he sim-
ula ed e y hemal i adiances, as o example was shown in
he s udy o G öbne e al. (2017). Based on he esul s o he
la e s udy we es ima e ha he simula ed i adiances migh
di e by up o 5% i a di e en ETS was used, making he
used ETS spec um a majo unce ain y ac o in UVIOS2
cloudless simula ions. Compa ison wi h he UVIs ha we e
simula ed wi h LUT o he p e ious e sion o he model
(i.e, whe e he a las plus mod an ETS was used o cons uc
he LUT) showed di e ences ha we e in all cases less han
2%. TOC is among he main egula o s o he UVI le els
a he su ace and hus using TOC alues ha ha e been e-
ie ed using di e en ozone abso p ion c oss sec ions ela-
i e o hose ha ha e been used o c ea e he LUTClick o
ap he e o en e ex . would esul in di e ences be ween
he measu ed and he simula ed UVI. Di e ences o 1%–
3% ha e been epo ed in he e ie ed TOC depending on
he used abso p ion c oss sec ions (F agkos e al., 2015; Re-
dondas e al., 2014), which may esul in di e ences o up
o ∼5% in he calcula ed UVI, depending mainly on he
used c oss sec ions, he SZA, ae osols load, and cloudiness
(Blum hale e al., 1995; Kim e al., 2013). In he domain
o which he sys em is commonly used (i.e., Eu ope, No h
A ica, Middle Eas ), a iabili y in SO2and NO2has a mino
impac on he UVI, and hus he o al concen a ion o hese
species has been se o ze o. The US s anda d a mosphe e
(Ande son e al., 1986) was used o desc ibe he p o iles o
a mosphe ic s a e and composi ion, and he su ace albedo
was se o 0.05. Adjus men o he su ace albedo o he local
condi ions when UVIOS is used o e mo e e lec i e e ains
(e.g., dese s, snow-co e ed su aces) is wi hin he model im-
p o emen s ha a e planned o he u u e since unde such
condi ions assuming a s anda d alue o 0.05 could esul in
la ge unce ain ies (e.g., Weihs e al., 2001).
The op ical p ope ies p o iles o libRad an de aul
ae osol model (She le, 1990) we e scaled o he alues o
AOD (spec ally using he co esponding Ångs öm Expo-
nen , AE) and SSA p o ided in Table 1. The AE and SSA
alues ha e been assumed o be in a ian wi h wa eleng h o
he simula ions. Thus, we also did no conside he spec al
dependence o he abso bing ae osol op ical dep h (as e.g.,
in he OMI and TROPOMI algo i hms, A ola e al., 2021),
which may induce inc eased unce ain ies o e pollu ed e-
gions (e.g., Roshan e al., 2020). Howe e , conside ing such
in o ma ion would inc ease signi ican ly he size o he LUT
and hus he compu a ional ime needed o he simula ions,
making he p o ision o he UVI in nea - eal ime o wide
a eas impossible.
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025

7456 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
The UV spec a we e weigh ed wi h he CIE (1998) ac-
ion spec um o he induc ion o e y hema in he human
skin (Webb e al., 2011) o calcula e he UVI. Since all sim-
ula ions ha e been pe o med o he a e age sea-su ace
le el (i.e., al i ude=0m, a mosphe ic p essu e =1013mb).
A co ec ion o he e ec o al i ude, assuming an inc ease
o 5% pe km (e.g., Zempila e al., 2017) has been applied
on he calcula ed UVI.
The cloud op ical hickness (COT) om he Spinning En-
hanced Visible and In a ed Image (SEVIRI) ins umen
aboa d he Me eosa Second Gene a ion (MSG) sa elli es has
been used o calcula e he Cloud Modi ica ion Fac o (CMF).
The COT p oduc is ex ac ed ope a ionally using he EU-
METSAT Sa elli e Applica ion Facili ies o Nowcas ing and
Ve y Sho -Range Fo ecas ing, NWC SAF so wa e package
(De ien and Le Gléau, 2005; Mé éo-F ance, 2016) and he
b oadcas ed MSG da a. A de ailed desc ip ion o he cloud
p oduc s by MSG can be ound in he ele an bibliog aphy
(Deneke e al., 2021; Mé éo-F ance, 2016). Using he MSG
COT alues and he SZA as inpu s o he mul ipa ame ic
equa ions desc ibed in Papach is opoulou e al. (2024) he
sho wa e CMF is calcula ed. Then, i is con e ed o CM-
FUV as desc ibed in S aige e al. (2008). Finally, he UVI is
calcula ed by mul iplying he cloudless-sky alues wi h he
CMFUV. Using wa eleng h dependen CMFUV o simula e
UV would be mo e accu a e (K o ko e al., 2001), bu would
inc ease compu a ional ime, and s ill he dominan unce -
ain y ac o ela ed wi h cloudiness would be he isibili y
o he sola disc.
To e alua e he me hodology used o he quan i ica ion o
he a enua ion o he UVI by clouds he all-sky UVI was
compa ed o QASUMEII measu emen s. Fu he mo e, he
all-sky UVIs we e compa ed o he co esponding alues ha
we e di ec ly simula ed by using cloud op ical p ope ies as
inpu s in he UVSPEC model o libRad an. I was assumed
ha all low-al i ude clouds o e Da os ex end om 4 o 5km
(wi h e e ence o he a.s.l.), and all high-al i ude clouds ex-
end om 7 o 8km. High-al i ude clouds we e in all cases
assumed o consis o ice c ys als wi h e ec i e adius equal
o 20µm and ice wa e con en (IWC) o 0.005g cm−3, while
low-al i ude clouds we e assumed o consis o wa e d ople s
wi h e ec i e adius equal o 10µm and liquid wa e con-
en (LWC) alue o 1 gcm−3. The COT a 550 nm p oduc
om MSG was used as an addi ional inpu , which leads o
an adjus men o he de aul LWC and IWC alues, using he
pa ame e iza ions by Hu and S amnes (1993) o wa e and
by Fu (1996; Fu e al., 1998) o ice clouds. The la e sim-
ula ions we e pe o med o he al i ude o he si e, while all
o he model se ings we e he same as hose used o p oduce
he cloudless LUT. The simula ions ha we e pe o med o
he al i ude o he si e we e also used o e alua e he assump-
ion ha he UVI inc eases by 5% pe km.
P ac ically he e a e wo ways o using UVIOS2. Fo pas
da a using he bes a ailable in o ma ion gi ing p io i y o
g ound based/sa elli e based/modelling based da a in his o -
de o p e e ence. Fo nowcas ing o sho e m o ecas ing
using any exis ing eal ime a ailable da a.
As shown in Table 2 he e a e basic di e ences bu also
common app oaches in he h ee UVI se ices. The main ad-
an age o UVIOS2 is ha i p o ides highe spa io empo al
esolu ion o nowcas ed o pas da a. Ne e heless, i u ilizes
CAMS and TEMIS o ecas s o AOD and ozone nowcas -
s/ o ecas s espec i ely. O e all, all he da a used a e going
h ough libRad an owa ds calcula ing UVI.
2.4 UVIOS2 inpu s
Di e en combina ions o model inpu s ha e been used o
assess he UVIOS2 accu acy when i is used o nowcas ing
and o clima ological analyses. In all cases, he modelled
cloudless-sky UVI alues we e de i ed by in e pola ing lin-
ea ly he elemen s o he 5-dimensional LUT. An o e iew
o he da a ha was used o in e pola e he UVI is p esen ed
in Table 3.
De aul alues o he ae osol asymme y pa ame e (ASY)
and he su ace albedo we e used o he simula ions. Analy-
ses o di e en AERONET da ase s show ha clima ological
ASY a 440nm usually a ies by abou ±0.03 a ound a ypi-
cal alue ha is sligh ly lowe han ∼0.7 (e.g., Foun oulakis
e al., 2019; Kazadzis e al., 2016; Kha i e al., 2016; Rap is
e al., 2018). Gi en ha ASY gene ally dec eases wi h wa e-
leng h i was assumed o be 0.7 in he UV. The eal ASY can
howe e di e occasionally by up o abou ±0.1 (e.g., Foun-
oulakis e al., 2019). We es ima ed ha a di e ence o 0.1
in he asymme y pa ame e can esul in di e ences o up o
∼2% in he simula ed UVI. Using a de aul su ace albedo
(0.05) also in oduces unce ain ies in he modelling o he
UV index. Su ace albedo changes spec ally and i s impac
di e s depending on ae osol load and p ope ies (e.g., Co
e al., 2009; Foun oulakis e al., 2019). Ne e heless, du ing
he snow- ee pe iod a Da os di e ences in su ace albedo
a e es ima ed o be wi hin ±0.03 (e.g., Feis e and G ewe,
1995) esul ing in di e ences ha a e o he o de o a ew
pe cen . Sensi i i y analysis e ealed ha he unce ain y in
he UVI simula ions o AOD ≤0.5 due o he combined e -
ec o using de aul ASY and su ace albedo alues (wi h
e o s o ±0.1 and ±0.03 espec i ely) is less han 3%.
Fo UVI nowcas ing, he ae osol p ope ies and TOC
ha we e used as model inpu s we e ei he o ecas s
(AOD, TOC) o clima ological alues (AE, SSA, ASY).
Speci ically, 1d ahead o ecas s o he TOC om TEMIS
(h ps://www. emis.nl/u adia ion/n /u index.php, las
access: 10 July 2025) and o he AOD a 550 nm om
CAMS (h ps://ads.a mosphe e.cope nicus.eu/cdsapp#!/
da ase /cams-global-a mosphe ic-composi ion- o ecas s?
ab=o e iew, las access: 10 July 2025), as well as mon hly
clima ological alues o he SSA and he AE ( ypical alues
o 0.9 and 1.5, espec i ely, ha e been es ima ed o Da os)
we e used o in e pola e he elemen s o he LUT. To al
ozone 5d o ecas s a e a ailable om TEMIS on a daily
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7457
Table 2. Inpu s o he UVIOS2, TEMIS, and CAMS se ices ha p o ide he UVI.
