Scene ofthe Tt aves 7235 (2020) 138541 Content ists available at ScioneoDirect Science of the Total Environment com/locate/scitatenv ELSEVIER journal homepage: www.else Impact of Covid-19 lockdown on PMjo, SOz and NO2 concentrations in ® Salé City (Morocco) aa Anas Otmani**, Abdelfettah Benchrif”*, Mounia Tahri°, Moussa Bounakhla”, El Mahjoub Chakir*, Mohammed El Bouch‘, M'hamed Krombi“ * ty of Sens Tf Ue ei. No © Rant re or Slr Ee, Se nd Tei (CNS Maro «tara Naoel de aes de Sree de Palo (INSP) Maaco "Deco Ron de Cte, Nitra Cate, ras niGHLIGHTs GRAPHICAL ABSTRACT ~ Ply NOsandS0; concensations were Daily average concentrations of SO, and NO, from March 1 th to April 2nd Teduced by more than half ing he ge ch Covi lack peri : + Covie19 countermeasures contribute : » ' to redce al plane concentrations * but with signin dleences among . * them | « + Langage tansportedaersoscons: og # af Mikaccaneternehaten: § on Sion reductions benefits elated tothe * lockdown, es ‘ ARTICLE INFO ADSTRACT ‘aia ovid-19 was ist repored in Morocco on March 2, 2020 Sa then, tm prevent is propagation the Moroccan fecebed 1 200 government declredastateofealth emergency. Ast frapidandstct couereasurs have ake ching, eco ined em 15 May2029 Tocking downs linn populations mobility and pohbing als al svokable acts. Inthe present ‘acepred 17 May 2000 “study, we attempted to evaluate the changes in levels of some air pollutants (mainly PM, NO, and SO,) in Salé ‘Avaible oie 19 May 2020 city (North-Western Morocco) during the lockdown measures. n ths context, a continuous measurement of PM\q $0, an NO; wascartied before and during the Covi-19 lockdown perind. Asa consequence ofthe securty cue measures and contro actions undertaken the emissions from vehicle exhaust and industrial production were Pile ‘significa reduced, which contribute to che decrease in the concentcations ofthe studied polatans. The ob- No tained results showed that the ffrence between tne concentrations recorded before and during the lockown 30, period were respectively 75% 49% and] 96% for PM; SO; and NOs, PM levels were much ess educed than NO loco ‘The tree-dimensional ar mass backward trajectories, sing the HYSPLT model, demonstrated the benetts of Nersn Po local emission reductions related to the lockdown were overhelimed by the contribution of long-range Itansported aerosols outside areas. tn addition, noteworthy diferences inthe air mass back trajectories and the meteorology between these two periods were evidenced (© 2020 sever BV. Al ight reserved. ~ Corresponding autor at National Cente for Nkar ney. Scene and Technology (CNESTEN) BP 182, 10001 Rt. Mero ‘ma edie enctweestenorgin (A Benet utps:tong 10.1016) scoters 2020139581 ‘aA 997962020 Elie BV. lights ese,2 ‘Aomanictal/ Scene of he al Eminent 735 (2020) 1354 1. Introduction Since the detection of realm's first case ofthe coronavirus (Cov 19), on March 2, Moroccan authorities, for preventing the spread of the pandemic and preserving the health and safety of Moroccan citi- zens, have decided to declare a “state of health emergency” and taken set of rapid and strict measures against Covid-19, All Moroccan cities \were locked down, and the majority of industrial and commercial aciv- ities have been forced t stop their activities until further notice. For cit- izens the authorities request them to stay at home and don't leave it except for very specifi reasons (buy essential supplies, medical ap- pointments...). In add to these safety actions, the authorities decide also to shut down transportation and travel in and out ofthe country, close down schools, universities, restaurants and mosques, and to ban the mass gatherings actoss Morocco (Maneesh and El Alaoui, 2020) However, some activities have been partially maintained as public ser- vice operation, pandemic prevention and contro, and residential living. needs. The Moroccan Government has decided to extend this sanitary state of emergency until 20 May 2020. While the costs ofthese draconian countermeasures are undoubt- edly enormous, several studies reported that reducing anthropogenic activities during Covid-19 outbreak plays a significant role in ar pllu- tion formation and could influence the total ambient air quality (He etal, 2020a:Isafan, 2020). However, Wang eta. (2020) argued that the severe air pollution events would not be avoided by reducing an- thropogenic activities (trafic and industrial activities) during Covid- 19 outbreak, particularly when meteorology is unfavourable. While the cities lockclown could significant improve the air quality its effect ‘may differ considerably across different types of cities. He et al. (2020a) find that more industrialized and colder cities experienced larger reduc- tions in air pollution levels. Furthermore, Croitoru and Sarraf (2017) re- ported that the rigorous options to reduce the ambient air pollution in ‘most Moroccan cities depending onthe emission soutces Fr instance airpollution rom the transport sector can be decreased through the im- provement of the transport system and/or promoting alternative modes of transport. Nevertheless, they highlighted that such options, among, others, need to be carefully