Twilight Mesospheric Clouds in Jezero as Observed by MEDA Radiation and Dust Sensor (RDS)

1. In oduc ion
The H2O and CO2 ice clouds on Ma s a e p ima y cons i uen s s udied o unde s anding he pas and p esen
clima e o he plane (Fo ge & Pie ehumbe ,1997; Mon messin e al.,2004). Cloud pa icles can a ec he
ene gy balance o he plane (e.g., Wol e al.,2019), and hus he a mosphe ic dynamics, as well as in luence
he e ical dis ibu ion o dus pa icles h ough dus sca enging. The dus sca enging by H2O clouds has c i ical
consequences in he wa e cycle o he plane ; o example, egions in he a mosphe e wi h insu icien quan i y o
dus pa icles (o condensa ion nuclei) can inhibi he o ma ion o H2O clouds (Mää änen e al.,2005; Mon messin
e al.,2002), and hus each wa e apo concen a ions in excess o sa u a ion (Mal aglia i e al.,2011; Na a o
Abs ac The Ma s En i onmen al Dynamics Analyze ins umen , on boa d NASA's Ma s 2020
Pe se e ance o e , includes a numbe o senso s o cha ac e ize he Ma ian a mosphe e. One o hese senso s
is he Radia ion and Dus Senso (RDS) ha measu es he sola i adiance a di e en wa eleng hs and
geome ies. We analyzed he RDS obse a ions made du ing wiligh o he pe iod be ween sol 71 and 492 o
he mission (Ls 39°–262°, Ma s Yea 36) o cha ac e ize he clouds o e he Pe se e ance o e si e. Using he
a io be ween he i adiance a zeni h a 450 and 750nm, we in e ed ha he main cons i uen o he de ec ed
high-al i ude ae osol laye s was ice om Ls=39°–150° (cloudy pe iod), and dus om Ls 150°–262°. A o al
o 161 wiligh s we e analyzed in he cloudy pe iod using a adia i e ans e code and we ound: (a) signa u es
o clouds/hazes in he signals in 58% o he wiligh s; (b) mos o he clouds had al i udes be ween 40 and
50km, sugges ing wa e ice composi ion, and had pa icle sizes be ween 0.6 and 2µm; (c) he cloud ac i i y
a sun ise is sligh ly highe ha a sunse , likely due o he di e ences in empe a u e; (d) he ime pe iod wi h
mo e cloud de ec ions and wi h he g ea es cloud opaci ies is du ing Ls 120°–150°; and (e) a no able dec ease
in he cloud ac i i y a ound aphelion, along wi h lowe cloud al i udes and opaci ies. This dec ease in cloud
ac i i y indica es lowe concen a ions o wa e apo o cloud condensa ion nuclei (dus ) a ound his pe iod in
he Ma ian mesosphe e.
Plain Language Summa y Du ing wiligh , g ound-based obse a ions o he i adiance allows
he de ec ion and cha ac e iza ion o high-al i ude clouds (abo e 30–35km). Because he sun is a o below he
ho izon, he cloud laye s e lec he di ec ligh ha only eaches he highe pa s o he a mosphe e, p oducing
an inc ease in he sky b igh ness wi h espec o he cloud- ee scena io. Mo eo e , he dec ease in he in ensi y
wi h he sola zeni h angle highly depends on he cloud al i ude and densi y. Using obse a ions made by
he Radia ion and Dus Senso , pa o he ins umen Ma s En i onmen al Dynamics Analyze on boa d
Pe se e ance o e , we p esen he e a s udy o he wiligh clouds de ec ed a he Pe se e ance landing si e o
he i s 490 sols o he mission (Ma s Yea 36). By modeling he i adiance a 450 and 950nm wi h adia i e
ans e simula ions, we cons ained he cloud al i ude, opaci y, and pa icle adius. The numbe o wiligh s
analyzed allowed us o s udy he seasonal end in he cloud ac i i y. Du ing he cloudy pe iod, Ls 39°–150°,
we ind a signi ican dec ease in he cloud ac i i y abo e 30–35km a ound aphelion (Ls∼70°). This implies
ha he seasonal dis ibu ion o clouds abo e 30–35km di e s om ha obse ed a lowe al i udes.
TOLEDO ETAL.
© 2023 The Au ho s.
This is an open access a icle unde
he e ms o he C ea i e Commons
A ibu ion-NonComme cial License,
which pe mi s use, dis ibu ion and
ep oduc ion in any medium, p o ided he
o iginal wo k is p ope ly ci ed and is no
used o comme cial pu poses.
Twiligh Mesosphe ic Clouds in Jeze o as Obse ed by MEDA
Radia ion and Dus Senso (RDS)
D. Toledo1 , L. Gómez1 , V. Apés igue1 , I. A uego1 , M. Smi h2 ,
A. Mungui a3 , G. Ma ínez4 , P. Pa el5 , A. Sanchez-La ega3 , M. Lemmon6 , L. Tamppa i5 ,
D. Viudez-Mo ei as7 , R. Hueso3 , A. Vicen e-Re o illo7 , C. Newman8 , R. Lo enz9 ,
M. Yela1 , M. de la To e Jua ez5 , and J. A. Rod iguez-Man edi7
1Ins i u o Nacional de Técnica Ae oespacial (INTA), Mad id, Spain, 2NASA Goda d Space Fligh Cen e , G eenbel , MD,
USA, 3Uni e sidad del País Vasco UPV/EHU, Bilbao, Spain, 4Luna and Plane a y Ins i u e, Uni e si ies Space Resea ch
Associa ion, Hous on, TX, USA, 5Je P opulsion Labo a o y, Cali o nia Ins i u e o Technology, Pasadena, CA, USA, 6Space
Science Ins i u e, Boulde , CO, USA, 7Cen o de A obiología (INTA-CSIC), To ejón de A doz, Mad id, Spain, 8Aeolis
Resea ch, Chandle , AZ, USA, 9Johns Hopkins Applied Physics Labo a o y, Lau el, MD, USA
Key Poin s:
• Mos o he cloud de ec ed a wiligh
be ween sol 71 and 492 o he Ma s
2020 mission (Ls 39°–262°) occu ed
a al i udes be ween 40 and 50km
• A ound aphelion (Ls∼70°) we ound
he minimum in cloud ac i i y and
lowe cloud opaci ies
• The cloud ac i i y a sun ise is sligh ly
s onge han a sunse and his is
likely due o he lowe empe a u es
Co espondence o:
D. Toledo,
[email p o ec ed]
Ci a ion:
Toledo, D., Gómez, L., Apés igue, V.,
A uego, I., Smi h, M., Mungui a, A.,
e al. (2023). Twiligh mesosphe ic
clouds in Jeze o as obse ed by MEDA
Radia ion and Dus Senso (RDS).
Jou nal o Geophysical Resea ch:
Plane s, 128, e2023JE007785. h ps://doi.
o g/10.1029/2023JE007785
Recei ed 11 FEB 2023
Accep ed 22 JUN 2023
10.1029/2023JE007785
RESEARCH ARTICLE
1 o 18
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
2 o 18
e al.,2014). Al hough o he mechanisms may be esponsible o he exis ence o supe sa u a ion on Ma s as well
(Fedo o a e al.,2020), i is well es ablished ha he o ma ion o clouds limi s he concen a ion o wa e apo
o alues below sa u a ion, and his pa ially con ols he amoun o wa e apo ha can be anspo ed o he
highe pa s o he a mosphe e, whe e he wa e can be pho odissocia ed in o i s ligh e componen s H and O.
One o he wo majo cloud egimes on Ma s is he aphelion cloud bel (ACB) (Clancy e al.,1996) occu ing in
he equa o ial egions be ween ∼10°S and ∼30°N and du ing he no he n sp ing and summe (Ls∼0°–180°).