Pa ame e UVIOS2 TEMIS CAMS
Pas / eanalysis da a
Cloud inpu s Based on MSG Cloud Op ical
hickness
Cloud co ec ion based on
sa elli e da a ( e lec i i y, cloud
co e ).
Dynamic cloud modeling wi h
eal- ime wea he o ecas s.
Spa ial 5km×5km o clouds
13km ×24 km o Ozone
∼80km ×40 km (GOME-2)
o 13km ×24 km (OMI).
0.4°×0.4° (∼44 km ×44km).
Tempo al E e y 15min Daily upda es (based on
sa elli e o e passes).
E e y 1h
Ae osol G ound based measu emen s o
CAMS AOD, based on
a ailabili y a he loca ion
unde s udy
Th ough cloud e lec i i y o
his o ical AOD
ad anced a mosphe ic models
and da a assimila ion om
sa elli e and g ound-based
obse a ions.
To al ozone B ewe i a ailable, mainly
based on OMI
Based on OMI Full a mosphe ic modeling
( anspo +chemis y).
Nowcas / o ecas da a
Cloud inpu s Based on MSG Cloud Op ical
Thickness and cloud mo ion
ec o s ( o o ecas )
No a ailable o ecas s. Only
Cloudless sky UV
Dynamic cloud modeling wi h
eal- ime wea he o ecas s.
Spa ial 5km ×5 km o clouds
13km ×24 km o Ozone
∼80km ×40 km (GOME-2)
o 13km ×24 km (OMI).
0.4°×0.4° (∼44 km ×44km).
Tempo al E e y 15min up o 3h Daily up o 7d E e y 3h up o 5 d
Ae osol Based on CAMS AOD
o ecas s
His o ical AOD CAMS o ecas ing
To al ozone TEMIS o ecas used (p e ious
day)
TEMIS o ecas : Based
on sa elli e obse a ions wi h
some basic ex apola ion
echniques
Uses mul iple sa elli e
sou ces+nume ical models.
Table 3. Combina ions o inpu da a o he UVIOS2 sys em o cloudless sky condi ions. The h ee di e en combina ions used o e alua e
he sys em as a ool o clima ological analysis a e e e ed o as CAMS, CAMS+OMI, GB.
Va iable Nowcas ing I (SAT) Clima ological I
(CAMS)
Clima ological II
(CAMS+OMI)
Nowcas ing II and
Clima ological III (GB)
AOD CAMS o ecas ed
AOD a 550 nm
CAMS eanalysis AOD
a 550nm
CAMS eanalysis AOD
a 550nm
Measu ed AOD a
500nm om CIMEL
TOC Fo ecas ed om
TEMIS
CAMS eanalysis OMI measu ed Measu ed om B ewe
AE Clima ological (1.5) Clima ological (1.5) Clima ological (1.5) Measu ed by CIMEL
(440–675nm)
empo al esolu ion. De ailed desc ip ion o TEMIS and he
a ailable p oduc s can be ound on he se ice web-page
(h ps://www. emis.nl/u adia ion/p oduc /index.php, las
access: 10 July 2025). The CAMS o ecas ed AOD is
a ailable o he ollowing 5 d, on a 1h esolu ion, and he
o ecas s a e upda ed e e y 12h.
Fo he calcula ion o he clima ological cloudless-sky
UVI, he h ee combina ions o inpu s p esen ed in Table 3
we e used:
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025
7458 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
1. Reanalysis TOC and AOD om CAMS (Inness e
al., 2019) ins ead o he co esponding o ecas ed
p oduc s. All o he pa ame e s we e kep he same
as o nowcas ing. The CAMS eanalysis, a ailable
om 2003 onwa ds, is he global eanalysis da ase
o a mosphe ic composi ion o he Eu opean Cen-
e o Medium-RangeWea he Fo ecas s (ECMWF),
consis ing o h ee-dimensional ime-consis en a mo-
sphe ic composi ion ields, including ae osols and
chemical species. I is based on ECMWF’s In eg a ed
Fo ecas Sys em (IFS), wi h se e al upda es o he
ae osol and chemis y modules desc ibed by Inness
e al. (2019). CAMS eanalysis p oduc s a e a ailable
om he Cope nicus A mosphe e Da a S o e (ADS,
h ps://ads.a mosphe e.cope nicus.eu/#!/home (las ac-
cess: 8 Augus 2024) on a 3-hou ly basis on a egu-
la 0.75°×0.75° la i ude/longi ude g id (ins ead o hei
na i e ep esen a ion). This da ase is e e ed o as
“CAMS”.
2. TOC ha has been e ie ed om he Ozone Moni o ing
Ins umen (OMI) aboa d Au a (Le el e al., 2006), and
eanalysis AOD om CAMS (Inness e al., 2019). All
o he pa ame e s we e again kep he same as o now-
cas ing. This da ase is e e ed o as “CAMS+OMI”.
3. TOC measu emen s om he B ewe spec o adiome e
wi h se ial numbe 163 (B ewe #163) (G öbne e al.,
2021), AOD a 500 nm, and AE (440–675nm) om he
CIMEL adiome e (Giles e al., 2019), and all o he pa-
ame e s he same as o he de aul nowcas ing se up.
The AOD and AE ha we e used o he s udy a e le el
1.5, e sion 3 AERONET di ec sun p oduc s. Le el 2
AERONET e ie als we e no used because hey we e
no a ailable a he ime o he analysis. Since hese in-
pu s a e p oduced in nea eal ime, we conside ha
hey could be po en ially used o UVI nowcas ing in
addi ion o clima ological s udies. Le el 2 AERONET
e ie als a e a ailable wi h a longe la ency and can be
used o eanalysis a a la e s age. This da ase is e-
e ed o as “GB”.
The cloudless-sky UVI LUT ou pu s we e in all cases pos
co ec ed o he e ec o he a ying Ea h-Sun dis ance
and o he su ace ele a ion (1596 m o Da os). The all-
sky UVI alues we e de i ed in all cases by mul iplying he
cloudless-sky UVI wi h he Cloud Modi ica ion Fac o in UV
(CMFUV), which was calcula ed as desc ibed in Sec . 2.1
om he MSG-SEVIRI COT.
The UVI was simula ed o he pe iod 1 July–20 Augus
2022 a he ime o he QASUMEII measu emen s (15min
empo al esolu ion). The MSG images, and hus he CM-
FUV, we e a ailable a he exac ime o he UV scans. All
he o he pa ame e s (AOD, TOC, e c) we e in e pola ed lin-
ea ly o he ime o he measu emen s.
Fo he analysis, measu emen s we e classi ied as clea -
sky (i.e., sun was no ully o pa ially co e ed by clouds) and
all-sky (i.e., o all cloudiness condi ions). In he ollowing,
clea -sky condi ions e e o unoccluded sola disc acco d-
ing o measu emen s (al hough clouds may be p esen on he
sky). Cloudless-sky condi ions e e o cloud- ee skies. To
classi y he measu emen s, he di ec componen o he o-
al sola i adiance, as i was measu ed by he py heliome e
ha was ope a ing a Da os du ing he campaign, was sim-
ula ed as desc ibed in Papach is opoulou e al. (2024), and
was hen compa ed o he measu ed di ec i adiance. When
di e ences be ween he wo componen s exceeded 10%, we
conside ed ha he sun was ( ully o pa ially) co e ed by
clouds.
3 Resul s
3.1 Assessmen o UVIOS2 o eal ime applica ions
In his sec ion we ied o assess he accu acy o he modelled
UVI when UVIOS2 is used o eal ime applica ions. Ini-
ially we compa ed he modelled and he measu ed UVI un-
de clea -sky and all-sky condi ions. The UVI was modelled
using he de aul inpu s and se up o he UVIOS2 (SAT), as
well as using high quali y g ound-based measu emen s (GB),
ha heo e ically can be a ailable a nea eal ime o he e-
ie al o a highe accu acy es ima e o he UVI.
3.1.1 Clea -sky UVI
Unde clea -skies, he a io be ween modelled UVI da ase s
and he co esponding measu ed UVI om QASUMEII was
hen calcula ed and he esul s a e shown in Fig. 2.
While he a e age a io is in bo h cases ∼0.99, he s an-
da d de ia ion is high, 0.063 and 0.057 o SAT and GB e-
spec i ely, i.e., only sligh ly lowe o GB. This esul shows
ha using highly accu a e inpu s o TOC, AOD a 500 nm,
and AE does no esul in a no iceable imp o emen in he
accu acy o he a e age modelled clea -sky UVI (s anda d
de ia ion dec eases by only a ew pe cen ), which means ha
o he ac o s a e also impo an o he calcula ion o he su -
ace UVI le els a Da os. The ole o each o he ac o s ha
we e ound o be he mos impo an is discussed in he ol-
lowing.
AOD and TOC
The AOD o ecas ed by CAMS is a a di e en wa eleng h
(550nm) ela i e o he AOD measu ed by he CIMEL
(500nm). To compa e he AOD om he wo di e en
sou ces, he AOD om he CIMEL was ex apola ed a
550nm using he measu ed AE (440–675 nm). The di e -
ences be ween he AOD a 550 nm om he CIMEL and
CAMS a e shown in he Appendix (Fig. A1). Di e ences in
AOD a e in mos cases wi hin ±0.1, wi h an a e age o ∼0,
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7459
Figu e 2. Ra ios be ween simula ed and measu ed clea -sky UVI. Red colou : a ios o simula ions pe o med using o ecas ed CAMS AOD
and TEMIS TOC. Blue colou : a ios o simula ions pe o med using measu ed AOD, AE (by CIMEL), and TOC (by B ewe ). Ra ios ha e
been calcula ed o SZA <75°. Dashed lines ep esen he mean, while do ed lines ep esen he ange o 2 s anda d de ia ion.
which explains di e ences o up o abou ±10 % be ween he
UVIs simula ed using he wo di e en da ase s. When di -
e ences in AOD a e la ge (e.g., in day o he yea (DOY)
201–202 CAMS has no cap u ed he la ge AOD le els o e
he si e and he AOD om CAMS is lowe by 0.15–0.25 el-
a i e o he AOD om CIMEL) hey esul in co espond-
ingly la ge di e ences be ween he a ios (o 10 %–20%).
