Ambient air particulate total lung deposited surface area (LDSA) levels in urban Europe

Ambien ai pa icula e o al lung deposi ed su ace a ea (LDSA) le els in u ban Eu ope Xiansheng Liu a, ⁎,Hadia ullah Hadia ullah b , Xun Zhang c,d, ⁎,Ped o T eche a a , Ma jan Sa adkoohi a,e , Me i xell Ga cia-Ma lès a, ,C is ina Reche a , Noemí Pé ez a , Da id C.S. Beddows h ,Im e Salma h , Wanda Thén i , Panayio is Kalka ou as j,k , Nikos Mihalopoulos j,k , Ch is oph Hueglin l , Da id C. G een m,n ,Anja H. T empe m , Benjamin Chazeau o,p , G égo y Gille q , Nicolas Ma chand o , Ja kko V. Niemi , Hanna E. Manninen , Ha i Po in , Nadezda Ziko a s , Jakub Ond acek s , Michael No man , Holge Ge wig u , Susanne Bas ian , Maik Me kel w , Kay Weinhold w , And ea Casans x , Juan And és Casque o-Ve a a,x , F ancisco J. Gómez-Mo eno y ,Begoña A íñano y , Ma ia Gini z , E angelia Diapouli z , Suzanne C umey olle aa , Vé onique Ri aul ab , Jean-Eudes Pe i ac , Oli ie Fa ez ad , Jean-Philippe Pu aud ae , Sebas iao Ma ins Dos San os ae , Hilkka Timonen a , Pasi P. Aal o ag , Ta eq Hussein ag,ah , Janne Lampilah i ag , Philip K. Hopke ai ,Al ed Wiedensohle w ,Roy M. Ha ison g,aj , Tuukka Pe äjä ag , Ma co Pandolfi a ,And és Alas uey a , Xa ie Que ol a a Ins i u e o En i onmen al Assessmen and Wa e Resea ch (IDAEA-CSIC), Ba celona, Spain b School o Pha maceu ical Science and Technology, Tianjin Uni e si y, Tianjin, China c Beijing Key Labo a o y o Big Da a Technology o Food Sa e y, School o Compu e Science and Enginee ing, Beijing Technology and Business Uni e si y,Beijing,China d Ho an No mal College. Ho an 848000, Xinjiang, China e Depa men o Mining, Indus ial and ICT Enginee ing (EMIT), Man esa School o Enginee ing (EPSEM), Uni e si a Poli ècnica de Ca alunya (UPC), 08242 Man esa, Spain Depa men o Applied Physics-Me eo ology, Uni e si y o Ba celona, Ba celona, Spain g Di ision o En i onmen al Heal h and Risk Managemen , School o Geog aphy, Ea h and En i onmen al Sciences Uni e si y o Bi mingham, Edgbas on, Bi mingham, Uni ed Kingdom h Ins i u e o Chemis y, Eö ös Lo ánd Uni e si y, Budapes , Hunga y i He esy Gyö gy Ph.D. School o Chemis y, Eö ös Lo ánd Uni e si y, Budapes , Hunga y j En i onmen al Chemical P ocesses Labo a o y, Depa men o Chemis y, Uni e si y o C e e, He aklion, G eece k Ins i u e o En i onmen al Resea ch &Sus ainable De elopmen , Na ional Obse a o y o A hens, A hens, G eece l Labo a o y o Ai Pollu ion and En i onmen al Technology, Swiss Fede al Labo a o ies o Ma e ials Science and Technology (EMPA), Duebendo , Swi ze land m MRC Cen e o En i onmen and Heal h, En i onmen al Resea ch G oup, Impe ial College London, UK n NIHR HPRU in En i onmen al Exposu es and Heal h, Impe ial College London, UK o Aix Ma seille Uni ., CNRS, LCE, Ma seille, F ance p Labo a o y o A mosphe ic Chemis y, Paul Sche e Ins i u e, Villigen, Swi ze land q A moSud, Regional Ne wo k o Ai Quali y Moni o ing o P o ence-Alpes-Cô e-d'Azu , Ma seille, F ance Helsinki Region En i onmen al Se ices Au ho i y (HSY), Helsinki, Finland s Ins i u e o Chemical P ocess Fundamen als, . .i. Academy o Sciences o he Czech Republic Roz ojo a, P ague, Czech Republic En i onmen and Heal h Adminis a ion, SLB-analys, S ockholm, Sweden u Ge man En i onmen Agency (UBA), Dessau-Roßlau, Ge many Saxon S a e O fice o En i onmen , Ag icul u e and Geology (L ULG), D esden, Ge many w Leibniz Ins i u e o T oposphe ic Resea ch (TROPOS), Leipzig, Ge many x Andalusian Ins i u e o Ea h Sys em Resea ch (IISTA-CEAMA), Uni e si y o G anada, G anada, Spain y Depa men o En i onmen , CIEMAT, Mad id, Spain z ENRACT, Ins i u e o Nuclea and Radiological Science &Technology, Ene gy &Sa e y, NCSR Demok i os, 15310 Ag. Pa aske i, A hens, G eece aa Uni . Lille, CNRS, UMR 8518 Labo a oi e d'Op ique A mosphé ique (LOA), Lille, F ance ab IMT No d Eu ope, Ins i u Mines-Télécom, Uni e si é de Lille, Cen e o Ene gy and En i onmen , 59000, Lille, F ance ac Labo a oi e des Sciences du Clima e de l'En i onnemen , CEA/O me des Me isie s, Gi -su -Y e e, F ance ad Ins i u na ional de l'en i onnemen indus iel e des isques (INERIS), Pa c Technologique Ala a BP2, Ve neuil-en-Hala e, F ance ae Eu opean Commission, Join Resea ch Cen e (JRC), Isp a, I aly a A mosphe ic Composi ion Resea ch, Finnish Me eo ological Ins i u e, Helsinki, Finland ag Ins i u e o A mosphe ic and Ea h Sys em Resea ch (INAR), Facul y o Science, Uni e si y o Helsinki, Finland ah En i onmen al and A mosphe ic Resea ch Labo a o y, Depa men o Physics, School o Science, The Uni e si y o Jo dan, Amman 11942, Jo dan ai Depa men o Public Heal h Sciences, Uni e si y o Roches e School o Medicine and Den is y, Roches e , NY, USA aj Depa men o En i onmen al Sciences, Facul y o Me eo ology, En i onmen and A id Land Ag icul u e, King Abdulaziz Uni e si y, Jeddah, Saudi A abia Science o he To al En i onmen 898 (2023) 165466 ⁎Co esponding au ho s. E-mail add esses: liuga @cid.csic.es (X. Liu), zhangxun@b bu.edu.cn (X. Zhang). h p://dx.doi.o g/10.1016/j.sci o en .2023.165466 Recei ed 11 Ap il 2023; Recei ed in e ised o m 16 June 2023; Accep ed 9 July 2023 A ailable online 13 July 2023 0048-9697/© 2023 The Au ho s. Publishedby Else ie B.V. This is an open access a icle unde heCC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4. 