The o he main cloud egime is he hoods o e he pola cap in bo h hemisphe es (Benson e al.,2010,2011)
du ing la e summe and he whole win e ( hese clouds a e no s udied in his wo k). As epo ed in many
p e ious wo ks, bo h e en s a e obse ed e e y Ma ian yea (MY). Al hough H2O and CO2 clouds o hazes
ha e been obse ed and s udied om he su ace o Ma s, di ec ly h ough images o indi ec ly h ough ae o-
sol opaci y measu emen s (e.g., Lemmon e  al., 2015; Lo enz e  al., 2020; P. H. Smi h & Lemmon, 1999),
he longes eco d o cloud e en s comes om ins umen a ion onboa d o bi e s (e.g., Mää änen e al.,2010;
McConnochie e al.,2010; Sánchez-La ega e al.,2018; Tamppa i e al.,2003; Wang & Inge soll,2002; Wol
e al.,2022). When hese o bi e obse a ions a e made a limb- iewing geome y, in o ma ion on he cloud
e ical p o iles can be de i ed (e.g., Rannou e al.,2006; M. D. Smi h e al.,2013). In hese pa icula cases he
cloud equency-o -occu ence o p ope ies (e.g., opaci y, pa icle adius) can be s udied as a unc ion o he al i-
ude. On he o he hand, i he o bi e obse a ions a e ob ained a nadi - iewing geome y, in gene al he cloud
e ical p o iles canno be de i ed and he o al ice column opaci y is p o ided (e.g., Giu anna e al.,2021; M. D.
Smi h,2009). While o bi al obse a ions p o ide a mo e comple e global co e age, landed obse a ions ep esen
a c i ically impo an componen o: (a) c oss alida e he o bi al obse a ions and e ie als; (b) s udy he diu nal
and seasonal a ia ions o he cloud ac i i y wi hou he impac o he o bi e spa ial and empo al sampling; and
(c) in es iga e he a mosphe ic con ex in which he clouds we e o med (i obse a ions o me eo ological ime
se ies a e a ailable).
On 18 Feb ua y 2021, he Ma s 2020 o e Pe se e ance success ully landed in Jeze o c a e (la i ude 18.44°N and
longi ude 77.45°E). To p o ide me eo ological con ex o o he obse a ions and o u u e human explo a ion,
Pe se e ance ca ies he Ma s En i onmen al Dynamics Analyze (MEDA) (Rod iguez-Man edi e al.,2021)
ins umen , which includes a se o senso s: wo wind senso s o in e wind di ec ion and speed, i e he mal
senso s a di e en loca ions and heigh s (ATS), an in a ed adiome e o measu e g ound and a mosphe ic
empe a u e as well as a mosphe ic IR luxes and e lec ed sola luxes (TIRS), a ela i i y humidi y senso , a
p essu e senso , and he Radia ion and Dus Senso (RDS) ha measu es he sola adia ion a di e en wa e-
leng hs anges om he UVA o he nea in a ed (Rod iguez-Man edi e al.,2021,2023). In his pape we ocus
on RDS obse a ions a wiligh , when he sola zeni h angle (SZA) is be ween 90° and 98°, o de ec and cha -
ac e ize high-al i ude clouds (abo e ∼30km) o he i s 492 sols o he mission (MY 36). We b ie ly desc ibe
he RDS in Sec ion2, as well as he obse a ions, he p inciple o measu emen and he adia i e ans e (RT)
modeling. In Sec ion3, we p esen ime se ies o high-al i ude ae osol laye s (ALs) de ec ed du ing wiligh , he
cloud e ie als and main esul s.
2. Obse a ions and Radia i e T ans e Modeling
2.1. RDS Ins umen
RDS measu es he sola i adiance a di e en spec al wa eleng hs and inciden geome ies. I is comp ised o
wo se s o pho ode ec o s (RDS-DP) and a came a poin ing a zeni h (RDS-SkyCam). The i s se o pho ode ec-
o s, he Top channels, co esponds o eigh zeni h-poin ed de ec o s which co e he ligh spec um om UVA o
Nea IR (Top-1 o Top-8: 255, 259, 250–400, 450, 650, 750, 190–1,100, and 950nm). Mos o he Top de ec o s
use in e e en ial il e s and mechanical masks (Apes igue e al.,2022) o cons ain hei ield o iew o ±15°
zeni h angle, while he Top-7 channel co e s he ull sky om 0° o 90° zeni h angle and o all azimu h angles.
The second se co esponds o he eigh La e al channels, which a e poin ed sideways a 20° (excep La -8, which
is 35°) abo e he o e deck and a e all a 750nm. The La -1 channel is blinded o s udy he pho ode ec o pe o -
mance deg ada ion. In his wo k only he obse a ions made by he Top senso s will be used. In gene al, MEDA
sampling is se a 1Hz wi h all senso s ope a ing o blocks o 1h and 5min. The disposi ion o he blocks
along he day a e selec ed o each sol based on a cadence ha al e na es e en and odd hou s, and he du a ion
and numbe o block some imes change depending on powe a ailabili y and da a olume cons ain s. Fo his
eason, no all he wiligh s a e co e ed by MEDA. Because o he low le els o i adiance expec ed du ing his
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
3 o 18
ime o he day, hese RDS obse a ions a e acqui ed wi h an ex a 40 gain ac o (Apes igue e al.,2022), which
is ac i a ed when SZA≥90°. No e ha because he RDS gain ac o was no ac i a ed un il sol 70, ou analysis
does no co e he i s sols o he mission.
2.2. P inciple o Measu emen o he De ec ion o Clouds
Since high-al i ude ALs (e.g., H2O clouds o de ached dus laye s) imply an inc ease in he i adiance du ing
wiligh , clouds can be de ec ed by looking a he e olu ion o he RDS Top obse a ions a SZAs be ween 90°
and 98° (∼30min long). Mo eo e , du ing his pe iod only he highe pa s o he a mosphe e ecei e Sun di ec
ligh (as o al i udes
𝐴𝐴𝐴
R
(
1+ an
2(90◦− SZA)
)0.5
, whe e R is he adius o he plane , he di ec ligh in e sec s
he plane su ace), making he a ia ion o he i adiance wi h SZA e y sensi i e o cloud p ope ies such as he
al i ude o he numbe densi y. As indica ed in Toledo, Rannou, Pomme eau, Sa kissian, and Foujols(2016), his
echnique allows de ec ing clouds wi h e y low opaci ies (sub isual opaci ies), as he pa hway o sunligh in a
ho izon ally homogeneous AL o h geome ical hickness is enhanced by a ac o >1/sin(SZA−90°). Figu e1
shows, as an example, RDS signals measu ed by Top-4 (450±10nm) and Top-8 (950±10nm) senso s a
Figu e 1. The uppe panels show a compa ison be ween Radia ion and Dus Senso (RDS) obse a ions a 450 (Top 4)
and 950nm (Top 8) made unde cloud- ee condi ions (sol 99) and unde he p esence o clouds (sol 271). The p esence
o he clouds esul in an inc ease in he i adiance (indica ed wi h he black a ows). No e ha each signal was no malized
by he signal alue a sola zeni h angle=90°. By doing so, we diminish he impac o he dus opaci y and pa icle adius
on he RDS signals, and make he compa ison be ween signals easie o in e p e . As we will see in he ollowing sec ion,
he no maliza ion o he signals also allow us o educe he numbe o ee pa ame e s in he adia i e ans e analysis. The
p esence o clouds (o hazes) on sol 271 a sun ise was also con i med by images aken by he Visual Moni o ing Came a
(VMC) (Sánchez-La ega e al.,2018) onboa d Ma s Exp ess (lowe panel). In he VMC images we see ha Jeze o c a e
(indica ed wi h he black a ow) was o e cas by b igh mo ning limb clouds o hazes.
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
4 o 18
wiligh o a cloud ee day and o a day wi h he p esence o clouds (also con i med by o bi e obse a ions).
We obse e ha he clouds p oduce an inc ease in he i adiance (indica ed by he black a ows) wi h espec
o he scena io wi hou clouds, and hus by compa ing each wiligh eco ded by he RDS wi h he signals o
he cloud- ee scena io, we can in e on wha sols o he mission he e we e clouds (o high dus laye s) p esen
du ing his pe iod o he day.