Di e ences in TOC (Fig. A2) a e gene ally wi hin ±25 DU,
wi h an a e age o abou 4DU (on a e age, TEMIS sligh ly
o e es ima es TOC o he pe iod o he campaign), bu oc-
casionally hey can each ±40DU. Di e ences o ±25DU
in TOC can explain di e ences o abou ±15 % in he UVI
modelled using he wo di e en da ase s (e.g., Kim e al.,
2013). The la ge di e ences be ween he a ios ha we e cal-
cula ed o he wo di e en UVI da ase s in DOY 194 a e
mos ly explained by di e ences in TOC (∼20DU du ing
mos o he day). Di e ences in DOYs 200–204, 206, and
223 a e mos ly explained by di e ences in AOD. The accu-
acy in he g ound-based measu emen s (∼0.02 o he AOD
(Giles e al., 2019) and be e han 2.5% o TOC (e.g., Ca -
lund e al., 2017; Foun oulakis e al., 2019) canno explain
he s anda d de ia ion o 0.057 in he a io be ween he mod-
elled UVI when GB measu emen s a e used as inpu s and he
measu ed UVI.
SSA
While a de aul SSA alue o 0.9 has been used o he sim-
ula ions, he eal SSA a he sho e UV wa eleng hs, which
con ibu e he mos in UVI, can di e signi ican ly, anging
om alues smalle han 0.8 (du ing e.g., e en s o dus , pol-
lu ion o biomass bu ning ae osols ha ha e been ans e ed
o e he si e) o alues exceeding 0.98 (e.g., o mix u es ha
a e domina ed by sul u ic ae osols). As discussed in K o ko
e al. (1998) he SSA has a e y signi ican impac on he
UVI. In hei s udy hey show ha assuming e y abso b-
ing ae osols (SSA=0.6) esul s in abou hal o he UVI
o highly e lec i e ae osols (SSA=0.99) o AOD =1 a
325nm. Con i ming he indings o K o ko e al. (1998), in
Fig. 3 we show ha he sensi i i y o he a ios o he inpu
SSA inc eases wi h inc easing AOD. Fo AOD be ween 0.3
and 0.4 a change o 0.1 (inc ease o dec ease) in SSA esul s
in a change o ∼0.1 in he a io (i.e., ∼10% in he simula ed
UVI) which is o simila magni ude wi h he change in UVI
due o a change o ∼0.1 in AOD. The e ec o changing SSA
becomes less signi ican as he AOD dec eases. Ne e heless,
e en o AOD o ∼0.1, a change o ∼0.1 in he SSA esul s
in a change o 0.05 in he a io (i.e., o ∼5% in he mod-
elled UVI). Gene ally, Figu e 3 deno es ha ae osol mix u es
o e Da os in he summe a e domina ed by ae osols ha a e
weak abso be s o he UV adia ion.
Changing he SSA om 0.8 o 0.99 esul s in mean a io
alues ha a e simila o each o he and close o uni y (see
Fig. A3). Using a simila analysis, K o ko e al. (1998) con-
cluded ha he SSA ha ga e he smalles g adien wi h AOD
is mo e ep esen a i e o To on o. Howe e , in ou case, he
analysis o he SSA a 440nm om AERONET o he pe-
iod o he campaign esul s in an a e age alue below 0.95,
and since o mo e ae osol species he SSA inc eases wi h
dec easing wa eleng h, we es ima e ha an SSA equal o 0.9
is mo e app op ia e o model he UVI a Da os.
The high alues o he a ios be ween modelled and mea-
su ed UVIs in DOY 197–199 can be possibly jus i ied by
eal SSA alues ha a e lowe han 0.9, and hus assum-
ing SSA=0.9 o he simula ions esul s in an o e es ima-
ion o he UVI. In hese days, he SSA a 440 nm om
AERONET was gene ally lowe han 0.9 ( alues be ween
0.77 and 0.92). As shown in Fig. 4, he low SSA alues may
be due o pollu ed ai masses o igina ing om low al i udes
o e Ge many. Du ing DOY 200–210 when a nega i e bias
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025
7466 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
Appendix A
Figu e A1. Di e ences be ween he 550 nm AOD om CAMS and he CIMEL (CAMS-CIMEL).
Figu e A2. Di e ences be ween he TOC om TEMIS and he B ewe (TEMIS-B ewe ).
Figu e A3. Densi y plo s o he a io be ween simula ed (using GB measu emen s) and measu ed clea -sky UVI when di e en SSA alues
a e used o he simula ions. Ve ical lines ep esen he mean and he 1σand 2σin e als i a no mal dis ibu ion ( ed line) is assumed.
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025

I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7467
Code and da a a ailabili y. Analy ical desc ip ion and ins uc ions
on how o access he model a e p o ided in Kosmopoulos e al.
(2021). All codes and da ase s ha a e necessa y o ep oduce he
esul s used in his pape a e a chi ed on Zenodo (Foun oulakis,
2025): h ps://doi.o g/10.5281/zenodo.16781118.
Au ho con ibu ions. Concep ualiza ion: SK, IF. Me hodology:
SK, IF and KP. Fo mal analysis: IF, KP, JG, GH, and DK. So -
wa e: IF, KP, and I-PR. Valida ion: IF, KP, SK, GH, and JG. In es-
iga ion: IF and KP. Resou ces: SK and CK. Da a cu a ion: IF, KP,
NK, JG and GH. Visualiza ion: IF, KP and GH. W i ing (o iginal
d a p epa a ion): IF, KP, and AM. Supe ision: SK. W i ing ( e-
iew and edi ing): IF, KP, SK, JG, GH, I-PR, DK, AM, CK, and CZ.
All au ho s ga e inal app o al o publica ion.
Compe ing in e es s. The con ac au ho has decla ed ha none o
he au ho s has any compe ing in e es s.
Disclaime . Publishe ’s no e: Cope nicus Publica ions emains
neu al wi h ega d o ju isdic ional claims made in he ex , pub-
lished maps, ins i u ional a ilia ions, o any o he geog aphical ep-
esen a ion in his pape . While Cope nicus Publica ions makes e -
e y e o o include app op ia e place names, he inal esponsibili y
lies wi h he au ho s. Also, please no e ha his pape has no e-
cei ed English language copy-edi ing. Views exp essed in he ex
a e hose o he au ho s and do no necessa ily e lec he iews o
he publishe .
Acknowledgemen s. This wo k was suppo ed by compu a ional
ime g an ed by he Na ional In as uc u es o Resea ch and Tech-
nology S.A. (GRNET) a he Na ional HPC acili y – ARIS –
unde p ojec ID pa210301-SO-LISIS. The Eu opean Commis-
sion p ojec “EXCELSIOR”: ERATOSTHENES: Excellence Re-
sea ch Cen e o Ea h Su eillance and Space-Based Moni o ing
o he En i onmen (g an no. 857510) is also acknowledged. C.
S. Ze e os would like o acknowledge CAMS2_82 P ojec : E al-
ua ion and Quali y Con ol (EQC) o global p oduc s. D. Kouk-
laki would like o acknowledge he PANGEA4CalVal p ojec (g an
no. 101079201) unded by he Eu opean Union.
S. Kazadzis acknowledges ACTRIS-CH (Ae osol, Clouds and
T ace Gases Resea ch In as uc u e – Swiss con ibu ion), unded
by he S a e Sec e a ia o Educa ion, Resea ch and Inno a ion.
SK, IF, KP would like o acknowledge HARMONIA (In e na-
ional ne wo k o ha moniza ion o a mosphe ic ae osol e ie als
om g ound-based pho ome e s; g an no. CA21119), suppo ed by
COST (Eu opean Coope a ion in Science and Technology).
Financial suppo . This p ojec has ecei ed unding om he Eu-
opean Union’s Ho izon 2020 esea ch and inno a ion p og amme
EIFFEL unde g an ag eemen no. 101003518.
Re iew s a emen . This pape was edi ed by Xiaohong Liu and e-
iewed by h ee anonymous e e ees.
Re e ences
Allaa , M., an Weele, M., Fo uin, P., and Kelde , H.: An empi -
ical model o p edic he UV-index based on sola zeni h an-
gles and o al ozone, Me eo ological Applica ions, 11, 59–65,
h ps://doi.o g/10.1017/S1350482703001130, 2004.
Ande son, G., Clough, S., Kneizys, F., Che wynd, J., and She -
le, E.: AFGL A mosphe ic Cons i uen P o iles (0.120km), Ai
Fo ce Geophysics Labo a o y, HANSCOM AFB, MA, 01731, 46
pp., h ps://apps.d ic.mil/s i/ /pd /ADA175173.pd (las access:
10 Oc obe 2025), 1986.
A ms ong, B. K. and Cus , A. E.: Sun exposu e and skin cance ,
and he puzzle o cu aneous melanoma: A pe spec i e on Fea s e
al. Ma hema ical models o age and ul a iole e ec s on he in-
cidence o skin cance among whi es in he Uni ed S a es, Ame -
ican Jou nal o Epidemiology 1977, Cance Epidemiol., 48, 147–
156, h ps://doi.o g/10.1016/j.canep.2017.04.004, 2017.
A ola, A., Kallisko a, S., den Ou e , P. N., Ed a dsen, K., Hansen,
G., Koskela, T., Ma in, T. J., Ma hijsen, J., Mee koe e , R.,
Pee e s, P., Seckmeye , G., Simon, P. C., Slape , H., Taalas,
P., and Ve debou , J.: Assessmen o ou me hods o es ima e
su ace UV adia ion using sa elli e da a, by compa ison wi h
g ound measu emen s om ou s a ions in Eu ope, Jou nal o
Geophysical Resea ch: A mosphe es, 107, ACL 11-1–ACL 11-
11, h ps://doi.o g/10.1029/2001JD000462, 2002.
A ola, A., Wandji Nyamsi, W., Lipponen, A., Kazadzis, S.,
K o ko , N. A., and Tamminen, J.: Re hinking he co ec-
ion o abso bing ae osols in he OMI- and TROPOMI-like
su ace UV algo i hms, A mos. Meas. Tech., 14, 4947–4957,
h ps://doi.o g/10.5194/am -14-4947-2021, 2021.