0/). Con en s lis s a ailable a ScienceDi ec Science o he To al En i onmen jou nal homepage: www.else ie .com/loca e/sci o en HIGHLIGHTS GRAPHICAL ABSTRACT •Lung deposi ed su ace a ea (LDSA) was in es iga ed in 25 EU and one USA si es. •Peaks o o al LDSA we e eco ded in 06:00–8:00 and 19:00–22:00 UTC in u ban. •The seasonal changes o LDSA showed sig- nifican di e ences in i s le els (p<0.01). •Ai ken and Accumula ion mode pa icles play a significan ole in he o al LDSA. ABSTRACTARTICLE INFO Edi o : P o . Pa los Kassomenos Keywo ds: To al lung deposi ed su ace a ea Pa icle numbe size dis ibu ion Spa ial a iabili y U ban en i onmen T a fic emissions This s udy aims o pic u e he phenomenology o u ban ambien o al lung deposi ed su ace a ea (LDSA) (including head/ h oa (HA), acheob onchial (TB), and al eola (ALV) egions) based on mul iple pa h pa icle dosime y (MPPD)model du ing 2017–2019 pe iod collec ed om u ban backg ound (UB, n= 15), a fic(TR,n= 6), subu ban backg ound (SUB, n= 4), and egional backg ound (RB, n = 1) moni o ing si es in Eu ope (25) and USA (1). B iefly, he spa ial- empo al dis ibu ion cha ac e is ics o he deposi ion o LDSA, including diel, weekly, and seasonal pa - e ns, we e analyzed. Then, he ela ionship be ween LDSA and o he ai quali y me ics a each moni o ing si e was in es iga ed. The esul showed ha he peak concen a ions o LDSA a UB and TR si es a e commonly obse ed in he mo ning (06:00–8:00 UTC) and la e e ening (19:00–22:00 UTC), coinciding wi h a fic ush hou s, biomass bu ning, and a mosphe ic s agna ion pe iods. The only LDSA nigh - ime peaks a e obse ed on weekends. Due o he a iabili y o emission sou ces and me eo ology, he seasonal a iabili y o he LDSA concen a ion e ealed sig- nifican di e ences (p= 0.01) be ween he ou seasons a all moni o ing si es. Meanwhile, he co ela ions o LDSA wi h o he pollu an me ics sugges ed ha Ai ken and accumula ion mode pa icles play a significan ole in he o al LDSA concen a ion. The esul s also indica ed ha he main p opo ion o o al LDSA is a ibu ed o he ALV ac ion (50 %), ollowed by he TB (34 %) and HA (16 %). O e all, his s udy p o ides aluable in o ma ion o LDSA as a p edic o in epidemiological s udies and o he fi s ime p esen ing o al LDSA in a a ie y o Eu opean u ban en i onmen s. 1. In oduc ion Wi h he apid de elopmen o he economy, ene gy consump ion has been on he ise, gene a ing a la ge olume o emissions in o he a mo- sphe e, causing se ious ai pollu ion issues (Khan e al., 2022). The a mo- sphe ic ae osol (pa icula e ma e , PM), cons i u es an impo an ac ion o ai pollu ion wi h majo e ec s on public heal h (Mahowald e al., 2014;Rao e al., 2018;Wehne e al., 2002). The complex chemical compo- si ion o he ae osol pa icles, and hei highly a iable pa icle numbe size dis ibu ions (PNSD) con e physicochemical p ope ies ha migh ha e a ious e ec s on human heal h (Baldau e al., 2016;McMu y, 2000). Epidemiological and oxicological s udies ha e sugges ed ha he smalle he pa icles, he mo e oxic hey migh be, especially ul afine pa - icula e ma e (UFP, o pa icles smalle han 100 nm) (Kwon e al., 2020). Because o hei small size, hese migh accessdeepe pa so he lungs, and e en he bloods eam h ough lung ansloca ion, hus causing sys emic oxici y (Cassee e al., 2019;Salma e al., 2015;Viei a and Kou akis, 2021). These UFPs also ha e a la ge su ace a ea o he same mass han accumula ion and coa se mode pa icles ha e. As a esul , hey can po en- ially ca y mo e ha m ul subs ances in o he al eoliand ci cula o y sys em, ha ing a g ea e oxic e ec han la ge diame e pa icles (Abdillah and Wang, 2022;Chen e al., 2016;Siou as e al., 2005;Sun e al., 2019). Now- adays, a widely used app oach o combine he lung deposi ion and su ace a ea o pa icles is o measu e o calcula e he concen a ions o lung depos- i ed su ace a ea (LDSA). The de e mina ion o LDSA is o high in e es o exposu e assessmen , as i eflec s he concep ha pa icle su ace a ea a ailable in he lung is a ele an exposu e me ic. LDSA has been p oposed as a c i ical p edic o o heal h ou comes om ae osol exposu e, as i appea s o be one o he mos ele an physical me ics o e alua ing exposu e o pa icles (Chang e al., 2022;Schmid and S oege , 2016). In p io s udies, LDSA concen a ions ha e been epo ed o ambien ae osols in di e en en i onmen s and se - e al ci iesin he wo ld. Fo ins ance, ambien measu emen sha e been con- duc ed in Ba celona (Spain) (Reche e al., 2015), Helsinki (Finland) (Fung e al., 2022;Kuulu ainen e al., 2016), Leices e (Hama e al., 2017), Bei- jing (China) (Xu e al., 2022), Augsbu g (Ge many) (Liu e al., 2022), and Minneapolis (USA) (Gonzalez e al., 2022). Howe e , compa a i e da a o o al LDSA concen a ions deposi ed in he all egions o lung o di e - en si e ypologies (u ban backg ound, UB; a fic, TR; subu ban back- g ound, SUB; and egional backg ound, RB, a eas) a e s ill e y limi ed and needed. He ein, we explo ed o al LDSA concen a ions wi hin Eu opean ci ies in he amewo k o he RI-URBANS p ojec (Resea ch In as uc u es Se - ices Rein o cing Ai Quali y Moni o ing Capaci ies in Eu opean U ban & Indus ial A eas). We conduc ed an analysis using hou ly PNSDs measu ed om 25 moni o ing si es in Eu ope and one in he USA be ween 2017 and 2019 ( he compiled da ase s o pa icle numbe size dis ibu ions, SI, Table S1) o es ima e he egional deposi ion o LDSA in he head/ h oa (HA), acheob onchial (TB), and al eola (ALV) egions o he human e- spi a o y sys em. We also in es iga ed he spa ial- empo al dis ibu ion cha ac e is ics o he deposi ion, including diel, weekly, and seasonal pa - e ns. In addi ion, he spa ial and seasonal he e ogenei y le els o he mon- i o ing s a ions we e analyzed using a coe ficien o di e gence (COD). Finally, he ela ionship be ween LDSA and o he ai quali y me ics, X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 2 namely pa icle numbe concen a ion (PNC) in di e en modes (nucle- a ion, Ai ken, accumula ion), black ca bon (BC), gaseous pollu an s (SO 2 , NO, NO 2 ,O 3 , and CO) and PM 1 ,PM 2.5 ,PM 10 a each moni o ing si e was in es iga ed o e eal he essen ial me ics associa ed wi h LDSA, iden i y- ing he mos closely me ics a e associa ed wi h LDSA, and hus p o iding insigh s in o he sou ces and anspo o LDSA and i s po en ial impac s on human heal h and he en i onmen . O e all, his s udy p o ides alu- able in o ma