An easy way o de e mine when cloud ea u es a e p esen in he RDS obse a ions is o compa e he Top-4 and
Top-8 signals o each wiligh wi h hose o he cloud- ee day o Figu e1 ( ed lines and e e ed he ea e as
he e e ence signals). Figu e2 shows, as an example, he co ela ion be ween he wiligh RDS signals a 450
and 950nm and he e e ence signals (a he same wa eleng hs and SZAs) o h ee di e en sun ises. The wo
co ela ion cu es (one o Top-4 and ano he o Top-8) de i ed o each sun ise we e i ed o a s aigh line,
whose slopes we e compa ed wi h he iden i y ela ion (g ay dashed line). I he slopes a e close o 1, hen he
obse a ions indica e ae osol condi ions simila o hose ound o he e e ence signals (cloud- ee wiligh ). On
he con a y, i he slopes a e <1, hen he obse a ions poin o he possible p esence o high ALs. No e ha
because he high ALs inc ease i adiance a su ace du ing wiligh and he e e ence signals a e plo ed in he
x-axis o Figu e2, he slopes a e expec ed o be smalle han 1 unde cloud condi ions. In he examples illus a ed
in Figu e2, we ob ained slope alues o 1.01, 0.95, and 0.69 a 450nm o sols 190, 178, and 311, espec i ely.
We es ima ed ha he slopes a e signi ican ly di e en o 1 when he slopes a e smalle han ∼0.97. This h esh-
old ep esen s he maximum slope alue below 1 o which signal di e ences ela i e o he e e ence signal
a e signi ican (accoun ing o signal e o s). The e o e, based on hese esul s we can in e ha high ALs we e
po en ially p esen a sun ise only o sols 178 and 311.
Fo de e mining he composi ion o he de ec ed high ALs (laye s made o ices o jus dus ), as bo h high-al i ude
clouds o de ached dus laye s a e expec ed o cause simila e ec s on he slope alues, we will make use o he
a io be ween he in ensi y a zeni h measu ed a wo di e en wa eleng hs. In pa icula , by choosing wo wa e-
leng hs a which he single sca e ing albedo (o he imagina y pa o he e ac i e index) o he dus pa icles
is e y di e en bu app oxima ely he same o wa e ice, hen he alue o he a io be ween he in ensi ies a
hese wo wa eleng hs highly depends on he ae osol composi ion. We can compu e hese a ios, de ined he e as
he colo index (CI), om he measu emen s made by di e en RDS Top channels; a CI om he a io be ween
Top 3 (250–400) and Top 6 (750nm) channels, and ano he CI om he a io be ween Top 4 (450) and Top 6
(750nm) channels. This selec ion o channels is based on he ac ha dus pa icles ha e a much g ea e imag-
ina y e ac i e index a 250–400 and 450nm han a 750nm (Wol e al.,2009,2010). Acco ding o his CI
de ini ion and since he single sca e ing albedo o he wa e (o CO2) ice pa icles is ∼1 in any o hese h ee RDS
channels, we expec g ea e alues o Top3 (250–400nm)/Top6 (750nm) and Top4 (450nm)/Top6 (750nm)
when he high-al i ude ALs a e made o wa e ice han when hey a e composed o only dus . Al hough simila
esul s would be ob ained by using he Top 8 (950nm) channel ins ead o he Top 6, we made his elec ion
because he Top 6 wa eleng h ange is he closes one o he minimum in he imagina y e ac i e index o he
dus (Wol e al.,2009). I is impo an o no e he e ha he CI is also sensi i e o a ia ions in he cloud pa icle
size. In pa icula , an inc ease in he pa icle adius o he high-al i ude ALs would also dec ease he CI alues
Figu e 2. Co ela ion be ween he Radia ion and Dus Senso signals a 450 (blue) and 950nm ( ed) measu ed o a cloud- ee day, ep esen ed in he x-axis, and
du ing he dawn o (a) sol 190, (b) 178, and (c) 311, ep esen ed in he y-axis. Fo each wiligh we de i ed wo co ela ion plo s, one pe channel, whose measu emen s
a e compa ed wi h he signals measu ed unde cloud- ee condi ions o he same sola zeni h angles. The co ela ion cu es we e i ed o a s aigh line (solid lines
in blue and ed o he Top 4 and Top 8 channels, espec i ely) whose slope is used o in e he p esence o ae osol laye . The g ay dashed line ep esen s he iden i y
ela ion.
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
5 o 18
( o a cons an opaci y). This is due o he dependence o he phase unc ion and single sca e ing albedo on he
pa icle adius.
2.3. A mosphe ic and Radia i e T ans e Modeling
The cloud p ope ies a e de i ed by modeling he RDS Top measu emen s wi h RT simula ions. We i s in es-
iga ed he sensi i i y o he RDS signals o di e en cloud p ope ies (e.g., al i ude, opaci y, geome ical hick-
ness, pa icle shape), and ound ha he cloud al i ude, numbe densi y, and pa icle adius a e he pa ame e s
wi h he g ea es impac . RT simula ions a wiligh a e made wi h a h ee-dimensional Mon e Ca lo RT model
in sphe ical geome y (since he plane-pa allel app oxima ion b eaks down o high SZAs) p e iously used o
cloud p ope ies e ie als on Ea h (Gomez-Ma in e al.,2021; Toledo, Rannou, Pomme eau, Sa kissian, &
Foujols,2016), Ti an (Rannou e al.,2016; Wes e al.,2016), and adap ed o he Ma ian a mosphe e (Toledo,
Rannou, Pomme eau, & Foujols,2016; Toledo e al.,2017). Since Mon e Ca lo RT simula ions ake a long
ime o calcula e, he e ie al p ocedu e makes use o a p e-compu ed se o look-up ables, o minimizing
he mean squa e di e ence be ween simula ed and obse ed RDS signals. Cloud sca e ing p ope ies a e
compu ed wi h Mie heo y and he e ac i e index o wa e ice (Wa en,1984). The cloud geome ical hick-
ness was ixed a 2km (we a ied his pa ame e up o alues o 6km and did no ound signi ican a ia ions
in he simula ions), and he cloud spa ial dis ibu ion densi y was de ined by a Gaussian heigh p o ile, scaled
o p oduce he desi ed opaci y. The dus sca e ing p ope ies we e de i ed om he empi ical o mula ion
p oposed by Pollack and Cuzzi(1980) and using he spec al e ac i e index gi en in Wol e al.(2009) is
used. The RDS signals we e no malized by he in ensi y measu ed a SZA=90° (o he minimum SZA o he
wiligh ) o educe he impac o he backg ound dus p ope ies (opaci y and e ) on he cloud e ie als. Fo
he e ical dis ibu ion o dus pa icles, we adop ed he modi ied Con a h p o ile (Con a h,1975) p oposed
by Fo ge e al.(1999).
𝜏𝜏
(z) = 𝜏𝜏0⋅𝜎𝜎(z) ⋅exp
[
𝜈𝜈⋅
(
1−𝜎𝜎(z)
−l)]
(1)
whe e τ0 is he e ical opaci y a su ace, σ(z) is he a io be ween he p essu e a z le el and he p essu e a
su ace (he e we assume p a ies wi h heigh as p=p0⋅exp(−z/H), whe e p0 is he p essu e a su ace and H he
scale heigh and equal o 11km), ν is a cons an se o 0.007 and l is he a ion be ween a e e ence heigh (se o
70km) and he al i ude o he op o he dus laye (Zmax). We in es iga ed he use o mo e complex dus e ical
dis ibu ions in ou RT simula ions. Based on p e ious wo ks, we simula ed he RDS signals using a dus e ical
p o ile esul ing om a Con a h- ype p o ile and a de ached dus laye (de ined by a Gaussian heigh p o ile)
wi h a iable al i ude. We ound no signi ican di e ences in he cloud e ie als using his non-mono onic dus
e ical dis ibu ions o de ached-dus laye al i udes less han o equal o 25km. Based on he esul s epo ed
in McCleese e al.(2010) and Hea ens e al.(2011a,2011b), which ound he maximum dus mass mixing a io
a al i udes be ween 15 and 25km ( o MY 28–29 and o mos o he no he n sp ing and summe ), we a o ed
he simple Con a h- ype p o iles o e mo e complex dus s uc u es o ou e ie al analysis.