De Backe , H., Koepke, P., Bais, A., de Cabo, X., F ei, T.,
Gillo ay, D., Hai e, C., Heikkilä, A., Kazan zidis, A., Koskela,
T., Ky ö, E., Lape a, B., Lo en e, J., Masson, K., Maye ,
B., Ple s, H., Redondas, A., Renaud, A., Schaube ge , G.,
Schmalwiese , A., Schwande , H., and Vanicek, K.: Compa -
ison o measu ed and modelled u indices o he assess-
men o heal h isks, Me eo ological Applica ions, 8, 267–277,
h ps://doi.o g/10.1017/S1350482701003024, 2001.
Bais, A. F., Ga dine , B. G., Slape , H., Blum hale , M., Be n-
ha d, G., McKenzie, R., Webb, A. R., Seckmeye , G., Kjeld-
s ad, B., Koskela, T., Ki sch, P. J., G öbne , J., Ke , J.
B., Kazadzis, S., Leszczynski, K., Wa dle, D., Jose sson,
W., B ogniez, C., Gillo ay, D., Reinen, H., Weihs, P., S e-
noe, T., E iksen, P., Kuik, F., and Redondas, A.: SUS-
PEN in e compa ison o ul a iole spec o adiome e s, Jou -
nal o Geophysical Resea ch: A mosphe es, 106, 12509–12525,
h ps://doi.o g/10.1029/2000JD900561, 2001.
Bais, A. F., Be nha d, G., McKenzie, R. L., Aucamp, P. J.,
Young, P. J., Ilyas, M., Jöckel, P., and Deushi, M.: Ozone–
clima e in e ac ions and e ec s on sola ul a iole adia-
ion, Pho ochemical & Pho obiological Sciences, 18, 602–640,
h ps://doi.o g/10.1039/C8PP90059K, 2019.
Bass, A. M. and Pau , R. J.: Abso p ion C oss-Sec ion o Ozone
as Func ion o Tempe a u e, in: Topical Mee ing on Spec-
oscopy in Suppo o A mosphe ic Measu emen s, Technical
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025
7468 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
Diges Se ies (Op ica Publishing G oup, 1980), pape WP15,
h ps://doi.o g/10.1364/SAM.1980.WP15, 1980.
Be nha d, G., Boo h, C. R., and Eh amjian, J. C.: Compa ison o
UV i adiance measu emen s a Summi , G eenland; Ba ow,
Alaska; and Sou h Pole, An a c ica, A mos. Chem. Phys., 8,
4799–4810, h ps://doi.o g/10.5194/acp-8-4799-2008, 2008.
Be nha d, G., A ola, A., Dahlback, A., Fiole o , V., Heikkilä,
A., Johnsen, B., Koskela, T., Lakkala, K., S endby, T., and
Tamminen, J.: Compa ison o OMI UV obse a ions wi h
g ound-based measu emen s a high no he n la i udes, A mos.
Chem. Phys., 15, 7391–7412, h ps://doi.o g/10.5194/acp-15-
7391-2015, 2015.
Be nha d, G. H., Bais, A. F., Aucamp, P. J., Klekociuk, A. R., Liley,
J. B., and McKenzie, R. L.: S a osphe ic ozone, UV adia ion,
and clima e in e ac ions, Pho ochemical & Pho obiological Sci-
ences, 22, 937–989, h ps://doi.o g/10.1007/s43630-023-00371-
y, 2023.
Blum hale , M.: UV Moni o ing o Public Heal h, In . J. En i-
on. Res. Public Heal h, h ps://doi.o g/10.3390/ije ph15081723,
2018.
Blum hale , M., Webb, A. R., Seckmeye , G., Bais, A. F., Hube , M.,
and Maye , B.: Simul aneous spec o adiome y: A s udy o sola
UV i adiance a wo al i udes, Geophys. Res. Le ., 21, 2805–
2808, h ps://doi.o g/10.1029/94GL02786, 1994.
Blum hale , M., Salzgebe , M., and Ambach, W.: Ozone and
Ul a iole -B i adiances: expe imen al de e mina ion o he a-
dia ion ampli ica ion ac o , Pho ochem Pho obiol, 61, 159–162,
h ps://doi.o g/10.1111/j.1751-1097.1995. b03954.x, 1995.
Blum hale , M., Ambach, W., and Ellinge , R.: Inc ease in sola UV
adia ion wi h al i ude, J. Pho ochem. Pho obiol. B, 39, 130–134,
h ps://doi.o g/10.1016/S1011-1344(96)00018-8, 1997.
B ogniez, C., Au iol, F., De oo, C., A ola, A., Kujanpää, J.,
Sau age, B., Kalakoski, N., Pi känen, M. R. A., Ca al amo,
M., Me zge , J.-M., Tou nois, G., and Da Conceicao, P.: Vali-
da ion o sa elli e-based noon ime UVI wi h NDACC g ound-
based ins umen s: in luence o opog aphy, en i onmen and
sa elli e o e pass ime, A mos. Chem. Phys., 16, 15049–15074,
h ps://doi.o g/10.5194/acp-16-15049-2016, 2016.
Caldwell, M. M., Bjö n, L. O., Bo nman, J. F., Flin , S.
D., Kulandai elu, G., Te amu a, A. H., and Te ini, M.:
E ec s o inc eased sola ul a iole adia ion on e es-
ial ecosys ems, J. Pho ochem. Pho obiol. B, 46, 40–52,
h ps://doi.o g/10.1016/S1011-1344(98)00184-5, 1998.
Ca lund, T., Kou eme i, N., Kazadzis, S., and G öbne , J.: Ae osol
op ical dep h de e mina ion in he UV using a ou -channel
p ecision il e adiome e , A mos. Meas. Tech., 10, 905–923,
h ps://doi.o g/10.5194/am -10-905-2017, 2017.
Casale, G. R., Siani, A. M., Diémoz, H., Agnesod, G., Pa isi,
A. V., and Colosimo, A.: Ex eme UV index and sola expo-
su es a Pla eau Rosà (3500m a.s.l.) in Valle d’Aos a Region,
I aly, Science o The To al En i onmen , 512–513, 622–630,
h ps://doi.o g/10.1016/j.sci o en .2015.01.049, 2015.
Chuba o a, N. and Zhdano a, Y.: The assessmen
o UV esou ces o e No he n Eu asia, 764–767,
h ps://doi.o g/10.1063/1.4804882, 2013.
Chuba o a, N., Zhdano a, Y., and Nez al, Y.: A new pa ame-
e iza ion o he UV i adiance al i ude dependence o clea -
sky condi ions and i s applica ion in he on-line UV ool
o e No he n Eu asia, A mos. Chem. Phys., 16, 11867–11881,
h ps://doi.o g/10.5194/acp-16-11867-2016, 2016.
CIE S: CIE S007/E-1998 E y hema e e ence ac ion
spec um and s anda d e y hema dose, Colo Res.
Appl., 24, 158–158, h ps://doi.o g/10.1002/(SICI)1520-
6378(199904)24:2<158::AID-COL11>3.0.CO;2-4, 1998.
Co , C. A., K o ko , N., Mad onich, S., Slusse , J. R., Holben, B.,
Gao, W., Flynn, J., Le e , B., and K eidenweis, S. M.: Re ie al
o ae osol single sca e ing albedo a ul a iole wa eleng hs a
he T1 si e du ing MILAGRO, A mos. Chem. Phys., 9, 5813–
5827, h ps://doi.o g/10.5194/acp-9-5813-2009, 2009.
Dahlback, A. and S amnes, K.: A new sphe ical model
o compu ing he adia ion ield a ailable o pho olysis
and hea ing a wiligh , Plane Space Sci., 39, 671–683,
h ps://doi.o g/10.1016/0032-0633(91)90061-E, 1991.
Dahlback, A., Gelso , N., S amnes, J. J., and Gjessing, Y.:
UV measu emen s in he 3000–5000m al i ude egion in
Tibe , Jou nal o Geophysical Resea ch: A mosphe es, 112,
h ps://doi.o g/10.1029/2006JD007700, 2007.
Deneke, H., Ba ien os-Velasco, C., Bley, S., Hüne bein, A., Lenk,
S., Macke, A., Mei ink, J. F., Sch oed e -Homscheid , M., Sen ,
F., Wang, P., We ne , F., and Wi huhn, J.: Inc easing he spa-
ial esolu ion o cloud p ope y e ie als om Me eosa SE-
VIRI by use o i s high- esolu ion isible channel: imple-
men a ion and examples, A mos. Meas. Tech., 14, 5107–5126,
h ps://doi.o g/10.5194/am -14-5107-2021, 2021.
De ien, M. and Le Gléau, H.: MSG/SEVIRI cloud mask and
ype om SAFNWC, In . J. Remo e Sens., 26, 4707–4732,
h ps://doi.o g/10.1080/01431160500166128, 2005.
Di ey, B. L.: Sola ul a iole adia ion e ec s on bi-
ological sys ems, Phys. Med. Biol., 36, 299–328,
h ps://doi.o g/10.1088/0031-9155/36/3/001, 1991.
Dü , B. and Philipona, R.: Au oma ic cloud amoun
de ec ion by su ace longwa e downwa d adia-
ion measu emen s, J. Geophys. Res., 109, D05201,
h ps://doi.o g/10.1029/2003JD004182, 2004.
D o kin, A. Y. and S einbe ge , E. H.: Modeling he al i ude
e ec on sola UV adia ion, Sola Ene gy, 65, 181–187,
h ps://doi.o g/10.1016/S0038-092X(98)00126-1, 1999.
Emde, C., Bu as-Schnell, R., Kylling, A., Maye , B., Gas eige , J.,
Hamann, U., Kylling, J., Rich e , B., Pause, C., Dowling, T.,
and Buglia o, L.: The libRad an so wa e package o adia-
i e ans e calcula ions ( e sion 2.0.1), Geosci. Model De ., 9,
1647–1672, h ps://doi.o g/10.5194/gmd-9-1647-2016, 2016.
E ickson III, D. J., Sulzbe ge , B., Zepp, R. G., and Aus in, A.