ion o use in an e alua ion o he heal h e ec s o a mo- sphe ic ae osol pa icles, including UFP. 2. Me hods 2.1. Ins umen a ion The ins umen a ion used o measu ing all me ics a he di e en s a- ions is desc ibed in Table S1. Di e en ins umen s and measu ing config- u a ions we e used o PNSD measu emen s a he di e en si es. B iefly, in his s udy he PNSDs we e measu ed by Mobili y Pa icle Size Spec ome- e s (MPSS) (Table S1). BC concen a ions we e moni o ed ei he by a Mul i-Angle Abso p ion Pho ome e (MAAP, The mo Scien ificmodel 5012) o a mul i-wa eleng h Ae halome e (Magee Scien ific, Model AE33/21, Slo enia) (Table S1). To ensu e da a quali y, we conduc ed a numbe o da a QA/QC p ocedu es. We applied a da a comple eness fil e o exclude da a wi h poo a ailabili y. Fu he mo e, we pe o med a senso ag eemen check o ensu e consis ency be ween he MPSS and o he pa i- cle ins umen s such as he Scanning Mobili y Pa icle Size s (SMPS). Concen a ions o PM X (PM 1 ,PM 2.5 ,PM 10 ) and o he abo e e e ed gaseous pollu an s we e measu ed wi h he con en ional ins umen a ion used in Eu ope and he US o ai quali y moni o ing. Specifically, PM X we e measu ed using a Tape ed Elemen Oscilla ing Mic obalance (TEOM, R&P 1400a, The mo Scien ific™), while gaseous pollu an s including NO/NO 2 ,O 3 ,SO 2 , and CO a e gene ally eco ded by comme cial ins umen s wi h he use o he chemiluminescence, UV-abso p ion, UV- fluo escence, and gas fil e co ela ion echniques, espec i ely. 2.2. Calcula ion o he size- ac iona ed lung-deposi ed su ace a ea (LDSA) The me hod commonly used o de e mine he geome ic su ace a ea concen a ion o sphe ical pa icles is based on he measu emen o PNSD (Asbach e al., 2009;Fie z e al., 2011). Pa icles p oduced om combus- ion ypically o m chain-like s uc u es (McDonald and Biswas, 2004), which can ha e a highe po osi y and su ace a ea han sphe ical pa icles. The e o e, he assump ion o sphe ical pa icle shape used in his s udy isan app oxima ion and may unde es ima e he o al su ace a ea. In he case o sphe ical pa icles, he PNSD can be easily con e ed in o su ace a ea size dis ibu ions (SASD) by means o he known ela ionship be ween a sphe e's su ace a eaand i s diame e . These dis ibu ions can be in eg a ed o e he size ange o in e es o ob ain he o al su ace a ea concen a- ions. Following ha , o ob ain LDSA concen a ions, he SASDs can be weigh ed wi h lung deposi ion cu es (Hussein e al., 2013)andin eg a ed o e he size ange o in e es . In his s udy, a model was applied o di ide he espi a o y ac in o h ee main egions, including HA, TB, and ALV e- gions, based on he mul iple pa h pa icle dosime y (MPPD) models (Asgha ian, 2022;Hussein e al., 2013;Hussein e al., 2015). In ou calcu- la ions, we used size-dependen egional deposi ion ac ion (DF) cu es ob ained om MPPD so wa e ( e sion 3.04, Chemical Indus y Ins i u e o Toxicology, Resea ch T iangle Pa k, NC; h ps://www.a a.com/mppd/ ). The o al LDSA alues we e calcula ed by fi s coun ing he numbe o pa icles in each size bin and hen mul iplying he PNC wi h he specificde- posi ion ac ion o he gi en pa icle size. To accu a ely assess he con i- bu ions o LDSA, i is impo an o conside he pa icle sizes in ol ed. As depic ed in Fig. S1, pa icles smalle han 10 nm ha e a ela i ely low weigh in he o e all calcula ion o LDSA con ibu ions (0.14 % ± 0.2 %), compa ed o o he pa icle sizes. In ou s udy we also compa ed he alues o ALV-LDSA by MPPD and lung-deposi ed Nanopa icle Su ace A ea Moni o (NSAM) (TSI, Model 3550, USA). The compa ison showed a e y good ag eemen o he esul s (Pea son's 2 = 0.927, Fig. S2), u he indica ing ha he MPPD can be used o es ima e he LDSA concen a ion. 2.3. S a is ical analysis The me ics a e epo ed as a e age concen a ion (AVE) ± s anda d de ia ion (STD). The co ela ions be ween LDSA and o he me ics o he di e en pe iods we e analyzed by calcula ion o Pea son's co ela ion coe - ficien . The s a is ically significan di e ences in LDSA concen a ions in a di e en ime (diel, weekly, and seasonal) and 26 moni o ing si es we e s udied using he K uskal-Wallis ANOVA on anks (K uskal and Wallis, 1952) and Duncan's Mul iple Range Tes (Duncan, 1955), which was pe - o med using SPSS So wa e (IBM SPSS S a is ics 25, Chicago, IL, USA). Subsequen ly, due o da a on ai pollu an s also exhibi he same cha ac e - is ics as geospa ial da a (i.e., spa ial he e ogenei y) (Yang e al., 2018), a co- e ficien o di e gence (COD) was applied o analyse he spa ial and seasonal he e ogenei y o ambien pollu an s (Fa idi e al., 2019). The COD jk me hod o iden i ying he di e ences was desc ibed in he SI (- Sec ion 1.3). Spa ially, i has been sugges ed ha a low COD jk alue (<0.2) indica es a high le el o homogenei y be ween si es, while a high COD jk alue (>0.2) indica es he e ogeneous si es (Wilson e al., 2005). 3. Resul and discussion 3.1. Gene al cha ac e is ics o o al LDSA a di e en measu emen si es The hou ly o al concen a ion o LDSA was examined o he yea s 2017–2019 using K uskal-Wallis ANOVA analysis and Duncan's me hod. The esul s e ealed ha some si es (e.g., BCN_UB, LEI_UB, and ATH_UB, among o he s) we e significan ly di e en om he o he ones, while some si es' pai s (e.g., ROC_UB and HEL_UB) sugges ed simila beha ious (Table S2 and S3). These di e ences indica e ha he e a e bo h spa ial he - e ogenei y and homogenei y ac oss he a ious moni o ing loca ions. Fu - he mo e, ou esul s show ha he a e age COD jk is 0.32 ± 0.06 a UB s a ions, indica ing he high spa ial he e ogenei y be ween hese moni o - ing si es (COD jk >0.2). The lowe COD jk alues a e ound in he SUB/RB (a e age COD jk , 0.28 ± 0.06) and TR (a e age COD jk , 0.26 ± 0.05) (SI, sec ion 2.1, Fig. S3), sugges ing a ce ain ex en o homogenei y be ween hese moni o ing si e. As a esul , hese findings may be use ul in iden i ying he po