3. Resul s
3.1. P esence o Clouds in he Pe iod Ls 39°–262°
The slope analysis desc ibed in 2.2 was pe o med o all he wiligh s a ailable up o sol 492 (Ls=262°), whose
esul s a e displayed in he uppe panel o Figu e3a. An inspec ion o he slope alues e eals 4 ob ious pe iods
o di e en high al i ude ae osol ac i i y:
1. Be ween Ls∼39° and 50°, high-al i ude ALs signa u es in he RDS signals a e ound o abou ∼40% o he
wiligh s co e ed by MEDA. In gene al, he slopes ob ained a sun ise a e smalle han hose du ing sunse ,
sugges ing g ea e opaci ies o al i udes. We canno es ablish he s a o his pe iod as no RDS da a wi h high
gain is a ailable be o e Ls=39°. In he ollowing sec ion and in AppendixA we will show ha hese ALs a e
a al i udes abo e ∼30km.
2. The second pe iod, be ween Ls∼50° and 114°, is cha ac e ized by a no able d op in he high-al i ude ae o-
sol de ec ions: in his pe iod, alues o he slopes a e close o 1. Only 27 wiligh s ou o 101 p esen slopes
smalle han 0.97 o Top 4 (450nm) channel, and 15 ou o 101 o Top 8 (950nm) channel. The pa icula
condi ions which led o his dec ease a e unclea and will be discussed in Sec ion3.3.
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License

Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
6 o 18
3. A e he second pe iod, we see a dec ease in he slopes de i ed om bo h channels, and hus an inc ease
in he numbe o wiligh s wi h he p esence o high-al i ude ALs. In gene al, he slopes du ing his pe iod
a e smalle han hose o he i s pe iod and dec ease wi h Ls (be o e eaching he minimum), indica ing
highe ae osol al i udes o opaci ies i we assume ha he slope in ou co ela ion plo s dec ease wi h hese
wo pa ame e s ( his will be demons a ed in he nex sec ion). The dec ease in he co ela ion slopes las s
up o Ls∼150°, which is when he minimum is ound. Du ing his pe iod o maximum high-al i ude ae osol
ac i i y (be ween Ls=114° and 162°) is when he came as o Pe se e ance o e and MEDA de ec ed he
Figu e 3. (a) Co ela ion slopes de i ed om he p ocedu e desc ibed in Figu e2 and he Radia ion and Dus Senso (RDS)
Top 4 (blue squa es) and Top 8 ( ed squa es) obse a ions o he wiligh s co e ed by RDS up o sol 492. The black dashed
line indica es he ime when a egional dus s o m was obse ed in Jeze o (Lemmon, Smi h, e al.,2022), and he black solid
line shows he 0.97 h eshold alue. (b) colo index signals used o disc imina e be ween dus and ice a e compu ed om he
a io be ween he RDS Top 3 (250–400nm) and Top 6 (750nm) obse a ions (blue do s) and om he a io be ween he RDS
Top 4 (450) and Top 6 (750nm) obse a ions ( ed do s).
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
7 o 18
o ma ion o a 22° sca e ing halo (Lemmon, Toledo, e al.,2022) a ound he Sun, Ls=142° and when a
egional dus s o m (MY36/2022A) was ac i ely aising dus in Jeze o c a e (Lemmon, Smi h, e al.,2022;
Sánchez-La ega e al.,2022; M. D. Smi h e al.,2023), Ls=153°–156° (indica ed by he black dashed line).
4. Fo Ls>∼156° (a e he dus s o m), he slopes become close o 1 bu wi h alues smalle han du ing he
second pe iod and highly a iable. The a iabili y du ing he las pe iod could indica e a change in he kind o
ae osol (ice o dus ) p esen in he highe pa s o he a mosphe e.
Al hough he esul s shown in Figu e3a may indica e a change in he ae osol- ype p esen du ing wiligh a e
he egional dus s o m MY36/2022A (dashed black line in Figu e3a), om he co ela ion slopes we canno
di ec ly disc imina e be ween high-al i ude de ached dus laye s o clouds. Indeed, in bo h scena ios he RDS
obse a ions a 450 and 950nm would show an inc ease espec o he e e ence signals, and hus ha e a slope
lowe han 1 in he co ela ion plo s. To e alua e he possible ae osol composi ion o he di e en de ec ed high
ALs, Figu e3b shows he a ios Top 3 (250–400nm)/Top 6 (750nm) and Top 4 (450nm)/Top 6 (750nm) (CI
de ined in Sec ion2.2) o he same wiligh s analyzed in Figu e3a. Fo each wiligh , we ep esen he a io
alues gi en a SZA∼90°. F om app oxima ely Ls=39° un il Ls=150°, which is wi hin he ACB season, he
CI does no show s ong a ia ions. A he ime a ound when he egional dus s o m passed o e he Pe se e ance
o e si e, he CI dec eased by a ac o o 2 in a ew sols. This is consis en wi h he p esence o dus laye s a high
al i ude and he inc ease in he dus opaci y as a esul o he dus s o m. In e es ingly, al hough he CI inc eased
once he dus s o m had anished, i ne e eco e ed he alues egis e ed be o e. Mo eo e , he CI ime-se ies
clea ly shows a nega i e end a e Ls∼172°. The e o e, based on hese esul s we can iden i y wo pe iods
wi h di e en ae osol scena ios du ing wiligh : (a) a i s pe iod om Ls∼39° o ∼150° wi h high and s able CI
alues likely p oduced by he p esence o p edominan ly wa e ice; (b) a second pe iod wi h lowe CI alues ha
a e dec easing wi h ime, mainly domina ed by dus . This is also consis en wi h he esul s using he obse a ions
made by MEDA-TIRS and epo ed in M. D. Smi h e al.(2023) ha a sys ema ic change in he diu nal end o
he ae osol opaci y occu ed a ound Ls 150°. In ha wo k, a diu nal and a seasonal componen in he ae osol
opaci y a iabili y was de i ed, and om ha i was in e ed ha a e he egional dus s o m he dus was he
ae osol domina ing he opaci y.
On he basis o hese esul s, we conclude ha high-al i ude ALs ound in he co ela ion slopes be o e Ls∼150°
we e mainly made o ice pa icles, while he cases a e ha da e co esponded o ALs whose opaci y was domi-
na ed by dus . F om he analysis o he co ela ion slopes and he CI we canno in e whe he he obse ed ice
pa icles consis ed o clouds (de ached laye s a a gi en al i ude) o hazes e ically ex ended o e se e al km.
In he nex sec ion, we will make use o RT simula ions o cons ain he cloud p ope ies o he de ec ion cases
be o e Ls 150°. As indica ed be o e, he CI alues a e also sensi i e o a ia ions in he cloud pa icle adius.
None heless, because he dec ease in CI coincides wi h he end o he ACB season and ou RT simula ions do no
indica e a sys ema ic change in he pa icle adius a ound Ls 150°, we conclude ha he d op in CI is p ima ily
due o he ae osol composi ion.
3.2. Cloud Al i ude, Opaci y, and Pa icle Size Re ie als
The cloud al i ude, numbe densi y and pa icle size we e de i ed by i ing he RDS Top 4 (450) and Top 8
(950nm) wiligh obse a ions simul aneously wi h he model desc ibed in Sec ion2.3. The cloud opaci y a
each wa eleng h is de i ed om he i ed cloud numbe densi y and he pa icle c oss sec ion, compu ed om
he i ed e and he e ac i e index o wa e ice. Only he wiligh s o which he RDS obse a ions co e ed he
minimum SZA ange o (91°–97°) we e conside ed in he analysis (a o al o 161 wiligh s). In AppendixA we
demons a e ha o clouds abo e ∼30km, ou e ie als a e no signi ican ly a ec ed by he e ical ex ension
o he main dus laye ( o his eason and o dec ease he numbe o ee pa ame e s, ou analysis is ocused on
al i udes abo e 30km). Assuming ha he e we e no de ached dus laye s abo e 25km in ou obse a ions o
he cloudy pe iod, we used a Zmax=45km o he dus p o iles. We did no ind signi ican di e ences in ou
cloud e ie als by a ying his pa ame e om 30 o 50km. We also pe o med a sensi i i y analysis o he dus
opaci y and e , de ailed in AppendixB, o e alua e he impac o hese pa ame e s on he cloud e ie als. We
ound ha o opaci y and e alues be ween 0.3 and 0.6, and be ween 1.2 and 1.4μm, espec i ely, ou cloud
e ie als a e no signi ican ly a ec ed. Fo his eason and based on he imes se ies o he dus opaci y e ie ed
om images aken egula ly by SkyCam (see AppendixB), hese pa ame e s a e ixed o 0.4 and 1.4μm. Rega d-
ing he cloud pa icle shape, we in es iga ed he impac in he RDS signals when using di e en shapes o he
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
8 o 18
han sphe es ( o see i adding an addi ional ee pa ame e was needed). In pa icula , simila simula ions we e
made bu using sphe oid and cylind ical pa icles and we did no ind signi ican a ia ions (see AppendixC o
mo e in o ma ion). The e o e, he only ee pa ame e s in ou in e sion analysis a e he cloud al i ude, opaci y
and pa icle adius.