T.: E ec s o s a osphe ic ozone deple ion, sola UV adia ion,
and clima e change on biogeochemical cycling: in e ac ions and
eedbacks, Pho ochemical & Pho obiological Sciences, 14, 127–
148, h ps://doi.o g/10.1039/C4PP90036G, 2015.
Feis e , U. and G ewe, R.: Spec al albedo measu emen s in he
u and isible egion o e di e en ypes o su aces, Pho-
ochem. Pho obiol., 62, 736–744, h ps://doi.o g/10.1111/j.1751-
1097.1995. b08723.x, 1995.
Feis e , U., Laschewski, G., and G ewe, R.-D.: UV in-
dex o ecas s and measu emen s o heal h-e ec i e
adia ion, J. Pho ochem. Pho obiol. B, 102, 55–68,
h ps://doi.o g/10.1016/j.jpho obiol.2010.09.005, 2011.
Fiole o , V. E., Kimlin, M. G., K o ko , N., McA hu , L. J. B.,
Ke , J. B., Wa dle, D. I., He man, J. R., Mel ze , R., Ma h-
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7469
ews, T. W., and Kau ola, J.: UV index clima ology o e he
Uni ed S a es and Canada om g ound-based and sa elli e es-
ima es, Jou nal o Geophysical Resea ch: A mosphe es, 109,
h ps://doi.o g/10.1029/2004JD004820, 2004.
Foun oulakis, I.: Da a and code o “Assessmen o he
accu acy in UV index modelling using he UVIOS2
sys em du ing he UVC-III campaign”, Zenodo [code],
h ps://doi.o g/10.5281/zenodo.16781118, 2025.
Foun oulakis, I., Siomos, N., Na si, A., D osoglou, T., and Bais, A.
F.: De i ing Ae osol Abso p ion P ope ies om Sola Ul a i-
ole Radia ion Spec al Measu emen s a Thessaloniki, G eece,
h ps://doi.o g/10.3390/ s11182179, 2019.
Foun oulakis, I., Diémoz, H., Siani, A. M., Hülsen, G., and
G öbne , J.: Moni o ing o sola spec al ul a iole i adi-
ance in Aos a, I aly, Ea h Sys . Sci. Da a, 12, 2787–2810,
h ps://doi.o g/10.5194/essd-12-2787-2020, 2020a.
Foun oulakis, I., Diémoz, H., Siani, A.-M., Laschewski, G., Fil-
ippa, G., A ola, A., Bais, A. F., De Backe , H., Lakkala, K.,
Webb, A. R., De Bock, V., Ka ppinen, T., Ga ane, K., Kap-
somenakis, J., Koukouli, M.-E., and Ze e os, C. S.: Sola UV
I adiance in a Changing Clima e: T ends in Eu ope and he
Signi icance o Spec al Moni o ing in I aly, En i onmen s,
h ps://doi.o g/10.3390/en i onmen s7010001, 2020b.
F agkos, K., Bais, A. F., Balis, D., Mele i, C., and Koukouli, M.
E.: The E ec o Th ee Di e en Abso p ion C oss-Sec ions and
hei Tempe a u e Dependence on To al Ozone Measu ed by a
Mid-La i ude B ewe Spec opho ome e , A mosphe e-Ocean,
53, 19–28, h ps://doi.o g/10.1080/07055900.2013.847816,
2015.
F agkos, K., Foun oulakis, I., Cha alampous, G., Pa-
pach is opoulou, K., Nisan zi, A., Hadjimi sis, D., and
Kazadzis, S.: Twen y-Yea Clima ology o Sola UV and
PAR in Cyp us: In eg a ing Sa elli e Ea h Obse a ions
wi h Radia i e T ans e Modeling, Remo e Sens (Basel), 16,
h ps://doi.o g/10.3390/ s16111878, 2024.
Fu, Q.: An Accu a e Pa ame e iza ion o he Sola Ra-
dia i e P ope ies o Ci us Clouds o Clima e Mod-
els, J. Clima e, 9, 2058–2082, h ps://doi.o g/10.1175/1520-
0442(1996)009<2058:AAPOTS>2.0.CO;2, 1996.
Fu, Q., Yang, P., and Sun, W. B.: An Accu a e Pa-
ame e iza ion o he In a ed Radia i e P ope -
ies o Ci us Clouds o Clima e Models, J. Cli-
ma e, 11, 2223–2237, h ps://doi.o g/10.1175/1520-
0442(1998)011<2223:AAPOTI>2.0.CO;2, 1998.
Ga ane, K., Bais, A. F., Kazadzis, S., Kazan zidis, A., and Mele i,
C.: Moni o ing o UV spec al i adiance a Thessaloniki (1990–
2005): da a e-e alua ion and quali y con ol, Ann. Geophys., 24,
3215–3228, h ps://doi.o g/10.5194/angeo-24-3215-2006, 2006.
Giles, D. M., Sinyuk, A., So okin, M. G., Scha e , J. S., Smi no ,
A., Slu ske , I., Eck, T. F., Holben, B. N., Lewis, J. R., Campbell,
J. R., Wel on, E. J., Ko kin, S. V., and Lyapus in, A. I.: Ad ance-
men s in he Ae osol Robo ic Ne wo k (AERONET) Ve sion 3
da abase – au oma ed nea - eal- ime quali y con ol algo i hm
wi h imp o ed cloud sc eening o Sun pho ome e ae osol op-
ical dep h (AOD) measu emen s, A mos. Meas. Tech., 12, 169–
209, h ps://doi.o g/10.5194/am -12-169-2019, 2019.
Go, S., Kim, J., Mok, J., I ie, H., Yoon, J., To es, O., K o ko ,
N. A., Labow, G., Kim, M., Koo, J.-H., Choi, M., and
Lim, H.: G ound-based e ie als o ae osol column abso p-
ion in he UV spec al egion and hei implica ions o
GEMS measu emen s, Remo e Sens. En i on., 245, 111759,
h ps://doi.o g/10.1016/j. se.2020.111759, 2020.
G öbne , J. and Spe eld, P.: Di ec aceabili y o he po able QA-
SUME i adiance scale o he p ima y i adiance s anda d o
he PTB, Me ologia, 42, 134–139, h ps://doi.o g/10.1088/0026-
1394/42/2/008, 2005.
G öbne , J., Sch ede , J., Kazadzis, S., Bais, A. F., Blum hale ,
M., Gö s, P., Tax, R., Koskela, T., Seckmeye , G., Webb,
A. R., and Rembges, D.: T a eling e e ence spec o adiome-
e o ou ine quali y assu ance o spec al sola ul a i-
ole i adiance measu emen s, Appl. Op ., 44, 5321–5331,
h ps://doi.o g/10.1364/AO.44.005321, 2005.
G öbne , J., Blum hale , M., Kazadzis, S., Bais, A., Webb, A.,
Sch ede , J., Seckmeye , G., and Rembges, D.: Quali y assu ance
o spec al sola UV measu emen s: esul s om 25 UV moni-
o ing si es in Eu ope, 2002 o 2004, Me ologia, 43, S66–S71,
h ps://doi.o g/10.1088/0026-1394/43/2/s14, 2006.
G öbne , J., K öge , I., Egli, L., Hülsen, G., Riechelmann, S.,
and Spe eld, P.: The high- esolu ion ex a e es ial sola spec-
um (QASUMEFTS) de e mined om g ound-based sola i -
adiance measu emen s, A mos. Meas. Tech., 10, 3375–3383,
h ps://doi.o g/10.5194/am -10-3375-2017, 2017.
G öbne , J., Schill, H., Egli, L., and S übi, R.: Consis ency
o o al column ozone measu emen s be ween he B ewe
and Dobson spec o adiome e s o he LKO A osa and
PMOD/WRC Da os, A mos. Meas. Tech., 14, 3319–3331,
h ps://doi.o g/10.5194/am -14-3319-2021, 2021.
Häde , D.-P.: E ec s o enhanced sola ul a iole adia ion
on aqua ic ecosys ems, in: Biophysics o pho o ecep o s
and pho omo emen s in mic oo ganisms, Sp inge , 157–172,
h ps://doi.o g/10.1007/978-1-4684-5988-3_12, 1991.
Häde , D.-P., Kuma , H. D., Smi h, R. C., and Wo es , R. C.: E ec s
on aqua ic ecosys ems, J. Pho ochem. Pho obiol. B, 46, 53–68,
h ps://doi.o g/10.1016/S1011-1344(98)00185-7, 1998.
He man, J. R., K o ko , N., Cela ie , E., La ko, D., and
Labow, G.: Dis ibu ion o UV adia ion a he Ea h’s su -
ace om TOMS-measu ed UV-backsca e ed adiances, Jou -
nal o Geophysical Resea ch: A mosphe es, 104, 12059–12076,
h ps://doi.o g/10.1029/1999JD900062, 1999.
Ho mann, G. and Me e , H.: Appa en con adic ion be ween
nega i e e ec s o UV adia ion and posi i e e ec s o sun ex-
posu e, Ge . Med. Sci., 3, Doc01, h ps://pubmed.ncbi.nlm.nih.
go /19675718/ (las access: 10 Oc obe 2025), 2005.
Holben, B. N., Eck, T. F., Slu ske , I., Tan é, D., Buis, J. P., Se -
ze , A., Ve mo e, E., Reagan, J. A., Kau man, Y. J., Naka-
jima, T., La enu, F., Jankowiak, I., and Smi no , A.: AERONET
– A Fede a ed Ins umen Ne wo k and Da a A chi e o
Ae osol Cha ac e iza ion, Remo e Sens. En i on., 66, 1–16,
h ps://doi.o g/10.1016/S0034-4257(98)00031-5, 1998.
Hu, Y. X. and S amnes, K.: An Accu a e Pa ame e iza ion o he Ra-
dia i e P ope ies o Wa e Clouds Sui able o Use in Clima e
Models, J. Clima e, 6, 728–742, h ps://doi.o g/10.1175/1520-
0442(1993)006<0728:AAPOTR>2.0.CO;2, 1993.
Hülsen, G. and G öbne , J.: Cha ac e iza ion and calib a-
ion o ul a iole b oadband adiome e s measu ing e y-
hemally weigh ed i adiance, Appl. Op ., 46, 5877–5886,
h ps://doi.o g/10.1364/AO.46.005877, 2007.