considered and adapted to the context of each city PMjo. NO and SO, are very common ambient air pollutants in urban, environments and various studies carried out have demonstrated their severe health-related problems like respiratory and cardiovascular dis- orders, blood pressure, and lung cancer (Koken et al, 2003; Le Tertre et al, 2002). These regulated pollutants arise from anthropogenic sources, including road traffic (eg, motor exhaust; brake, wear and ‘oad erosion; resuspension due to wheel-generated turbulence) and dustrial activities (eg, meta processing industry and waste manage ment) (Thorpe and Harrison, 2008; He et al, 2020b). Besides the above emission sources, lng-range pollution also plays an important role in certain worldwide countries. In Morocco, Benchri tal (2018) "underlined the predominant influence ofthe trans-boundary pollution onto the northern cities ai quality. In tis context, the objective of the present paper isto assess the variations of PMya. NO> and SO; concen- trations in the ambient air of Sal city before and during few days into the implementation ofthe lockdown measures against the propagation of Covid-19 pandemic 2.Materials and methods A continuous PMyo, SO» and NOp measurements were carried out after and during the lockdown in Salé city, NW Morocco, located about 7 km northwest ofthe Kingdom's capital (Rabat), in an urban- residential area. The measurements were cartied aut from 11 March {to 02 April 2020 on Medersa des Merinides (Fig. 1), located about 1.5 km west ofthe historical city centre (Médina). The instruments ‘were set up on a terraced roo, approximately 10 m above ground level (34.04°N, 683°W). Ambient ar PMyo (aerosols ~ 10 ym in diame- ter) were collected on pre-baked (at 300°C during 2h) 150 mm diam- ‘eter QM-A Quartz fiber filters (Whatman®). Daily samples were ‘ollecied with high-volume (DA80, Diitel) sampler during 24 h with, ‘an operating flow rate of 30 m"/h, To determine PMyo mass concentra tions, filters were weighed before and after sampling using a Mettler Microbalance (Mode! MS3045) with 0.1 mg precision. Filter weighing was achieved, in an acclimatized room for 24h at an average tempera- ture of 20 + 1 °C with atmospheric relative humidity maintained below 30%. The uncertainty for each weighing was typically of the order of 20 pg, which represents an average uncertainty of 0.1% for PMio ‘measurements, Low-cost electrochemical sensors, by ENVEA Caitpol. were used, to collect high-resolution temporal data in real-time of SO, and NO. Moderate inter-comparison against reference measurements ‘was performed. Before in-situ measurements, the sensors were co located at a regulatory urban site in urban background conditions, where the performance ofthe calibration algorithms was explored. The uncertainty was typically of the order of 1 ppb for each gas, measurement. Since the variability of pollutants is closely related tothe immediate history ofthe air masses before ativing atthe sampling site (Salvador ‘etal, 2008), three-dimensional air mass backward trajetories were cal- culated using version 47 of the Air Resources Laboratory's Hybrid ‘Single-Particle Lagrangian Integrated Trajectory (HYSPLIT™) model (Draxler and Rolph, 2003; Draxleret al, 2005), Three days backward Uuajectories arriving over sampling point were computed at 12:00 UTC at an elovation of $00 m above ground level. Then, cluster analysis tech nique was carried out to categorize the trajectories into distinct clusters (Benchrifet al, 2018). Furthermore, for investigating the ole othe me- teorology onthe concentrations of studied urban air pollutants, meteo- ‘ological data records were obtained from the weather station Rabat- Sal6 Airport (3405, ~6.75), about 12 km from our sampling site (hutps:/;www.wunderground.com). The recorded metrological data ‘were: wind speed (m/s), relative humidity (2), precipitation (mm) and air temperature (°C). Finally, in an attempt to point out the main changes in PMyo SO and No, concentrations, their dally averages were calculated forthe periods before (March 11th to 20th) and dusing the lockdown (March 21st 10 April 2nd). The objective was to evaluate the relative variation (in %) ‘and the difference in the mean concentration (in jg/nw) between both periods of sampling. 53. Results Comparisons between meteorological parameters prior to and dur ing the lockdown period highlighted the existence of substantial dilfer- ‘ences between the two sampling periods (Table 1). Actually the first period (before lockdown) was characterized by non-significant vatia- tions in temperature, relative humidity and low amount of rainy days (0.4 mm on average). However, the wind speed was steadily increased nti the second period. For the second period (during the lock down), a high meteorology variation was observed: decrease in temperature (difference between averages: -03 °C), more humidity (+1.9%), windy (+19 mys), more amount of precipitation (+0.1 mm), and ‘more rainfall days (percentage of the rainy days in the second period reached 692). ‘The inter-period variability ofthe final mean back trajectories resulting from the clustering process was examined for the periods