Figu e4 shows h ee o ou bes i s o he da a acqui ed by Top-4 and Top-8 senso s du ing wiligh o sol
78, 99, and 292, and whe e we can see ha we ma ch he no malized signals e y well ( educed Chi-squa e
unc ion, χ
2, alues<1.2). Fo hese cases, he co ela ion slopes a e 0.97, 1.01, and 0.92 a 450nm, and 0.96,
0.99, and 0.95 a 950nm, espec i ely. Fo sol 99, whose slope is ∼1, he obse a ions could be i ed wi hou
a cloud in he RT model. This is consis en wi h ou assump ion in Sec ion3.1 ha slopes ∼1 a e indica i e o
skies ee o high-al i ude ALs. Fo he i ed signals o sol 78, whose in ensi ies a e g ea e han hose on sol
99 o he same SZAs, we used he cloud model desc ibed abo e and we de i ed a cloud al i ude, opaci y and e
o 42.8±4.1km, 0.011±0.004 and 1.14±0.21μm, espec i ely. We o iginally a emp ed o i he da a using
only he dus Con a h p o iles, wi h Zmax ea ed as a ee pa ame e , bu could no achie e a i wi h a easonably
good χ
2 using his model ( educed χ
2≫1). Howe e , o he same obse a ions (sol 78), a educed χ
2 simila
o ha ob ained wi h he de ached cloud model was achie ed by using a cloud e ically ex ended o e se e al
kilome e s and wi h e <0.4μm. In his ae osol model, e e ed as he haze model he ea e , he laye o ice
pa icles is ex ended o e 30km (o mo e) and cen e ed a an al i ude o 40km. The e o e, o hese pa icula
obse a ions we could no in e i he de ia ion wi h espec o he e e ence signals was p oduced by he p esence
o de ached clouds o hazes. Fo he h ee wiligh s, he obse a ions made on sol 292 show he highe de ia ions
wi h espec o he e e ence signals, and he cloud model p o ides i ed cloud al i ude, opaci y and e alues o
42.0±2.0km, 0.031±0.006 and 1.34±0.52μm, espec i ely. Fo hese obse a ions, nei he he haze model
no he dus model could i he da a wi h a easonably good χ
2, hus indica ing unequi ocally he p esence o
clouds. Al hough he opaci ies de i ed o sol 78 and 292 a e small, i is impo an o no e ha hese opaci ies
ep esen he a e age o e he senso 's FOV. I , du ing he de ec ion, he clouds co e ed only a ew pe cen o
he FOV, hen ou e ie ed cloud opaci y would be smalle han ha de i ed om an ins umen (e.g. a came a)
whose FOV is ully co e ed by he cloud. Ano he poin o no e is ha om his analysis we can only in e he
cloud o haze opaci y abo e ∼30km (see AppendixA), and hus he opaci y con ibu ion om clouds o hazes
below his le el a e no included in he cloud opaci y e ie als.
A simila analysis was pe o med o all he wiligh s co e ed by MEDA up o Ls=150°, which is he ime when
he d op in he CI is obse ed (Figu e3). Fo he comple e da a se (a o al o 161 wiligh s analyzed wi h he RT
model), he signals shown in Figu e4 a e ep esen a i e examples. In 54 wiligh s, RDS obse a ions indica ed
he p esence o clouds, whose i ed pa ame e s a e displayed in he le panels o Figu e5 (a, c, and e). In hese
cases, he cloud model achie ed educed χ
2 alues<1.2, and simila esul s we e no ob ained (in e ms o χ
2)
by eplacing he cloud laye by a e ically ex ended haze. Tha is o say, hese cases a e like he wiligh on sol
292 analyzed in Figu e4. On he o he hand, o a o al o 40 wiligh s, we ound ha bo h he cloud and haze
models i ed he da a wi h educed χ
2 alues<1.2. The esul s ob ained o hese cases using he cloud model
Figu e 4. Compa ison be ween simula ions and obse a ions a 450 (le ) and 950nm ( igh ) o he wiligh s o sols 78, 99,
and 292. The shaded a eas ep esen he e o s and he ed dashed lines he simula ions using he cloud pa ame e s i ed o
each case. Fo each wiligh , he obse a ions a 450 and 950nm we e i ed simul aneously.
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
9 o 18
a e displayed in he igh panels o Figu e5 (b, d, and ). Fo he es o he wiligh s, we ha e: (a) 42 cases wi h
cloud opaci ies below ou limi o de ec ion (∼0.004), de ined in his wo k as he minimum opaci y o p oduce a
a ia ion o a leas 5% espec o he e e ence signals a SZA=93°; (b) 25 cases o which none o he models
achie ed a good i (likely due o changes in he cloud opaci y du ing he wiligh pe iod o o complex ae osol
scena ios). The e o e, in he 58% o he wiligh s analyzed we ound signa u es o clouds o hazes in he RDS
signals. In mos o he cases he clouds we e ound a al i udes be ween 40 and 50km. Based on hese al i udes,
we assume hese clouds a e made o wa e ice. Howe e , we poin ha om he modeling o he RDS signals, we
canno di ec ly disc imina e be ween clouds made o CO2 o H2O ice. The e o e, we can no ule ou he possibil-
i y ha some o he clouds shown in Figu e5 a e made o CO2 ice (in pa icula hose wi h he highes al i udes).
In gene al, he cloud pa icle sizes we e in he ange be ween e =0.6 and 2μm (accoun ing o he e o s in his
Figu e 5. Cloud al i ude, opaci y, and e ec i e adius ( e ) e ie ed om Radia ion and Dus Senso (RDS) Top-4 and Top-8 wiligh obse a ions up o Ls=150°
using he cloud model desc ibed in Sec ion2.3. The le panels (a, c, and e) ep esen he cloud cases o which only he cloud model could i he da a wi h a educed
χ
2<1.2, while he igh ones (b, d, and ) he cases o which bo h he cloud and haze models achie ed i s wi h educed χ
2<1.2. The pu ple do s indica e he wiligh
Ls da es o which he minimum sola zeni h angles ange (91°–97°) was co e ed by he RDS obse a ions.
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
16 o 18
Da a A ailabili y S a emen
All Pe se e ance da a used in his s udy a e publicly a ailable ia he Plane a y Da a Sys em (Rod iguez-Man edi
& de la To e Jua ez,2021). The slope and CI analyses, adia i e ans e simula ions, cloud e ie als, empe a-
u es, and MCD da a o Figu es1-8,A1,B1,B2,C1, andC2 a e a ailable in an a chi e loca ed a Toledo(2023).