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025
7470 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
Hülsen, G. and G öbne , J.: Repo o he Thi d In e na ional So-
la UV Radiome e Calib a ion Campaign (UVC-III), GAW e-
po No. 284, h ps://lib a y.wmo.in /idu l/4/68642 (las access:
10 July 2025), 2023.
Hülsen, G., G öbne , J., Ne as, S., Spe eld, P., Egli, L., Po o ec-
chio, G., and Smid, M.: T aceabili y o sola UV measu emen s
using he Qasume e e ence spec o adiome e , Appl. Op ., 55,
7265–7275, h ps://doi.o g/10.1364/AO.55.007265, 2016.
Hülsen, G., G öbne , J., Bais, A., Blum hale , M., Diémoz, H., Bol-
sée, D., Diaz, A., Foun oulakis, I., Na anen, E., Sch ede , J.,
S e ania, F., and Gue e o, J. M. V.: Second sola ul a iole a-
diome e compa ison campaign UVC-II, Me ologia, 57, 35001,
h ps://doi.o g/10.1088/1681-7575/ab74e5, 2020.
Inness, A., Ades, M., Agus í-Pana eda, A., Ba é, J., Benedic-
ow, A., Blechschmid , A.-M., Dominguez, J. J., Engelen, R.,
Eskes, H., Flemming, J., Huijnen, V., Jones, L., Kipling, Z.,
Massa , S., Pa ing on, M., Peuch, V.-H., Razinge , M., Remy,
S., Schulz, M., and Su ie, M.: The CAMS eanalysis o a -
mosphe ic composi ion, A mos. Chem. Phys., 19, 3515–3556,
h ps://doi.o g/10.5194/acp-19-3515-2019, 2019.
Juzeniene, A. and Moan, J.: Bene icial e ec s o UV adia ion o he
han ia i amin D p oduc ion, De ma oendoc inol., 4, 109–117,
h ps://doi.o g/10.4161/de m.20013, 2012.
Juzeniene, A., B ekke, P., Dahlback, A., Ande sson-Engels, S.,
Reich a h, J., Moan, K., Holick, M. F., G an , W. B., and
Moan, J.: Sola adia ion and human heal h, Repo s on
P og ess in Physics, 74, 66701, h ps://doi.o g/10.1088/0034-
4885/74/6/066701, 2011.
Kazadzis, S., Bais, A., A ola, A., K o ko , N., Kou eme i, N.,
and Mele i, C.: Ozone Moni o ing Ins umen spec al UV i a-
diance p oduc s: compa ison wi h g ound based measu emen s
a an u ban en i onmen , A mos. Chem. Phys., 9, 585–594,
h ps://doi.o g/10.5194/acp-9-585-2009, 2009.
Kazadzis, S., Rap is, P., Kou eme i, N., Ami idis, V., A ola, A.,
Ge asopoulos, E., and Schus e , G. L.: Ae osol abso p ion e-
ie al a ul a iole wa eleng hs in a complex en i onmen , A -
mos. Meas. Tech., 9, 5997–6011, h ps://doi.o g/10.5194/am -9-
5997-2016, 2016.
Ke , J. B.: The B ewe Spec opho ome e BT – UV Radia ion in
Global Clima e Change: Measu emen s, Modeling and E ec s
on Ecosys ems, edi ed by: Gao, W., Slusse , J. R., and Schmold ,
D. L., Sp inge Be lin Heidelbe g, Be lin, Heidelbe g, 160–191,
h ps://doi.o g/10.1007/978-3-642-03313-1_6, 2010.
Ke , J. B., McEl oy, C. T., Wa dle, D. I., Ola son, R. A., and
E ans, W. F. J.: The Au oma ed B ewe Spec opho ome e , in:
A mosphe ic Ozone, edi ed by: Ze e os, C. S. and Ghazi, A.,
Sp inge , Do d ech , h ps://doi.o g/10.1007/978-94-009-5313-
0_80, 1985.
Kha i, P., Takamu a, T., Nakajima, T., Es ellés, V., I ie, H., Kuze,
H., Campanelli, M., Sinyuk, A., Lee, S. -M., Sohn, B. J., Pan-
di hu ai, G., Kim, S. -W., Yoon, S. C., Ma inez-Lozano, J. A.,
Hashimo o, M., De a a, P. C. S., and Manago, N.: Fac o s o
inconsis en ae osol single sca e ing albedo be ween SKYNET
and AERONET, Jou nal o Geophysical Resea ch: A mosphe es,
121, 1859–1877, h ps://doi.o g/10.1002/2015JD023976, 2016.
Kim, J., Cho, H.-K., Mok, J., Yoo, H. D., and Cho, N.:
E ec s o ozone and ae osol on su ace UV adia-
ion a iabili y, J. Pho ochem. Pho obiol. B, 119, 46–51,
h ps://doi.o g/10.1016/j.jpho obiol.2012.11.007, 2013.
Kinne, S.: The MAC 2 ae osol clima ology, Tellus
B: Chemical and Physical Me eo ology, 71, 1–21,
h ps://doi.o g/10.1080/16000889.2019.1623639, 2019.
Kosmopoulos, P., Kou ou sidis, D., Papach is opoulou, K., Rap-
is, P. I., Masoom, A., Sain -D enan, Y.-M., Blanc, P., Kon oes,
C., and Kazadzis, S.: Sho -Te m Fo ecas ing o La ge-Scale
Clouds Impac on Downwelling Su ace Sola I adia ion, En-
e gies, h ps://doi.o g/10.3390/en13246555, 2020.
Kosmopoulos, P. G., Kazadzis, S., Schmalwiese , A. W., Rap is, P.
I., Papach is opoulou, K., Foun oulakis, I., Masoom, A., Bais,
A. F., Bilbao, J., Blum hale , M., K eu e , A., Siani, A. M., Ele -
he a os, K., Topaloglou, C., G öbne , J., Johnsen, B., S endby, T.
M., Vilaplana, J. M., Dopple , L., Webb, A. R., Khazo a, M., De
Backe , H., Heikkilä, A., Lakkala, K., Ja oslawski, J., Mele i, C.,
Diémoz, H., Hülsen, G., Klo z, B., Rimme , J., and Kon oes, C.:
Real- ime UV index e ie al in Eu ope using Ea h obse a ion-
based echniques: sys em desc ip ion and quali y assessmen , A -
mos. Meas. Tech., 14, 5657–5699, h ps://doi.o g/10.5194/am -
14-5657-2021, 2021.
K o ko , N. A., Bha ia, P. K., He man, J. R., Fiole o , V., and
Ke , J.: Sa elli e es ima ion o spec al su ace UV i adiance in
he p esence o oposphe ic ae osols: 1. Cloud- ee case, Jou -
nal o Geophysical Resea ch: A mosphe es, 103, 8779–8793,
h ps://doi.o g/10.1029/98JD00233, 1998.
K o ko , N. A., He man, J. R., Bha ia, P. K., Fiole o , V., and
Ahmad, Z.: Sa elli e es ima ion o spec al su ace UV i a-
diance: 2. E ec s o homogeneous clouds and snow, Jou -
nal o Geophysical Resea ch: A mosphe es, 106, 11743–11759,
h ps://doi.o g/10.1029/2000JD900721, 2001.
Kylling, A., Albold, A., and Seckmeye , G.: T ansmi ance
o a cloud is wa eleng h-dependen in he UV- ange:
Physical in e p e a ion, Geophys. Res. Le ., 24, 397–400,
h ps://doi.o g/10.1029/97GL00111, 1997.
Lakkala, K., A ola, A., Heikkilä, A., Kau ola, J., Koskela, T., Ky ö,
E., Lind o s, A., Meinande , O., Tanskanen, A., G öbne , J.,
and Hülsen, G.: Quali y assu ance o he B ewe spec al UV
measu emen s in Finland, A mos. Chem. Phys., 8, 3369–3383,
h ps://doi.o g/10.5194/acp-8-3369-2008, 2008.
Lakkala, K., Kujanpää, J., B ogniez, C., Hen io , N., A ola, A.,
Aun, M., Au iol, F., Bais, A. F., Be nha d, G., De Bock, V., Ca al-
amo, M., De oo, C., Diémoz, H., Egli, L., Fo es ie , J.-B., Foun-
oulakis, I., Ga ane, K., Ga cia, R. D., G öbne , J., Hassinen, S.,
Heikkilä, A., Hende son, S., Hülsen, G., Johnsen, B., Kalakoski,
N., Ka anikolas, A., Ka ppinen, T., Lamy, K., León-Luis, S. F.,
Lind o s, A. V., Me zge , J.-M., Min ielle, F., Muska el, H. B.,
Po a aix, T., Redondas, A., Sanchez, R., Siani, A. M., S endby,
T., and Tamminen, J.: Valida ion o he TROPOsphe ic Moni o -
ing Ins umen (TROPOMI) su ace UV adia ion p oduc , A -
mos. Meas. Tech., 13, 6999–7024, h ps://doi.o g/10.5194/am -
13-6999-2020, 2020.
Le el , P. F., Hilsen a h, E., Leppelmeie , G. W., an den Oo d, G.
H. J., Bha ia, P. K., Tamminen, J., de Haan, J. F., and Vee kind, J.
P.: Science objec i es o he ozone moni o ing ins umen , IEEE
T ansac ions on Geoscience and Remo e Sensing, 44, 1199–
1208, h ps://doi.o g/10.1109/TGRS.2006.872336, 2006.
Lind o s, A. V., Kujanpää, J., Kalakoski, N., Heikkilä, A., Lakkala,
K., Mielonen, T., Sneep, M., K o ko , N. A., A ola, A., and Tam-
minen, J.: The TROPOMI su ace UV algo i hm, A mos. Meas.
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7471
Tech., 11, 997–1008, h ps://doi.o g/10.5194/am -11-997-2018,
2018.
Long, C. S., Mille , A. J., Lee, H.-T., Wild, J. D., P zy-
wa y, R. C., and Hu o d, D.: Ul a iole Index Fo e-
cas s Issued by he Na ional Wea he Se ice, B. Am.