en ial sou ces o LDSA and de eloping common bu also si e-specific s a egies o mi iga e i s impac s on human heal h and he en i onmen . In addi ion, compa ing all moni o ing si es (Table S4), o UB a eas, he highes annual a e age o al LDSA concen a ion is ound in BUD_UB (85 ± 53 μm 2 /cm 3 ), ollowed by GRA_UB and MAR_UB (63 ± 47, 63 ± 48 μm 2 /cm 3 ), while he lowes is in ROC_UB (23 ± 15 μm 2 /cm 3 )and HEL_UB (24 ± 16 μm 2 /cm 3 ). Fo TR si e, he highes annual a e age o al LDSA is ound in LDN_TR (68 ± 41 μm 2 /cm 3 ), ollowed by LEI2_TR (65 ± 39 μm 2 /cm 3 ), while he lowes is in STO_TR (30 ± 16 μm 2 /cm 3 ) and HEL_TR (39 ± 26 μm 2 /cm 3 ). Fo SUB a eas, he highes annual a e - age o al LDSA is ound in PRA_SUB (59 ± 81 μm 2 /cm 3 ), ollowed by LIL_SUB (50 ± 34 μm 2 /cm 3 ), while he lowes is in ATH_SUB (43 ± 24 μm 2 /cm 3 ) and PAR_SUB (30 ± 22 μm 2 /cm 3 ). The high STD associa ed wi h he annual a e age o al LDSA concen a ions could be mainly a ib- u ed o diel a ia ions in emissions sou ces, me eo ological condi ions, and anspo pa e ns. The di e ences could pa ly also be influenced by he co - ec ionme hodapplied o pa icledi usion losses in he MPSSs. In Budapes , a conse a i e size-independen me hodisappliedwhich esul edinaco ec- ion ac o o 1.22–1.27 o e he 3 yea s (Salma e al., 2016). In e es ingly, ISP_RB, loca ed in an ai pollu ion ho spo egion o Eu ope ( he Po Valley) also had he second highes o al LDSA concen a ion (75 ± 61 μm 2 /cm 3 ) o all moni o ing si es. O e all, he highes concen a ion o o al LDSA is ound in UB (BUD_UB) and he lowes is also ound in UB (HEL_UB), which is compa able wi h he end o PM 2.5 (Fig. 1, Fig. S4, and Table S2). Mo eo e , he con ibu ions o he deposi ion o annual a e age o al LDSA in di e en egions o he espi a o y ac , including HA, TB, and X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 3 ALV, a e also es ima ed a all esea ch si es du ing he obse ed pe iod (Table S2). The esul s o he o al LDSA concen a ion calcula ions show ha LDSA is p ima ily accumula ed in ALV (50 %), which is conside ed o p esen he g ea es po en ial heal h isk (Fung e al., 2022;Salo e al., 2021a,Salo e al., 2021b), and lowe con ibu ions we e ob ained o HA (16 %) and TB (34 %) deposi ion, consis en wi h he s ong size- dependence o deposi ion (Kuma e al., 2014;Volio is and Sama a, 2018). A p e ious s udy epo ing an es ima ion o he inhaled deposi ed dose a e du ing common exposu e scena ios o UB ae osols in an Eas e n Medi e anean ci y (Amman, Jo dan), epo ed HA, TB, and ALV inhaled deposi ed doses eaching 7–16 %, 16–28 %, and 56–76 %, espec i ely (Hussein e al., 2022). In compa ison, a p io s udy in he Helsinki me o- poli an a ea ound ha he annual mean ALV-LDSA concen a ions a ied be ween 9 and 22 μm 2 /cm 3 a di e en si es (Kuula e al., 2020). The ange was 9–12 μm 2 /cm 3 o u ban backg ound and de ached housing a ea si es, which a e qui e clean a eas in Helsinki (Kuula e al., 2020), o alues up o ~189 μm 2 /cm 3 ypically obse ed nea PM sou ces such as u ban oads in A hens, G eece (Che is anidis e al., 2020). This ange was confi med by ou esul s whe e he ALV-LDSA annual a e ages ange om 12 ± 8 o 41 ± 29 μm 2 /cm 3 , poin ing o a significan he e ogenei y o LDSA ac oss u ban Eu ope. In addi ion, a ecen s udy on he di e en lo- ca ions o BC in he lungs, alsoshowed BCwas majo ly deposi ed in he ALV egion, accoun ing o 49.0–53.2 % o he o al BC deposi ion dose, less in he TB egion (35.6–37.2 %) and e en less in he HA egion, 11.2–13.8 % (Liu e al., 2023). 3.2. To al LDSA ime ends 3.2.1. Diel a ia ions Diel a ia ion in o al LDSA can be used as an indica o o he ole o local human ac i i ies, e ec s o me eo ology and exposu e o ai pollu ion isks (Reddy e al., 2012;Zhou e al., 2018). The esul s showed ha he diel ends o o al LDSA e ealed some simila ea u es among all moni o ing si es (Fig. 2 and Fig. S5). A he UB si es (Fig. 2a and Fig. S5a), he mo ning peak o o al LDSA occu ed a ~06:00–09:00 UTC, app oxima ely one hou a e local sun ise. The o me coincided wi h he mo ning a fic ush hou s (Teinilä e al., 2019). Fo example, he o al LDSA concen a- ions in BUD_UB a e significan ly (p= 0.0001) highe han in he o he ci - ies, which may be closely ela ed o he esiden ial and household emissions, indus ial sou ces, and some o - oad anspo , ypes o ehicles, a ficflow, and long- ange anspo o ai masses in ha ci y (Salma e al., 2017;Salma e al., 2020;Thén and Salma, 2022). A p io s udy also e- po ed ha du ing ush hou , abou 74 % o LDSA is a ibu ed o a fic emi ed pa icles (Chang e al., 2022). Fu he mo e, i is also wo h no ing ha he mo ning peak o LDSA a TR si es s a s ea lie (4:30–6:00 UTC), is mo e p onounced, emains high un il he e ening a fic peak (~19:00 UTC), and hen sligh ly dec eases (Fig. 2c&S5c), u he illus a ing he impac o a fic on o al LDSA (Reche e al., 2015;Wie zbicka e al., 2014). Then, a e he mo ning peak, a UB si es, he o al LDSA is educed due o a dec ease in emissions and ela i e humidi y (RH), and inc eased empe a u e (T), wind speed (WS), and adia ion (RAD) (Fig. S6). The Fig. 1. 