Re e ences
Ande son, E., & Leo y, C. (1978). Ma ine 9 ele ision limb obse a ions o dus and ice hazes on Ma s. Jou nal o he A mosphe ic Sciences,
35(4), 723–734. h ps://doi.o g/10.1175/1520-0469(1978)035<0723:m lood>2.0.co;2
Apes igue, V., Gonzalo, A., Jiménez, J. J., Boland, J., Lemmon, M., de Mingo, J. R., e al. (2022). Radia ion and dus senso o Ma s en i on-
men al dynamic analyze onboa d M2020 o e . Senso s, 22(8), 2907. h ps://doi.o g/10.3390/s22082907
A wood, S. A., Smi h, M. D., Bad i, K., Edwa ds, C. S., Ch is ensen, P.R., Wol , M. J., e al. (2022). Diu nal a iabili y in emi s day ime
obse a ions o wa e ice clouds du ing Ma s aphelion-season. Geophysical Resea ch Le e s, 49(15), e2022GL099654. h ps://doi.
o g/10.1029/2022gl099654
Benson, J. L., Kass, D. M., & Kleinböhl, A. (2011). Ma s' no h pola hood as obse ed by he Ma s clima e sounde . Jou nal o Geophysical
Resea ch, 116(E3), E03008. h ps://doi.o g/10.1029/2010je003693
Benson, J. L., Kass, D. M., Kleinböhl, A., McCleese, D. J., Scho ield, J. T., & Taylo , F. W. (2010). Ma s' sou h pola hood as obse ed by he
Ma s clima e sounde . Jou nal o Geophysical Resea ch, 115(E12), E12015. h ps://doi.o g/10.1029/2009je003554
Clancy, R. T., G ossman, A., Wol , M., James, P., Rudy, D., Billawala, Y., e al. (1996). Wa e apo sa u a ion a low al i udes a ound Ma s
aphelion: A key o Ma s clima e? Ica us, 122(1), 36–62. h ps://doi.o g/10.1006/ica .1996.0108
Clancy, R. T., Wol , M. J., & Ch is ensen, P.R. (2003). Ma s ae osol s udies wi h he MGS TES emission phase unc ion obse a ions: Op ical
dep hs, pa icle sizes, and ice cloud ypes e sus la i ude and sola longi ude. Jou nal o Geophysical Resea ch, 108(E9), 5098. h ps://doi.
o g/10.1029/2003je002058
Clancy, R. T., Wol , M. J., Whi ney, B. A., Can o , B. A., & Smi h, M. D. (2007). Ma s equa o ial mesosphe ic clouds: Global occu ence and
physical p ope ies om Ma s global su eyo he mal emission spec ome e and Ma s o bi e came a limb obse a ions. Jou nal o Geophys-
ical Resea ch, 112(E4), E04004. h ps://doi.o g/10.1029/2006je002805
Con a h, B. J. (1975). The mal s uc u e o he Ma ian a mosphe e du ing he dissipa ion o he dus s o m o 1971. Ica us, 24(1), 36–46. h ps://
doi.o g/10.1016/0019-1035(75)90156-6
Fedo o a, A. A., Mon messin, F., Ko able , O., Luginin, M., T okhimo skiy, A., Belyae , D. A., e al. (2020). S o my wa e on Ma s: The dis i-
bu ion and sa u a ion o a mosphe ic wa e du ing he dus y season. Science, 367(6475), 297–300. h ps://doi.o g/10.1126/science.aay9522
Fo ge , F., Hou din, F., Fou nie , R., Hou din, C., Talag and, O., Collins, M., e al. (1999). Imp o ed gene al ci cula ion models o he Ma ian
a mosphe e om he su ace o abo e 80 km. Jou nal o Geophysical Resea ch, 104(E10), 24155–24175. h ps://doi.o g/10.1029/1999je001025
Fo ge , F., & Pie ehumbe , R. T. (1997). Wa ming ea ly Ma s wi h ca bon dioxide clouds ha sca e in a ed adia ion. Science, 278(5341),
1273–1276. h ps://doi.o g/10.1126/science.278.5341.1273
Giu anna, M., Wolkenbe g, P., G assi, D., A onica, A., Aoki, S., Scaccaba ozzi, D., e al. (2021). The cu en wea he and clima e o Ma s:
12 yea s o a mosphe ic moni o ing by he plane a y Fou ie spec ome e on Ma s exp ess. Ica us, 353, 113406. h ps://doi.o g/10.1016/j.
ica us.2019.113406
Gomez-Ma in, L., Toledo, D., P ados-Roman, C., Adame, J. A., Ochoa, H., & Yela, M. (2021). Pola s a osphe ic clouds de ec ion a Belg ano
ii An a c ic s a ion wi h isible g ound-based spec oscopic measu emen s. Remo e Sensing, 13(8), 1412. h ps://doi.o g/10.3390/ s13081412
Guzewich, S. D., & Smi h, M. (2019). Seasonal a ia ion in Ma ian wa e ice cloud pa icle size. Jou nal o Geophysical Resea ch: Plane s,
124(2), 636–643. h ps://doi.o g/10.1029/2018je005843
Hea ens, N., Richa dson, M., Kleinböhl, A., Kass, D., McCleese, D., Abdou, W., e al. (2011a). Ve ical dis ibu ion o dus in he Ma ian
a mosphe e du ing no he n sp ing and summe : High-al i ude opical dus maximum a no he n summe sols ice. Jou nal o Geophysical
Resea ch, 116(E1), E01007. h ps://doi.o g/10.1029/2010je003692
Hea ens, N., Richa dson, M., Kleinböhl, A., Kass, D., McCleese, D., Abdou, W., e al. (2011b). The e ical dis ibu ion o dus in he Ma ian
a mosphe e du ing no he n sp ing and summe : Obse a ions by he Ma s clima e sounde and analysis o zonal a e age e ical dus p o iles.
Jou nal o Geophysical Resea ch, 116(E4), E04003. h ps://doi.o g/10.1029/2010je003691
Figu e C2. Radia ion and Dus Senso Top-8 signals simula ed o a cloud laye a an al i ude o 40km and opaci y o 0.005,
and he phase unc ions illus a ed in Figu eC1. Fo he dus pa icles, we used he model desc ibed in Sec ion2.3 (o in
Figu e4).
Acknowledgmen s
This wo k has been unded by he
Spanish Minis y o Economy and
Compe i i eness, h ough he p ojec s
no. ESP2014-54256-C4-1-R (also
ESP2014-54256-C4-2-R, ESP2014-
54256-C4-3-R, and ESP2014-54256-
C4-4-R), Spanish Minis y o Science,
Inno a ion and Uni e si ies, p ojec s
no. ESP2016-79612-C3-1-R (also
ESP2016-79612-C3-2-R and ESP2016-
79612-C3-3-R), Spanish Minis y o
Science and Inno a ion/S a e Agency
o Resea ch (10.13039/501100011033),
p ojec s no. PID2021-126719OB-C41,
ESP2016-80320-C2-1-R, RTI2018-
098728-B-C31 (also RTI2018-098728-
B-C32 and RTI2018-098728-B-C33),
RTI2018-099825-B-C31. RH and ASL
we e suppo ed by he Spanish p ojec
PID2019-109467GB-I00 unded by
MCIN/AEI/10.13039/50110001103 and
by G upos Gobie no Vasco IT1742-22.
The US co-au ho s pe o med hei wo k
unde sponso ship om NASA’s Ma s
2020 p ojec , om he Game Changing
De elopmen p og amme wi hin he
Space Technology Mission Di ec o-
a e and om he Human Explo a ion
and Ope a ions Di ec o a e. Pa o
his esea ch was ca ied ou a he
Je P opulsion Labo a o y, Cali o nia
Ins i u e o Technology, unde a con ac
wi h he Na ional Ae onau ics and Space
Adminis a ion (80NM0018D0004).
G.M. acknowledges JPL unding om
USRA Con ac Numbe 1638782. ML
is suppo ed by con ac 15-712 om
A izona S a e Uni e si y and 1607215
om Cal ech-JPL. A. V-R. is suppo ed
by he Comunidad de Mad id P ojec
S2018/NMT-4291 (TEC2SPACE-CM).