Me eo ol. Soc., 77, 729–748, h ps://doi.o g/10.1175/1520-
0477(1996)077<0729:UIFIBT>2.0.CO;2, 1996.
Lucas, R. M., Yaza , S., Young, A. R., No al, M., de G uijl, F.
R., Takizawa, Y., Rhodes, L. E., Sinclai , C. A., and Neale,
R. E.: Human heal h in ela ion o exposu e o sola ul a-
iole adia ion unde changing s a osphe ic ozone and cli-
ma e, Pho ochemical & Pho obiological Sciences, 18, 641–680,
h ps://doi.o g/10.1039/C8PP90060D, 2019.
Malino ic-Milice ic, S. and Mihailo ic, D. T.: The use o
NEOPLANTA model o e alua ing he UV index in he
Voj odina egion (Se bia), A mos. Res., 101, 621–630,
h ps://doi.o g/10.1016/j.a mos es.2011.04.008, 2011.
Maye , B. and Kylling, A.: Technical no e: The libRad an so -
wa e package o adia i e ans e calcula ions - desc ip ion
and examples o use, A mos. Chem. Phys., 5, 1855–1877,
h ps://doi.o g/10.5194/acp-5-1855-2005, 2005.
Maye , B., Seckmeye , G., and Kylling, A.: Sys ema ic long- e m
compa ison o spec al UV measu emen s and UVSPEC model-
ing esul s, Jou nal o Geophysical Resea ch: A mosphe es, 102,
8755–8767, h ps://doi.o g/10.1029/97JD00240, 1997.
McKenzie, R., Be nha d, G., Liley, B., Dis e ho , P., Rhodes, S.,
Bais, A., Mo gens e n, O., Newman, P., Oman, L., B ogniez,
C., and Simic, S.: Success o Mon eal P o ocol Demons a ed
by Compa ing High-Quali y UV Measu emen s wi h “Wo ld
A oided” Calcula ions om Two Chemis y-Clima e Models,
Sci. Rep., 9, 12332, h ps://doi.o g/10.1038/s41598-019-48625-
z, 2019.
McKenzie, R. L. and Lucas, R. M.: Reassessing Impac s o Ex-
ended Daily Exposu e o Low Le el Sola UV Radia ion,
Sci. Rep., 8, 13805, h ps://doi.o g/10.1038/s41598-018-32056-
3, 2018.
Mé éo-F ance: Algo i hm Theo e ical Basis Documen
o he Cloud P oduc P ocesso s o he NWC/-
GEO (GEO-CMA- 4.0, GEO-CT- 3.0, GEO-CTTH-
3.0, GEO-CMIC- 1.0), Technical Repo NWC/C-
DOP2/GEO/MFL/SCI/ATBD/Cloud, Issue 1, Re . 1, Mé éo-
F ance, h ps://www.nwcsa .o g/documen s/20182/30773/
NWC-CDOP2-GEO-MFL-SCI-ATBD-Cloud_ 1.1.pd (las
access: 10 Oc obe 2025), 2016.
Mok, J., K o ko , N. A., A ola, A., To es, O., Je h a, H.,
And ade, M., Labow, G., Eck, T. F., Li, Z., Dicke son, R.
R., S enchiko , G. L., Osipo , S., and Ren, X.: Impac s o
b own ca bon om biomass bu ning on su ace UV and ozone
pho ochemis y in he Amazon Basin, Sci. Rep., 6, 36940,
h ps://doi.o g/10.1038/s ep36940, 2016.
Mok, J., K o ko , N. A., To es, O., Je h a, H., Li, Z., Kim, J., Koo,
J.-H., Go, S., I ie, H., Labow, G., Eck, T. F., Holben, B. N., He -
man, J., Loughman, R. P., Spinei, E., Lee, S. S., Kha i, P., and
Campanelli, M.: Compa isons o spec al ae osol single sca e -
ing albedo in Seoul, Sou h Ko ea, A mos. Meas. Tech., 11, 2295–
2311, h ps://doi.o g/10.5194/am -11-2295-2018, 2018.
Molina, L. T. and Molina, M. J.: Absolu e abso p ion c oss sec-
ions o ozone in he 185- o 350-nm wa eleng h ange, Jou -
nal o Geophysical Resea ch: A mosphe es, 91, 14501–14508,
h ps://doi.o g/10.1029/JD091iD13p14501, 1986.
Mo gens e n, O., B aesicke, P., Hu wi z, M. M., O’Conno , F. M.,
Bushell, A. C., Johnson, C. E., and Pyle, J. A.: The Wo ld
A oided by he Mon eal P o ocol, Geophys. Res. Le ., 35,
h ps://doi.o g/10.1029/2008GL034590, 2008.
Newman, P. A. and McKenzie, R.: UV impac s a oided by he Mon-
eal P o ocol, Pho ochemical & Pho obiological Sciences, 10,
1152–1160, h ps://doi.o g/10.1039/C0PP00387E, 2011.
Papach is opoulou, K., Foun oulakis, I., Bais, A. F., Psiloglou,
B. E., Papadimi iou, N., Rap is, I.-P., Kazan zidis, A., Kon-
oes, C., Ha zaki, M., and Kazadzis, S.: E ec s o clouds and
ae osols on downwelling su ace sola i adiance nowcas ing
and sho - e m o ecas ing, A mos. Meas. Tech., 17, 1851–1877,
h ps://doi.o g/10.5194/am -17-1851-2024, 2024.
Pa isi, A., Igoe, D., Downs, N., Tu ne , J., Ama , A., and
A Jeba , M.: Sa elli e Moni o ing o En i onmen al So-
la Ul a iole A (UVA) Exposu e and I adiance: A Re-
iew o OMI and GOME-2, Remo e Sens (Basel), 13, 752,
h ps://doi.o g/10.3390/ s13040752, 2021.
Peuch, V.-H., Engelen, R., Rixen, M., Dee, D., Flemming, J., Su -
ie, M., Ades, M., Agus í-Pana eda, A., Ananasso, C., Ande -
sson, E., A ms ong, D., Ba é, J., Bousse ez, N., Dominguez,
J. J., Ga igues, S., Inness, A., Jones, L., Kipling, Z., Le e e-
Danczak, J., Pa ing on, M., Razinge , M., Ribas, R., Ve moo e,
S., Yang, X., Simmons, A., Ga cés de Ma cilla, J., and Thépau ,
J.-N.: The Cope nicus A mosphe e Moni o ing Se ice: F om
Resea ch o Ope a ions, B. Am. Me eo ol. Soc., 103, E2650–
E2668, h ps://doi.o g/10.1175/BAMS-D-21-0314.1, 2022.
P ei e , M. T., Koepke, P., and Reude , J.: E ec s o al i ude and
ae osol on UV adia ion, Jou nal o Geophysical Resea ch: A -
mosphe es, 111, h ps://doi.o g/10.1029/2005JD006444, 2006.
Piacen ini, R. D., Cede, A., and Bá cena, H.: Ex eme sola o-
al and UV i adiances due o cloud e ec measu ed nea he
summe sols ice a he high-al i ude dese ic pla eau Puna o
A acama (A gen ina), J. A mos. Sol. Te . Phys., 65, 727–731,
h ps://doi.o g/10.1016/S1364-6826(03)00084-1, 2003.
Rap is, I.-P., Kazadzis, S., Ele he a os, K., Ami idis, V.,
and Foun oulakis, I.: Single Sca e ing Albedo’s Spec-
al Dependence E ec on UV I adiance, A mosphe e,
h ps://doi.o g/10.3390/a mos9090364, 2018.
Redondas, A., E ans, R., S uebi, R., Köhle , U., and We-
be , M.: E alua ion o he use o i e labo a o y-de e mined
ozone abso p ion c oss sec ions in B ewe and Dobson e-
ie al algo i hms, A mos. Chem. Phys., 14, 1635–1648,
h ps://doi.o g/10.5194/acp-14-1635-2014, 2014.
Reude , J. and Schwande , H.: Ae osol e ec s on
UV adia ion in nonu ban egions, Jou nal o Geo-
physical Resea ch: A mosphe es, 104, 4065–4077,
h ps://doi.o g/10.1029/1998JD200072, 1999.
Riede , H. E., S aehelin, J., Weihs, P., Vuilleumie , L., Maede , J.
A., Holawe, F., Blum hale , M., Lind o s, A., Pe e , T., Simic,
S., Spich inge , P., Wagne , J. E., Walke , D., and Riba e , M.:
Rela ionship be ween high daily e y hemal UV doses, o al
ozone, su ace albedo and cloudiness: An analysis o 30yea s
o da a om Swi ze land and Aus ia, A mos. Res., 98, 9–20,
h ps://doi.o g/10.1016/j.a mos es.2010.03.006, 2010.
Roshan, D. R., Koc, M., Abdallah, A., Ma in-Poma es,
L., Isai an, R., and Foun oukis, C.: UV Index Fo e-
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025

7472 I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling
cas ing unde he In luence o Dese Dus : E alua ion
agains Su ace and Sa elli e-Re ie ed Da a, A mosphe e,
h ps://doi.o g/10.3390/a mos11010096, 2020.
Schenzinge , V., K eu e , A., Klo z, B., Schwa zmann, M., and
G öbne , J.: On he p oduc ion and alida ion o sa elli e
based UV index maps, A mos. Meas. Tech. Discuss. [p ep in ],
h ps://doi.o g/10.5194/am -2023-188, 2023.
Schmalwiese , A. W., G öbne , J., Blum hale , M., Klo z, B., De
Backe , H., Bolsée, D., We ne , R., Tomsic, D., Me elka, L.,
E iksen, P., Jepsen, N., Aun, M., Heikkilä, A., Dup a , T., Sand-
mann, H., Weiss, T., Bais, A., To h, Z., Siani, A.-M., Vacca o,
L., Diémoz, H., G i oni, D., Zipoli, G., Lo enze o, G., Pe ko ,
B. H., di Sa a, A. G., Massen, F., Yousi , C., Aculinin, A. A.,
den Ou e , P., S endby, T., Dahlback, A., Johnsen, B., Biszczuk-
Jakubowska, J., K zyscin, J., Hen iques, D., Chuba o a, N., Ko-
la ž, P., Mija o ic, Z., G oselj, D., P ibullo a, A., Gonzales, J.