2017–2019 a e age concen a ions o o al LDSA a 25 Eu opean si es and one om he USA. X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 4 lowes concen a ions a e usually de ec ed in he midday-a e noon (12:00–16:00 UTC) a all backg ound si es due o less om a fic (o bio- mass bu ning in win e ) o hese hou s, highe WS and inc eased mixing dep h (Fig. S6d, g), hus inc easing dispe sion condi ions. A e 16:00 UTC, which majo ly lea e om wo k, a fic inc eases and a b oad LDSA e ening peak egula ly is obse ed be ween ~19:00–22:00 UTC a UB and TR si es (Fig. 2). Howe e , a HEL_TR, i is ha dly a ec ed by he ush hou a e ening. This di e ence is mainly due o he ac ha he si e is less a ec ed by esiden ial hea ing and he homecoming a fic a e ening (Helin e al., 2018;Hie ikko e al., 2018). I was also ound in he p e ious s udy o diel a ia ion in black ca bon a his si e (Helin e al., 2018). The la e o al LDSA peaks du ing hese hou s migh be linked o inc eased ehicula emissions associa ed wi h e ening a fic ush hou (Chang e al., 2022), cooking (Allan e al., 2010), and biomass bu ning (pa - icula ly du ing win e ) (Ba i e al., 2010;Liu e al., 2022), as well as un a- o able me eo ological condi ions (e.g., low WS and g ea e s abili y and educed mixing dep h, Fig. S6) o dispe sion (Zhu e al., 2018). The sligh dec easing end om midnigh o ea ly mo ning (23:00 o 03:00 UTC, Fig. 2) is a ibu ed o he p og essi e educ ion o an h opogenic emis- sions, including a ficflow. Meanwhile, du ing his pe iod, he e is he po- en ial o hyg oscopic pa icles o abso b wa e when he RH inc eases, which can inc ease he pa icle size, leading o a dec ease in o al LDSA (Li e al., 2021;Wiegand e al., 2011). Fo he SUB and RB si es, he diel a ia ion o o al LDSA is sligh ly di - e en compa ed o UB and TR si es. A low peak o o al LDSAis ound in he mo ning and e ening a fic ush hou s (05:00–06:00 and 19:00–20:00 UTC). A ound hose SUB si es, he e a e smalle oads wi h ligh e a fic. Thus, a hese loca ions, measu emen s a e no significan ly a ec ed by di- ec and local esh a fic emissions and ends a e like he ones epo ed p e iously o his en i onmen (Kuulu ainen e al., 2016). Thus, LDSA ends o be qui e cons an wi h e y na ow a ia ions (±5 μm 2 /cm 3 ) along heday(Fig. 2b). In con as , a ISP_RB, he o al LDSA concen a ion g adually dec eases om 0:00–15:00 UTC and inc eases du ing he e ening (21:00–0:00 UTC), wi h a ± 16 μm 2 /cm 3 a e age hou ly a ia ion, which may be caused by ma ked noc u nal he mal in e sions in he Po Valley, cha ac e ized by ex emely low winds, due o he Alps' p o ec ion agains no he n winds, and high win e ime biomass bu ning domes ic emissions. 3.2.2. Weekly a ia ions The weekly a e age a ia ions o he o al LDSA concen a ions a UB and TR si es (Fig. 3) a e cha ac e ized by highe wo kday (Monday-F iday) alues, when he highes o al LDSA concen a ion is measu ed du ing mo ning and e ening ush hou s (6:00–9:00 and 18:00–21:00 UTC) due o he high ehicle-exhaus emissions (Reche e al., 2015). In he weekend (Sa u day and Sunday), o u ban a eas, he e is a significan dec ease (Fig. 3a, c). Du ing weekend mo nings, he e a e no o weak o al LDSA peaks (Fig. 3a, c) because o he low a fic densi y, excluding GRA_UB si es. Howe e , only a la e peak is ound on Sa u day and Sunday e enings (~20:00–22:00 UTC) compa ed o he co esponding peak on weekdays, due o high ec ea ional mobili y and highe emissions om cooking and Fig. 2. 2017–2019 a e age o no malized hou ly o al LDSA o he 26 s udied si es. a, u ban backg ound (UB) si es; b, subu ban backg ound (SUB) and egional backg ound (ISP_RB) si es; and c, a fic (TR) si es (Min-Max no maliza ion was used in a da ase o all wi hin a specified ange, ypically be ween 0 and 1. I is calcula ed by sub ac ing he minimum alue in he da ase om each alue and di iding by he ange o he da ase ). X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 5 domes ic hea ing when people s ay a home du ing weekends (Allan e al., 2010). Addi ional weekend con ibu ions o o al LDSA s em om e en s such as ba becues and la e-nigh ac i i ies on Sa u days, esul ing in a peak occu ing la e in he day and las ing o longe (Hama e al., 2017). A he SUB/RB si es (Fig. 3b), he weekly diel end o o al LDSA has a weak mo ning peak and a mo e p onounced e ening peak, which is compa- able wi h u ban si es (see he highes a e aged o al LDSA concen a ions on Sa u day midnigh a ISP_RB, Fig. 3b, p obably due o ec ea ional do- mes ic biomass bu ning o longe use o hea ing). 3.2.3. Seasonal a ia ions Fig. 4 shows he 2017–2019 a e age hou ly o al LDSA concen a ions by season (Sp ing, Ma -May; Summe , Jun-Aug; Au umn, Sep-No ; Win e , Dec-Feb) o each si e. O e all, he o al LDSAconcen a ions a all moni o - ing si es showed subs an ial seasonal di e ences (p= 0.05, Table S5). The ex en o which he e we e s a is ically significan di e ences in o al LDSA concen a ions be ween seasons a ied sligh ly ac oss di e en moni o ing si es. Fo example, BUD_UB, ISP_RB, LAN_UB, and MAD_UB we e signifi- can ly di e en be ween he ou seasons (p= 0.0001), while he o al LDSA concen a ions a ATU_SUB, HEL_UB, HEL_TR, LER2_TR, and ROC_UB we e no di e en be ween he au umn and win e seasons, and ATH_SUB, GRA_UB, LDSA concen a ions a he ATH_SUB, GRA_UB, LER2_TR, and PAR_SUB, si es did no di e significan ly be ween he sp ing and summe seasons, u he sugges ing ha LDSA is influenced by local pollu ion sou ces. In addi ion, combined wi h he COD analysis o seasonal a iabili y, he seasonal he e ogeneous among all si es we e mainly ound be ween win e and o he h ee seasons, specially be ween win e and sum- me (0.29 ± 0.12 o SUB, 0.30 ±0.06 o UB, and 0.12 ± 0.04 o TR) (SI, sec ion 2.2, Fig. S7), indica ing ha me eo ological condi ions (e.g., empe a u e) has a s ong influence on ai pollu ion a UB and SUB (COD jk >0.2). Howe e , i is wo h no ing ha a TRsi es, he seasonal a - ia ion o ai pollu ion shows homogenei y (COD jk <0.2), which indica es ha TR si es a e a ela i ely s able sou ce o pollu ion, and he equency and ou es o ehicle ope a ions do no change much, so he s abili y o he pollu ion sou ce may lead o li le di e ence in seasonal a ia ion. In win e , he highes a e age exposu es o LDSA me ics a e ound a ATH_UB, BUD_UB, GRA_UB, LDN2_UB, MAD_UB, ZUR_UB, DRE_TR, ISP_RB, PAR_SUB, and PRA_SUB. The inc ease in o al LDSA concen a ion in win e may be influenced by he emissions om esiden ial hea ing, a - fic emissions, and me eo ological condi ions a o ing s agna ion (Liu e al., 2021a,Liu e al., 2021b). Fo example, (Ka sanos e al., 2019) epo