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License

Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
17 o 18
He nández-Be nal, J., Sánchez-La ega, A., del Río-Gaz elu u ia, T., Hueso, R., Ra anis, E., Ca desín-Moinelo, A., e al. (2021). A long- e m
s udy o Ma s mesosphe ic clouds seen a wiligh based on Ma s exp ess VMC images. Geophysical Resea ch Le e s, 48(7), e2020GL092188.
h ps://doi.o g/10.1029/2020GL092188
Jaquin, F., Gie asch, P., & Kahn, R. (1986). The e ical s uc u e o limb hazes in he Ma ian a mosphe e. Ica us, 68(3), 442–461. h ps://doi.
o g/10.1016/0019-1035(86)90050-3
Lemmon, M. T., Smi h, M. D., Viudez-Mo ei as, D., de la To e-Jua ez, M., Vicen e-Re o illo, A., Mungui a, A., e  al. (2022). Dus ,
sand, and winds wi hin an ac i e Ma ian s o m in Jeze o C a e . Geophysical Resea ch Le e s, 49(17), e2022GL100126. h ps://doi.
o g/10.1029/2022gl100126
Lemmon, M. T., Toledo, D., Apes igue, V., A uego, I., Wol , M. J., Pa el, P., e  al. (2022). Hexagonal p isms o m in wa e -ice clouds
on Ma s, p oducing halo displays seen by pe se e ance o e . Geophysical Resea ch Le e s, 49(17), e2022GL099776. h ps://doi.
o g/10.1029/2022gl099776
Lemmon, M. T., Wol , M. J., Bell, J. F., III, Smi h, M. D., Can o , B. A., & Smi h, P.H. (2015). Dus ae osol, clouds, and he a mosphe ic op i-
cal dep h eco d o e 5 Ma s yea s o he Ma s explo a ion o e mission. Ica us, 251, 96–111. h ps://doi.o g/10.1016/j.ica us.2014.03.029
Lo enz, R. D., Lemmon, M. T., Maki, J., Ban ield, D., Spiga, A., Cha alambous, C., e al. (2020). Scien i ic obse a ions wi h he insigh sola
a ays: Dus , clouds, and eclipses on Ma s. Ea h and Space Science, 7(5), e2019EA000992. h ps://doi.o g/10.1029/2019ea000992
Mää änen, A., Mon messin, F., Gonde , B., Schol en, F., Ho mann, H., González-Galindo, F., e al. (2010). Mapping he mesosphe ic CO2
clouds on Ma s: MEx/omega and MEx/HRSC obse a ions and challenges o a mosphe ic models. Ica us, 209(2), 452–469. h ps://doi.
o g/10.1016/j.ica us.2010.05.017
Mää änen, A., Pé o , K., Mon messin, F., & Haucheco ne, A. (2013). Mesosphe ic clouds on Ma s and on Ea h. Compa a i e Clima ology o
Te es ial Plane s, 393–413. h ps://doi.o g/10.2458/azu_uap ess_9780816530595-ch16
Mää änen, A., Vehkamäki, H., Lau i, A., Me ikallio, S., Kauhanen, J., Sa ijä i, H., & Kulmala, M. (2005). Nuclea ion s udies in he Ma ian
a mosphe e. Jou nal o Geophysical Resea ch, 110(E2), E02002. h ps://doi.o g/10.1029/2004je002308
Madeleine, J.-B., Fo ge , F., Millou , E., Mon abone, L., & Wol , M. (2011). Re isi ing he adia i e impac o dus on Ma s using he LMD
global clima e model. Jou nal o Geophysical Resea ch, 116(E11), E11010. h ps://doi.o g/10.1029/2011je003855
Mal aglia i, L., Mon messin, F., Fedo o a, A., Ko able , O., Fo ge , F., & Be aux, J.-L. (2011). E idence o wa e apo in excess o sa u a ion
in he a mosphe e o Ma s. Science, 333(6051), 1868–1871. h ps://doi.o g/10.1126/science.1207957
McCleese, D., Hea ens, N., Scho ield, J., Abdou, W., Band ield, J., Calcu , S., e al. (2010). S uc u e and dynamics o he Ma ian lowe and
middle a mosphe e as obse ed by he Ma s clima e sounde : Seasonal a ia ions in zonal mean empe a u e, dus , and wa e ice ae osols.
Jou nal o Geophysical Resea ch, 115(E12), E12016. h ps://doi.o g/10.1029/2010je003677
McConnochie, T., Bell, J., III, Sa ansky, D., Wol , M., Toigo, A., Wang, H., e al. (2010). Themis- is obse a ions o clouds in he Ma ian
mesosphe e: Al i udes, wind speeds, and decame e -scale mo phology. Ica us, 210(2), 545–565. h ps://doi.o g/10.1016/j.ica us.2010.07.021
Mishchenko, M. I. (1991). Ligh sca e ing by andomly o ien ed axially symme ic pa icles. Jou nal o he Op ical Socie y o Ame ica A, 8(6),
871–882. h ps://doi.o g/10.1364/josaa.8.000871
Mishchenko, M. I., & T a is, L. D. (1994). T-ma ix compu a ions o ligh sca e ing by la ge sphe oidal pa icles. Op ics Communica ions,
109(1–2), 16–21. h ps://doi.o g/10.1016/0030-4018(94)90731-5
Mon abone, L., Spiga, A., Kass, D. M., Kleinböhl, A., Fo ge , F., & Millou , E. (2020). Ma ian yea 34 column dus clima ology om Ma s
clima e sounde obse a ions: Recons uc ed maps and model simula ions. Jou nal o Geophysical Resea ch: Plane s, 125(8), e2019JE006111.
h ps://doi.o g/10.1029/2019je006111
Mon messin, F., Fo ge , F., Rannou, P., Cabane, M., & Habe le, R. M. (2004). O igin and ole o wa e ice clouds in he Ma ian wa e cycle as
in e ed om a gene al ci cula ion model. Jou nal o Geophysical Resea ch, 109(E10), E10004. h ps://doi.o g/10.1029/2004je002284
Mon messin, F., Quéme ais, E., Be aux, J.-L., Ko able , O., Rannou, P., & Lebonnois, S. (2006). S ella occul a ions a UV wa eleng hs by
he SPICAM ins umen : Re ie al and analysis o Ma ian haze p o iles. Jou nal o Geophysical Resea ch, 111(E9), E09S09. h ps://doi.
o g/10.1029/2005je002662
Mon messin, F., Rannou, P., & Cabane, M. (2002). New insigh s in o Ma ian dus dis ibu ion and wa e -ice cloud mic ophysics. Jou nal o
Geophysical Resea ch, 107(E6), 4-1–4-14. h ps://doi.o g/10.1029/2001JE001520
Mungui a, A., Hueso, R., Sánchez-La ega, A., To e-Jua ez, M. D. L., Ma ínez, G., Newman, C., e al. (2022). Nea su ace a mosphe ic
empe a u es a Jeze o om Ma s 2020 MEDA measu emen s. Jou nal o Geophysical Resea ch: Plane s, 128, e2022JE007559. h ps://doi.
o g/10.1029/2022JE007559
Na a o, T., Madeleine, J.-B., Fo ge , F., Spiga, A., Millou , E., Mon messin, F., & Mää änen, A. (2014). Global clima e modeling o he
Ma ian wa e cycle wi h imp o ed mic ophysics and adia i ely ac i e wa e ice clouds. Jou nal o Geophysical Resea ch: Plane s, 119(7),
1479–1495. h ps://doi.o g/10.1002/2013je004550
Pollack, J. B., & Cuzzi, J. N. (1980). Sca e ing by nonsphe ical pa icles o size compa able o a wa eleng h: A new semi-empi ical
heo y and i s applica ion o oposphe ic ae osols. Jou nal o he A mosphe ic Sciences, 37(4), 868–881. h ps://doi.
o g/10.1175/1520-0469(1980)037<0868:sbnpos>2.0.co;2
Rannou, P., Pe ie , S., Be aux, J.-L., Mon messin, F., Ko able , O., & Rebe ac, A. (2006). Dus and cloud de ec ion a he Ma s limb wi h UV
sca e ed sunligh wi h SPICAM. Jou nal o Geophysical Resea ch, 111(E9), E09S10. h ps://doi.o g/10.1029/2006je002693
Rannou, P., Toledo, D., La as, P., D’a e sa, E., Mo iconi, M., Ad iani, A., e al. (2016). Ti an's su ace spec a a he Huygens landing si e and
Shang i-la. Ica us, 270, 291–306. h ps://doi.o g/10.1016/j.ica us.2015.09.016
Rod iguez-Man edi, J. A., & de la To e Jua ez, M. (2021). Ma s 2020 pe se e ance o e Ma s en i onmen al dynamics analyze (MEDA)
expe imen da a eco d (EDR) and educed da a eco d (RDR) da a p oduc s a chi e bundle [Da ase ]. PDS A mosphe es Node. Re ie ed om
h ps://a cna .psi.edu/u n:nasa:pds:ma s2020_meda:b owse_skycam
Rod iguez-Man edi, J. A., De la To e Juá ez, M., Alonso, A., Apés igue, V., A uego, I., A ienza, T., e al. (2021). The Ma s en i onmen al
dynamics analyze , MEDA. A sui e o en i onmen al senso s o he Ma s 2020 mission. Space Science Re iews, 217(3), 1–86. h ps://doi.