R. M., Bilbao, J., Gue e o, J. M. V., Se ano, A., Ande sson, S.,
Vuilleumie , L., Webb, A., and O’Hagan, J.: UV Index moni o -
ing in Eu ope, Pho ochemical & Pho obiological Sciences, 16,
1349–1370, h ps://doi.o g/10.1039/C7PP00178A, 2017.
Schmucki, D. A. and Philipona, R.: UV adia ion in he
Alps: he al i ude e ec , P oc. SPIE 4482, 234–239,
h ps://doi.o g/10.1117/12.452923, 2002.
Schulz, M., Ch is ophe, Y., Ramone , M., Wagne , A., Eskes, H.
J., Basa , S., Benedic ow, A., Bennouna, Y., Blechschmid , A.-
M., Chab illa , S., Cue as, E., El Yazidi, A., Flen je, H., Hansen,
K. M., Im, U., Kapsomenakis, J., Lange ock, B., Rich e , A.,
Suda chiko a, N., Thou e , V., Wa neke, T., and Ze e os, C.:
Valida ion epo o he CAMS nea - eal- ime global a mo-
sphe ic composi ion se ice: Pe iod Decembe 2018 -Feb ua y
2019, Cope nicus A mosphe e Moni o ing Se ice (CAMS) e-
po , CAMS84_2018SC1_D1.1.1_DJF2019_ 1.pd , June 2019,
h ps://doi.o g/10.24380/7 h6- k72, 2022.
She le, E.: Models o ae osols, clouds, and p ecipi a ion o a -
mosphe ic p opaga ion s udies, AGARD Con . P oce., 1, h ps:
//ui.adsabs.ha a d.edu/abs/1990apu .aga .....S/abs ac (las ac-
cess: 10 Oc obe 2025), 1990.
Siani, A. M., Casale, G. R., Diémoz, H., Agnesod, G., Kimlin, M.
G., Lang, C. A., and Colosimo, A.: Pe sonal UV exposu e in
high albedo alpine si es, A mos. Chem. Phys., 8, 3749–3760,
h ps://doi.o g/10.5194/acp-8-3749-2008, 2008.
Sola, Y., Lo en e, J., Campmany, E., de Cabo, X., Bech, J.,
Redaño, A., Ma ínez-Lozano, J. A., U illas, M. P., Alados-
A boledas, L., Olmo, F. J., Díaz, J. P., Expósi o, F. J., Ca-
cho o, V., So ibas, M., Labajo, A., Vilaplana, J. M., Sil a,
A. M., and Badosa, J.: Al i ude e ec in UV adia ion du -
ing he E alua ion o he E ec s o Ele a ion and Ae osols
on he Ul a iole Radia ion 2002 (VELETA-2002) ield cam-
paign, Jou nal o Geophysical Resea ch: A mosphe es, 113,
h ps://doi.o g/10.1029/2007JD009742, 2008.
S aige , H., den Ou e , P. N., Bais, A. F., Feis e , U., Johnsen,
B., and Vuilleumie , L.: Hou ly esol ed cloud modi ica ion
ac o s in he ul a iole , A mos. Chem. Phys., 8, 2493–2508,
h ps://doi.o g/10.5194/acp-8-2493-2008, 2008.
Tanskanen, A., K o ko , N. A., He man, J. R., and A ola,
A.: Su ace ul a iole i adiance om OMI, IEEE T ansac-
ions on Geoscience and Remo e Sensing, 44, 1267–1271,
h ps://doi.o g/10.1109/TGRS.2005.862203, 2006.
TEMIS: h ps://www. emis.nl/u adia ion/UVa chi e.php, las ac-
cess: 10 Oc obe 2025.
U illas, M. P., Ma ín, M. J., Es e e, A. R., Salaza , G.,
Sua ez, H., Cas illo, J., and Ma ínez-Lozano, J. A.:
UVER and UV index a high al i ude in No hwes e n
A gen ina, J. Pho ochem. Pho obiol. B, 163, 290–295,
h ps://doi.o g/10.1016/j.jpho obiol.2016.08.012, 2016.
an Dijk, A., Slape , H., den Ou e , P. N., Mo gens e n, O.,
B aesicke, P., Pyle, J. A., Ga ny, H., S enke, A., Dame is,
M., Kazan zidis, A., Tou pali, K., and Bais, A. F.: Skin can-
ce isks a oided by he Mon eal P o ocol–wo ldwide mod-
eling in eg a ing coupled clima e-chemis y models wi h a
isk model o UV, Pho ochem. Pho obiol., 89, 234–246,
h ps://doi.o g/10.1111/j.1751-1097.2012.01223.x, 2013.
Vanicek, K., F ei, T., Li ynska, Z., and Schmalwiese , A.: UV-
Index o he Public, Publica ion o he Eu opean Com-
muni ies, B ussels, Belgium, h ps://api.seman icschola .o g/
Co pusID:46521251 (las access: 10 Oc obe 2025), 2000.
Ve debou , J.: A me hod o gene a e su ace UV adia ion maps
o e Eu ope using GOME, Me eosa , and ancilla y geophysical
da a, Jou nal o Geophysical Resea ch: A mosphe es, 105, 5049–
5058, h ps://doi.o g/10.1029/1999JD900302, 2000.
Vi , R., Laschewski, G., Bais, A., Diémoz, H., Foun oulakis, I.,
Siani, A.-M., and Ma za akis, A.: UV-Index Clima ology o
Eu ope Based on Sa elli e Da a, A mosphe e (Basel), 11, 727,
h ps://doi.o g/10.3390/a mos11070727, 2020.
Webb, A. R. and Engelsen, O.: Calcula ed Ul a iole Expo-
su e Le els o a Heal hy Vi amin D S a us, Pho ochem.
Pho obiol., 82, 1697–1703, h ps://doi.o g/10.1111/j.1751-
1097.2006. b09833.x, 2006.
Webb, A. R., Slape , H., Koepke, P., and Schmalwiese ,
A. W.: Know You S anda d: Cla i ying he CIE E y-
hema Ac ion Spec um, Pho ochem. Pho obiol., 87, 483–486,
h ps://doi.o g/10.1111/j.1751-1097.2010.00871.x, 2011.
Webb, A. R., Kazan zidis, A., Ki , R. C., Fa a , M. D., Wilkinson,
J., and Rhodes, L. E.: Colou Coun s: Sunligh and Skin Type as
D i e s o Vi amin D De iciency a UK La i udes, Nu ien s, 10,
h ps://doi.o g/10.3390/nu10040457, 2018.
Webb, A. R., Te ene skaya, I. P., Holick, M. F., an Dijk,
A., McKenzie, R. L., Lucas, R. M., Young, A. R.,
Philipsen, P. A., and de G uijl, F. R.: P e i amin D ac-
ion spec um: Challenging CIE owa ds a s anda d,
Ligh ing Resea ch & Technology, 14771535221122936,
h ps://doi.o g/10.1177/14771535221122937, 2022.
Weihs, P. and Webb, A. R.: Accu acy o spec al UV model calcu-
la ions: 2. Compa ison o UV calcula ions wi h measu emen s,
Jou nal o Geophysical Resea ch: A mosphe es, 102, 1551–
1560, h ps://doi.o g/10.1029/96JD02621, 1997.
Weihs, P., Lenoble, J., Blum hale , M., Ma in, T., Seckmeye ,
G., Philipona, R., De la Casinie e, A., Se gen , C., G öbne ,
J., Cabo , T., Masse o , D., Pichle , T., Pouga ch, E., Ren-
ga ajan, G., Schmucki, D., and Simic, S.: Modeling he e -
ec o an inhomogeneous su ace albedo on inciden UV
adia ion in moun ainous e ain: De e mina ion o an e -
ec i e su ace albedo, Geophys. Res. Le ., 28, 3111–3114,
h ps://doi.o g/10.1029/2001GL012986, 2001.
WHO: En i onmen al Heal h C i e ia 160 – Ul a iole
Radia ion, h ps://www.icni p.o g/en/publica ions/a icle/
ehc-160-u -1994.h ml (las access: 10 Oc obe 2025), 1994.
Geosci. Model De ., 18, 7451–7473, 2025 h ps://doi.o g/10.5194/gmd-18-7451-2025
I. Foun oulakis e al.: Assessmen o he accu acy in UV index modelling 7473
Za a i, F., Fo no, R. N., Ga cía Fuen es, J., and And ade, M. F.:
E y hemally weigh ed UV a ia ions a wo high-al i ude lo-
ca ions, Jou nal o Geophysical Resea ch: A mosphe es, 108,
h ps://doi.o g/10.1029/2001JD000918, 2003.
Zempila, M.-M., Koukouli, M.-E., Bais, A., Foun oulakis, I.,
A ola, A., Kou eme i, N., and Balis, D.: OMI/Au a UV
p oduc alida ion using NILU-UV g ound-based measu e-
men s in Thessaloniki, G eece, A mos. En i on., 140, 283–297,
h ps://doi.o g/10.1016/j.a mosen .2016.06.009, 2016.
Zempila, M.-M., an Ge en, J. H. G. M., Taylo , M., Foun-
oulakis, I., Koukouli, M.-E., an Weele, M., an de A, R.
J., Bais, A., Mele i, C., and Balis, D.: TEMIS UV p od-
uc alida ion using NILU-UV g ound-based measu emen s in
Thessaloniki, G eece, A mos. Chem. Phys., 17, 7157–7174,
h ps://doi.o g/10.5194/acp-17-7157-2017, 2017.
Ze e os, C., Foun oulakis, I., Ele he a os, K., and Kazan zidis,
A.: The long- e m a iabili y o human heal h ela ed sola
ul a iole -B adia ion doses o he 1980s o he end o 21s cen-
u y, Physiol. Re ., h ps://doi.o g/10.1152/phys e .00031.2022,
2023.
h ps://doi.o g/10.5194/gmd-18-7451-2025 Geosci. Model De ., 18, 7451–7473, 2025