ed a significan enhancemen o biomass bu ning ae osols a he ATH_UB si e in win e (12/2016–02/2017) and e ealed ha o ganic ma e con ib- u ed app oxima ely hal o he submic on mass in win e , which also influ- enced wi h lowe WS (<3 m/s), enhancing he win e ae osol load. The analysis o seasonal me eo ological ac o s indica es ha lowe win e empe a u es in ci ies such as BUD_UB, ISP_RB, and MAD_UB, which a e p edominan ly a ec ed by empe a u e, can esul in ele a ed su ace ai pollu ion. This is mainly due o inc eased hea ing o homes du ing colde seasons combined wi h las ing Tin e sion laye s occu ing in he whole Ca pa hian Basin in win e . The la e phenomenon es ic s he e ical and ho izon al ai mixing and esul s in poo ai quali y o e ex ended a eas o he basin in la ge and smalle ci ies as well as in u al a eas. Addi- ionally, an ea lie s udy conduc ed in Budapes e ealed ha biomass bu ning is he p ima y sou ce o ai pollu ion in win e , accoun ing o ca. 70 % o he o al ca bon emissions in he PM 2.5 size ac ion (Salma e al., 2020). While in Mad id's win e e enings, empe a u e in e sion aps pol- lu ed ai unde a wa m ai laye , causing smog and poo ai quali y (Siegmann and Gomez-Mo eno, 2022). The highes a e age exposu e o o al LDSA me ics a di e en si es a y seasonally, possibly due o di e ences in PNSD. Du ing summe , si es such as BCN_UB, HEL_UB, LAN_UB, LDN_UB, ROC_UB, LIL_SUB, HEL_TR, and STO_TR ha e he highes a e age o al LDSA, while du ing sp ing, DRE_UB, LEI_UB, LEI_TR, LEI2_TR, and LDN_TR ha e he highes a - e age o al LDSA. P e ious esea ch by (Masiol e al., 2018) ound Accumu- la ion mode PNC peaked a he ROC_UB si e in summe due o inc eased pho ochemical ac i i y leading o mo e seconda y pa icle o ma ion, his phenomenon was also obse ed in Ge many (e.g., D esden, Be lin, e c.) (Junke mann e al., 2016). In addi ion, in Helsinki, Finland (e.g., HEL_UB, HEL_TR), i may be due o long- ange a mosphe ic anspo and e ical downwa d anspo om high al i ude a mosphe ic laye s en iched in nu- clea ion and Ai ken mode pa icles (Lampilah i e al., 2021;Niemi e al., 2009;Pe zold e al., 2008). Finally, in Ba celona (BCN_UB), ac o s such as bu ning and o es fi es, shipping, and a ia ion may con ibu e o high o al LDSA in summe (Pe zold e al., 2008;Ri as e al., 2020). Due o he lack o comple e annual da a o BIR_UB and MAR_UB, no compa ison o seasonal exposu e has been made. 3.3. To al LDSA s o he me ics 3.3.1. To al LDSA s fine modes In ambien ai , pa icles a e usually dis ibu ed ac oss se e al o e lap- ping “modes”, i.e., he nuclea ion (1–25 nm), Ai ken (25–100 nm), accumu- la ion (100–1000 nm) and coa se (1000–10,000 nm) modes, a he han e enly dis ibu ed (Ha ison e al., 2000;Whi by e al., 1972)(Ha ison e al., 2000;Whi by e al., 1972). Pa icles om hese modes ha e di e en o ma ion mechanisms (o sou ces) as well as di e ing ae odynamicbeha - io and a e. The e o e, his s udy analyzes he co ela ions be ween o al LDSA and fine modes, mo e specifically wi h nuclea ion (o e all 10–25 nm, bu a ying om 10 o 25 nm o 20–25 nm, depending o he si e, Table S1), Ai ken (25–100 nm), and accumula ion (100–500 nm) mode PNCs, and o al PNC (10–500 nm). As shown in Fig. 5, all he abo e modes had highly significan co ela- ions wi h o al LDSA (p= 0.01) in all si es. Howe e , he co ela ions a e highe o he accumula ion mode ( = 0.93 ± 0.04, max 0.98, min 0.84), ollowed by he Ai ken mode ( = 0.85 ± 0.05, max 0.94, min 0.77), N 10–500 (PNC_ o ) ( = 0.84 ± 0.09, max 0.95, min 0.58), and lowe o he nuclea ion mode ( = 0.42 ± 0.17, max 0.89, min 0.13), sug- ges ing ha fine modes in he Ai ken and accumula ion ha e he g ea es e ec on he lung deposi ion. Acco ding o Sein eld and Pandis (Sein eld and Pandis, 2008) he mos common size anges o pa icle su ace a ea is in he ange o 100–500 nm, which was compa able wi h ou s udy (Fig. S1). To measu e he LDSA, one needs o measu e he size dis ibu ion o pa icles and hen calcula e he pa icle su ace in each size bin based on i s lung-deposi ion p obabili y. (Todea e al., 2015) conduc ed ho ough measu emen s o se e al LDSA senso s and ins umen s (DiSCmini, Tes o Inc.; nanoT ace , Philips Ae asense Inc.; Pa ec o , Naneos L d.; NSAM 3550, TSI Inc.; Ae o ak, TSI Inc.). Addi ionally, p e ious s udies by (Asbach e al., 2009)and(Hamme e al., 2019) ha e shown ha pa icles smalle han he minimum o he deposi ion cu es a e mo e ele an o heal h e ec s udies, and LDSA concen a ions a e mainly influenced by Fig. 3. 2017–2019 hou ly o al LDSA a e aged concen a ions pe day o he week o he 26 s udied si es. a, u ban backg ound (UB) si es; b, subu ban backg ound (SUB) and egional backg ound (ISP_RB) si es; and c, a fic (TR) si es. X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 6 Fig. 4. 2017–2019 seasonal a ia ions o o al LDSA concen a ions o 24 ou o he 26 s udied si es. a, u ban backg ound (UB) si es; b, subu ban backg ound (SUB) and egional backg ound (ISP_RB) si es; and c, a fic (TR) si es. X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 7 pa icles smalle han 400 nm, which is consis en wi h ou findings. Hence, we selec ed 400 nm as he uppe size limi o ensu e compa abili y o da ase s. P e ious s udies ha e also confi med ha u ban ehicle emissions a e he main sou ce o fine mode in he u ban a mosphe e, wi h pa icles mainly in he o m o soo ha ing p e alen Ai ken and accumula ion mode sizes (30–500 nm) (Shi e al., 2000;Vu e al., 2015). Meanwhile, Fig. S1 indica es ha he con ibu ion o pa icles smalle han 10 nm o LDSA is ela i ely low (weigh : 0.14 % ± 0.2 %). Ou findings show ha he e is a s onge co ela ion be ween LDSA and PNC in he nuclea ion and Ai ken mode a TR si es compa ed o wha is obse ed a UB and SUB s a ions. This co ela ion sugges s ha he e is a s onge associa ion be ween LDSA and a fic- ela ed pa icle sou ces a u ban a