o g/10.1007/s11214-021-00816-9
Rod iguez-Man edi, J. A., de la To e Jua ez, M., Sanchez-La ega, A., Hueso, R., Ma inez, G., Lemmon, M., e al. (2023). The di e se me eo ol-
ogy o jeze o c a e o e he i s 250 sols o pe se e ance on Ma s. Na u e Geoscience, 16, 1–10. h ps://doi.o g/10.1038/s41561-022-01084-0
Sánchez-La ega, A., Chen-Chen, H., O donez-E xebe ia, I., Hueso, R., del Rio-Gaz elu u ia, T., Ga o, A., e al. (2018). Limb clouds and dus
on Ma s om images ob ained by he isual moni o ing came a (VMC) onboa d Ma s exp ess. Ica us, 299, 194–205. h ps://doi.o g/10.1016/j.
ica us.2017.07.026
Sánchez-La ega, A., Rio-Gaz elu u ia, T. D., Hueso, R., Juá ez, M. D. L. T., Ma ínez, G., Ha i, A.-M., e al. (2022). Ma s 2020 pe se e -
ance o e s udies o he Ma ian a mosphe e o e Jeze o om p essu e measu emen s. Jou nal o Geophysical Resea ch: Plane s, 128,
e2022JE007480. h ps://doi.o g/10.1029/2022JE007480
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
Jou nal o Geophysical Resea ch: Plane s
TOLEDO ETAL.
10.1029/2023JE007785
18 o 18
Smi h, M. D. (2009). Themis obse a ions o Ma s ae osol op ical dep h om 2002–2008. Ica us, 202(2), 444–452. h ps://doi.o g/10.1016/j.
ica us.2009.03.027
Smi h, M. D., Bad i, K., A wood, S. A., Edwa ds, C. S., Ch is ensen, P.R., Wol , M. J., e al. (2022). EMIRS obse a ions o he aphelion-season
Ma s a mosphe e. Geophysical Resea ch Le e s, 49(15), e2022GL099636. h ps://doi.o g/10.1029/2022gl099636
Smi h, M. D., Ma ínez, G. M., Sebas ián, E., Lemmon, M. T., Wol , M. J., Apés igue, V., e al. (2023). Diu nal and seasonal a ia ions o ae osol
op ical dep h obse ed by MEDA/TIRS a Jeze o C a e , Ma s. Jou nal o Geophysical Resea ch: Plane s, 128(1), e2022JE007560. h ps://
doi.o g/10.1029/2022je007560
Smi h, M. D., Wol , M. J., Clancy, R. T., Kleinböhl, A., & Mu chie, S. L. (2013). Ve ical dis ibu ion o dus and wa e ice ae osols om CRISM
limb-geome y obse a ions. Jou nal o Geophysical Resea ch: Plane s, 118(2), 321–334. h ps://doi.o g/10.1002/jg e.20047
Smi h, P.H., & Lemmon, M. (1999). Opaci y o he Ma ian a mosphe e measu ed by he image o Ma s pa h inde . Jou nal o Geophysical
Resea ch, 104(E4), 8975–8985. h ps://doi.o g/10.1029/1998je900017
S che binine, A., Mon messin, F., Vincendon, M., Wol , M. J., Vals, M., Ko able , O., e al. (2022). A wo Ma ian yea s su ey o wa e ice clouds
on Ma s wi h ACS onboa d TGO. Jou nal o Geophysical Resea ch: Plane s, 127(12), e2022JE007502. h ps://doi.o g/10.1029/2022je007502
Tamppa i, L. K., Zu ek, R., & Paige, D. (2003). Viking-e a diu nal wa e -ice clouds. Jou nal o Geophysical Resea ch, 108(E7), 5073. h ps://
doi.o g/10.1029/2002je001911
Toledo, D. (2023). Twiligh mesosphe ic clouds in Jeze o as obse ed by MEDA adia ion and dus senso (RDS) [Da ase ]. Zenodo. h ps://doi.
o g/10.5281/zenodo.7932426
Toledo, D., A uego, I., Apés igue, V., Jiménez, J., Gómez, L., Yela, M., e al. (2017). Measu emen o dus op ical dep h using he sola i a-
diance senso (SIS) onboa d he exoma s 2016 EDM. Plane a y and Space Science, 138, 33–43. h ps://doi.o g/10.1016/j.pss.2017.01.015
Toledo, D., Rannou, P., Pomme eau, J.-P., & Foujols, T. (2016). The op ical dep h senso (ODS) o column dus opaci y measu emen s and cloud
de ec ion on Ma ian a mosphe e. Expe imen al As onomy, 42(1), 61–83. h ps://doi.o g/10.1007/s10686-016-9500-7
Toledo, D., Rannou, P., Pomme eau, J.-P., Sa kissian, A., & Foujols, T. (2016). Measu emen o ae osol op ical dep h and sub- isual cloud de ec-
ion using he op ical dep h senso (ODS). A mosphe ic Measu emen Techniques, 9(2), 455–467. h ps://doi.o g/10.5194/am -9-455-2016
Wang, H., & Inge soll, A. P. (2002). Ma ian clouds obse ed by Ma s global su eyo Ma s o bi e came a. Jou nal o Geophysical Resea ch,
107(E10), 8-1–8-16. h ps://doi.o g/10.1029/2001JE001815
Wa en, S. G. (1984). Op ical cons an s o ice om he ul a iole o he mic owa e. Applied Op ics, 23(8), 1206–1225. h ps://doi.o g/10.1364/
ao.23.001206
Wes , R., Del Genio, A., Ba ba a, J., Toledo, D., La as, P., Rannou, P., e al. (2016). Cassini imaging science subsys em obse a ions o Ti an's
sou h pola cloud. Ica us, 270, 399–408. h ps://doi.o g/10.1016/j.ica us.2014.11.038
Wol , M. J., Clancy, R. T., Goguen, J. D., Malin, M. C., & Can o , B. A. (2010). Ul a iole dus ae osol p ope ies as obse ed by MARCI.
Ica us, 208(1), 143–155. h ps://doi.o g/10.1016/j.ica us.2010.01.010
Wol , M. J., Clancy, R. T., Kah e, M. A., Habe le, R. M., Fo ge , F., Can o , B. A., & Malin, M. C. (2019). Mapping wa e ice clouds on Ma s
wi h MRO/MARCI. Ica us, 332, 24–49. h ps://doi.o g/10.1016/j.ica us.2019.05.041
Wol , M. J., Fe nando, A., Smi h, M. D., Fo ge , F., Millou , E., A wood, S. A., e al. (2022). Diu nal a ia ions in he aphelion cloud bel as obse ed
by he emi a es explo a ion image (EXI). Geophysical Resea ch Le e s, 49(18), e2022GL100477. h ps://doi.o g/10.1029/2022gl100477
Wol , M. J., Smi h, M., Clancy, R., A idson, R., Kah e, M., Seelos I , F., e al. (2009). Wa eleng h dependence o dus ae osol single sca e ing
albedo as obse ed by he compac econnaissance imaging spec ome e . Jou nal o Geophysical Resea ch, 114(E2), E00D04. h ps://doi.
o g/10.1029/2009je003350
21699100, 2023, 7, Downloaded om h ps://agupubs.onlinelib a y.wiley.com/doi/10.1029/2023JE007785 by Uni e sidad Del Pais Vasco, Wiley Online Lib a y on [22/12/2023]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License