fics a ions. Con e sely, in he accumula ion mode, he e is a highe co ela ion be- ween LDSA and PNC a SUB si es compa ed o wha is obse ed a UB and TR si es. This implies ha he e is a s onge co ela ion be ween LDSA and pa icle sou ces associa ed wi h indus ial and o he human ac- i i ies a SUB s a ions. 3.3.2. To al LDSA s BC In 2012, he In e na ional Agency o Resea ch on Cance (IARC) classi- fied BC-con aining diesel soo as g oup 1, ca cinogenic o humans (IARC, 2012). Al hough p e ious s udies ha e analyzed he co ela ion be ween BC and ALV-LDSA (Liu e al., 2022;Reche e al., 2015), he co ela ion be- ween he wo indica o s a di e en moni o ing si es has no been in es i- ga ed in de ail. Howe e , i has also been epo ed ha he ela ionship be ween BC and LDSA is impo an o unde s anding pa icle heal h im- pac s and o sou ce appo ionmen (Lepis ö e al., 2022). In his s udy, he concen a ions o o al LDSA and BC a e co ela ed (p= 0.01) a all 26 moni o ing s a ions (Fig. 5), and pa icula ly a he si es ISP_RB, PRA_SUB, ATH_UB, BUD_UB, GRA_UB, LAN_UB, LDN_UB, and ROC_UB, whe e he co ela ion coe ficien s a e >0.8. This esul sugges s ha a sig- nifican ac ion o BCis loca ed in he UFPs, and hey may ha e simila p e- ailing sou ces and he same pa icle size ange (Chang e al., 2022;Kuma e al., 2010;Lepis ö e al., 2022). While a si e LEI_UB, he co ela ion be ween LDSA and BC was ela i ely low, wi h an 2 o 0.096. This was p i- ma ily due o he ac ha he highes PNC was eco ded du ing non-peak a fic hou s, a ound 11–12 h, and an i-co ela ed wi h BC. Specifically, PNC eached i s maximum a midday, whe eas BC was a i s lowes poin du ing midday and peaked du ing a fic ush hou s. The obse ed pa e n could be a ibu ed o egional o u ban pho o-nuclea ion and umiga ion om highe a mosphe ic laye s, which a e en iched in nuclea ion mode and O 3 and deple ed in BC as he plane a y bounda y laye g ows h ough con ec i e dynamics (T eche a e al., 2023). In u ban a eas o Eu ope, ae osols con aining BC mos ly a ise om bo h aged and esh oad a fic pa icles and include long- ange anspo ed as well as locally gene a ed a ficpa icles(Saa ikoski e al., 2021;Zeka e al., 2006). The e o e, he high co ela ion o be ween o al LDSA and BC implies ha he a fic emission made a high con ibu ion o o al LDSA. 3.3.3. To al LDSA s gaseous pollu an s A co ela ion analysis be ween o al LDSA concen a ions a each si e and he concen a ions o gaseous pollu an s (SO 2 ,CO,NO,NO 2 ,andO 3 ) has been ca ied ou . The esul s show a mode a e and posi i e co ela ion be ween o al LDSA and SO 2 (0.00 ≤ 2 ≤0.52), NO (0.05 ≤ 2 ≤0.56), NO 2 (0.21 ≤ 2 ≤0.64), and CO (0.02 ≤ 2 ≤0.79) (Fig. 5). Since CO, NO, and NO 2 mos ly a ise om he combus ion o ossil uels (all ou pollu - an s), including ha om mo o ehicle emissions (mainly NO, NO 2 and CO) (Liu e al., 2021a,Liu e al., 2021b), which poin s o o al LDSA as a ace o oad a ficemissions.Toconfi m his phenomenon, he co ela- ion be ween LDSA and gaseous pollu an s we e compa ed in di e en ype a ea. We ound he highe co ela ion a a fic si es, suppo ing wi h ou finding. In pa icula , a si es ISP_RB and LIL_SUB, LDSA s CO, he 2 alues we e 0.79 and 0.76, espec i ely, implying ha he main sou ce o LDSA a hese wo si es is om ehicle emissions. To al LDSA concen a- ions a e nega i ely co ela ed wi h O 3 concen a ions a all si es (−0.63 ≤ ≤−0.03). These nega i e alues a e likely he esul o he in- e se ela ionship be ween O 3 and NO X due o he i a ion o O 3 by NO (Cao e al., 2022). Fig. 5. Pea son's co ela ion coe ficien s ( ) o o al LDSA wi h fine modes o he nuclea ion (<25 nm), Ai ken (25–100 nm), and smalle size o he accumula ion (100–500 nm), PNC_ o (N10–500), BC, gaseous pollu an s (SO 2 , NO, NO 2 ,O 3 , and CO), and PM X (PM 1 ,PM 2.5 , and PM 10 ) indi idually o he 26 s udied si es o he 2017–2019 hou ly-a e aged concen a ions (**p= 0.01, and whi e bands ep esen missing da a). X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 8 3.3.4. To al LDSA s PM X The co ela ions be ween o al LDSA and ambien PM X (PM 1 ,PM 2.5 ,and PM 10 ) a e highly a iable among he si es (Fig. 5). Fo example, in DRE_UB, ISP_RB, LIL_UB, LDN2_UB, and MAR_UB, is >0.60, and eaches 0.88 in one case, while in MAD_UB, DRE_UB and GRA_UB, is <0.20. The esul s show clea ly ha le els o PM X do no necessa ily co- a y wi h hose o LDSA and hence i is possible ha he measu emen s o he la e can signi - ican ly imp o e he abili y o ai quali y moni o ing o e alua e he ha m ul e ec s o pa icula e pollu ion (Salo e al., 2021a,Salo e al., 2021b). 4. Conclusions and limi a ions This s udy analyzes he long- e m (2017–2019) spa ial- empo al cha - ac e is ics o o al LDSA a 25 Eu opean and one USA moni o ing si es, in- cluding 15 u ban backg ounds, UB; ou subu ban backg ounds, SUB; one egional backg ound, RB; and six a fic, TR, si es. The concen a ion o o al LDSA a ies ac oss u ban Eu ope. The annual ange o concen a ion o all si es is be ween 20 and 85 μm 2 /cm 3 . The e a e lowe UB concen a- ions in No he n Eu ope and highe concen a ions in Sou he n Eu ope, wi h a end o TR >UB >SUB. The diel, weekly, and seasonal a iabili y in o al LDSA shows significan di e ences (p= 0.01) a di e en ypes o si es, which may be caused by he changes in he sou ces o UFP and la ge pa icles, emission a es, a fic olume, and me eo ological ac o s, including hose a o ing new pa icle o ma ion, s agna ion, long- ange anspo , e ical anspo o ae osols, and influence om plumes om pollu ion ho spo s. Based on he co ela ions o o al LDSA wi h o he pol- lu an me ics, i is sugges ed ha he PNC in he Ai ken and accumula ion modes a e mainly associa ed wi h he o al LDSA concen a ion. The esul s also indica ed ha he main p opo ion o LDSA is a ibu ed o he ALV (50 % on a e age a all si es), ollowed by he TB (34 %) and HA (16 %) ac ions. O e all, he esul s o his s udy p o ide aluable in o ma ion on he o al LDSA concen a ions in di e en ypes o loca ions and high- ligh s he impo ance o conside ing LDSA as an addi ional ai quali y indi- ca o due o i s possible ela ionship wi h heal h isks. The e alua ion o o al LDSA as an exposu e me ic in ime se ies o epidemiological s udies should be u he in es iga ed. The di ec compa ison o he da ase s was a majo limi a ion due o di - e en p o ocols used o PNSD measu emen s and quali y assu ance. Some da ase s (DRE_UB, LAN_UB, LEI_UB, DRE_TR, LEI and LEI2_TR) we e ob- ained ollowing he ACTRIS p o ocols o measu emen s o PNSD and he ins umen s a e equen ly passing quali y assu ance exe cises by he Wo ld Calib a ion Cen e o Ae osol Physics (WCCAP, h ps://www. eu ochamp.o g/calib a ion-cen es/wccap), while o he s (ATH_UB, BCN_UB, GRA_UB, HEL_UB, MAD_UB, HEL_TR, ATH_SUB, LIL_SUB, PRA_SUB, ISP_RB) a e p oduced by ACTRIS collabo a o s o pa ne s and a e ob ained wi h p o ocols ending o ollow he ones om ACTRIS, bu no ecen ly o ne e passed he quali ychecks by WCCAP. Finally, da ase s o BIR_UB, BUD_UB, LND and LND2_UB, MAR_UB, ROC_UB, ZUR_UB, PAR_SUB, LND_TR and STO_TR a e ob ained by highly specialized esea ch eams, bu da a unce ain y emains unknown. The lack o ha moniza ion has caused a high a iabili y o he lowe size de ec ion limi , wi h some PNSD measu emen s s a ing om 3 nm, while o he s s a a 20 nm. Ac- co dingly, cau ion is equi ed when compa ing LDSA alues om hese end membe si es. Howe e , one should akein o accoun ha LDSA con i- bu ions below 10 nm a e ela i ely small. CRediT au ho ship con ibu ion s a emen Xiansheng Liu: Da a cu a ion, Me hodology, So wa e, W i ing –o ig- inal d a . Hadia ullah Hadia ullah: Me hodology, W i ing –o iginal d a . Xun Zhang: So wa e. Ped o T eche a: Da a cu a ion. Ma jan Sa adkoohi: Da a cu a ion. Me i xell Ga cia-Ma lès: Da a cu a ion. C is ina Reche: Da a cu a ion. Noemí Pé ez: Da a cu a ion. Da id C.S. Beddows: Da a cu a ion. Im e Salma: Da a cu a ion. Wanda Thén: Da a cu a ion. Panayio is Kalka ou as: Da a cu a ion. Nikos Mihalopoulos: Da a cu a ion. Ch is oph Hueglin: Da a cu a ion. Da id C. G een: Da a cu a ion. Anja H. T empe : Da a cu a ion. Benjamin Chazeau: Da a cu a ion. G égo y Gille: Da a cu a ion. Nicolas Ma chand: Da a cu a ion. Ja kko V. Niemi: Da a cu a ion. Hanna E. Manninen: Da a cu a ion. Ha i Po in: Da a cu a ion. Nadezda Ziko a: Da a cu a ion. Jakub Ond acek: Da a cu a ion. Michael No man: Da a cu a ion. Holge Ge wig: Da a cu a ion. Susanne Bas ian: Da a cu a ion. Maik Me kel: Da a cu a ion. Kay Weinhold: Da a cu a ion. And ea Casans: Da a cu a ion. Juan And és Casque o-Ve a: Da a cu a ion. F ancisco J. Gómez-Mo eno: Da a cu a ion. Begoña A íñano: Da a cu a ion. Ma ia Gini: Da a cu a ion. E angelia Diapouli: Da a cu a ion. Suzanne C umey olle: Da a cu a ion. Vé onique Ri aul : Da a cu a ion. Jean- Eudes Pe i : Da a cu a ion. Oli ie Fa ez: Da a cu a ion. Jean-Philippe Pu aud: Da a cu a ion. Sebas iao Ma ins Dos San os: Da a cu a ion. Hilkka Timonen: Da a cu a ion. Pasi P. Aal o: Da a cu a ion. Ta eq Hus- sein: Da a cu a ion. Janne Lampilah i: Da a cu a ion. Philip K. Hopke: W i ing – e iew &edi ing. Al ed Wiedensohle : W i ing – e iew & edi ing. Roy M. Ha ison: W i ing – e iew &edi ing. Tuukka Pe äjä: W i ing – e iew &edi ing. Ma co Pandolfi:W i ing – e iew &edi ing. And és Alas uey: W i ing – e iew &edi ing. Xa ie Que ol: Supe ision, W i ing – e iew &edi ing. Da a a ailabili y Da a will be made a ailable on eques . Decla a ion o compe ing in e es The au ho s decla e ha hey ha e no known compe ing financial in e - es s o pe sonal ela ionships ha could ha e appea ed o influence he wo k epo ed in his pape . Acknowledgemen s This s udy is suppo ed by he RI-URBANS p ojec (Resea ch In as uc- u esSe icesRein o cingAi Quali yMoni o ing Capaci ies in Eu opean U - ban &Indus ial A eas, Eu opean Union's Ho izon 2020 esea ch and inno a ion p og am, G een Deal, Eu opean Commission, con ac 101036245). This s udy is also suppo ed by Na ional Na u al Science Foun- da ion o China (42101470, 72242106) and in pa by he Chunhui P ojec Founda ion o he Educa ion Depa men o China unde G an HZKY20220053. This s udy benefi ed om he Ae osol, Clouds and T ace Gases Resea ch In as uc u e (ACTRIS), especially he so-called ACTRIS-2 H2020 esea ch p ojec (g an no 654109), and he au ho s would like o hank ACTRIS (The Ae osol, Clouds and T ace Gases Resea ch In as uc u e), especially he ACTRIS in si u EBAS Da a Cen e (EBAS), o p o iding da ase s o he s udy. This s udy is also pa ly unded by he Na ional Ins i u e o Heal h Resea ch (NIHR) Heal h P o ec ion Resea ch Uni in En i onmen al Exposu es and Heal h, a pa ne ship be ween UK Heal h Secu i y Agency (UKHSA) and Impe ial College London, and he UK Na u al En i onmen Re- sea ch Council, and he iews exp essed a e hose o he au ho (s) and no necessa ily hose o he NIHR, UKHSA o he Depa men o Heal h and Social Ca e. The esea ch was also suppo ed by he Hunga ian Resea ch, De elop- men and Inno a ion O fice (g an no. K132254). We hank also he suppo om “Agencia Es a al de In es igación” om he Spanish Minis y o Science and Inno a ion, and FEDER unds unde he p ojec s CAIAC (PID2019- 108990RB-I00); and he Gene ali a de Ca alunya (AGAUR 2017 SGR41) and he Di ecció Gene al de Te i o i. IMT No d Eu ope and LOA acknowl- edge financial suppo om he Labex CaPPA p ojec , unded by he F ench Na ional Resea ch Agency (ANR-11-LABX-0005-01), and he CLIMIBIO and ECRIN p ojec s, bo h financed by he Regional Council “Hau s-de-F ance” and he Eu opean Regional De elopmen Fund (ERDF). Appendix A. Supplemen a y da a Supplemen a y da a o his a icle can be ound online a h ps://doi. o g/10.1016/j.sci o en .2023.165466. X. Liu e al. Science o he To al En i onmen 898 (2023) 165466 9