prior to (March 11th to 20th) and following the lockdown (March 21st to April 2nd). For each period, the relative occurrence, direction‘Aunt Senco the Tet Erne 725 (202) 13854 a ee: Ss Fie 1. Mapsbonig the lation fhe samp ste (2404, 685W). and mean altitude ofthe five identified clusters are presented in Fig 2 The period following the lockdown is dominated by air masses coming from the Norther and Western sectors generally accounted {or 2/3 of the period of concern, while the Eastern sector over the local continental and coastal environments for 1/3 Particularties raised in the first period (before lockdown) accounting for strong winds from the North (across Portugal and Northern Morocco) then occurted over short periods of time. Overall, relevant cluster patter differences were evidenced between both periods. Fig. 3 provides daily variation of PMjo, SO» and NOp conceatra- tions before an after the lockdown. A dramatic decrease inthe con- centrations of NO; (from 5.6 to 02 yg/m") and PMio (from 114 to 28.3 jm’) was eecorded, which corresponds to a variation of = 96% and —75% respectively. For 50, concentrations. hey dropped by3.2 g/m? (from 6.6103: wm) accounted for —49%. The low SO, concentrations recorded in the fl study period (4.75 g/m on aver- 2g2) do not allow evidencing a definite trend. Although all pollutant levels did not exceed the Moroccan air quality standard values (de- cree n'2-09-286 seting standards fr air quality andthe modalities for air monitoring. 2009) during te lockdown period, PMyo concen trations, inthe first period, are higher than the daily reference value of 50 g/m. ‘rable ‘Mean concentrations MSO, NOs and averaged meteorological parameters observe ‘sale iy betwen March 115 20 (before he lawn an arch 213 Ap 2 rng he eho). “Ar palluantincteooegial Sele Dung erence parameter leekdown lockdown [baron in Pie) 1146 233 863 |-75) No ii) Ea 02 54|-26) 80. gin?) ss 33 32-48) Temperature 0) Ms 12 03-2) Preciptation (om) oo ot a1 +38) ‘Wind sped 0s) so 73 19[4241 Fela hom (2) ms ia 19[421 4.Discussion_ From the obtained results, we could note thatthe government deci- sions in response to Covid-19 have impacted the air pollution in Salé city. So, the concentrations of PM, NO, and SO;, in the studied area, were decreased, respectively, by 75%, 96% and 49% within few days after implementation of Covid-19 countermeasures. Several studies (Ocak and Turalioglu, 2008; Dragomir etal, 2015) argued the effect of ‘meteorology on the atmospheric concentrations of traffic-related pol- lutants. While the role ofthe meteorological parameters is quite evident inthis study, they are not quantified The most significant variation was observed for NO, a common, tracer of urban ait pollution/industeial activity. tis primarily resulted from high temperature combustion processes of fossil fuels, especially diesel, thermal power generation, industrial emissions, automobile ex- hhaust and shipping (Burmett et al, 2004; Tobias et al, 2020). Biswas etal. (2019) reported that ambient NO is linked to many health haz- ards and its high concentrations can induce the formation of nitrate aerosols and acid rain. NO» shoved a significant decrease compared to the first period (before lockdown) when social and industrial activities were operating. normally. Its rate reduction was much higher than those observed for PM and SO,. This could be explained by the fact thar the emergency measures established by the Moroccan authorities, related tothe cessation of industrial and transportation activites, had asa consequence a limitation in NO; emission from both industrial pro- {duction and vehicle exhaust, which has implicated a sharp decrease in [NO concentrations during this period, 0, is one of the important indicators of air pollutants that are strongly related to the combustion of coal, petroleurn, and chemical fuel emissions. It isthe major precursor of nucleation formation of ‘new particles in the atmosphere: and when these processes occur in populated regions, they could increase the human exposure to ultra- fine particles (Kulmala etal, 2004), For the present study the variations Jn SO, concentrations were not higher compared to those of PMyo and [NO> because: i) its levels appeared quite variable and significantly hhave decreased since the use, in 20095, of fuel and lubricant oil with low sulphur contents; i) commercial ships and incineration activities4 ‘Aomanictl/ Scene of he al Eminent 735 (2020) 1354 at KOHN GeOW a MON GOW Source Seu Meter AGL Meter AGL Fig 2 Mean ajc avn at Sal frm March 11 o 20 (tse) and rom Nach 21s Apna (gh sid) cise ino 5 wajectry estes (industrial zone and pottery complex around the sampling site) clo- urban particulate emissions from a numberof local sources, However, sures due to pandemic prevention and control actions contribute signif the back trajectory analysis portrayed that air quality in Salé urban icantly inthe reduction of SO, emissions. area is mainly affected by PM originating from the Mediterranean, ‘The pattern of PM, derived principally from road traffic industrial Basin (characterized by marine vessel emissions out of Western exhaust emissions, construction works and dust entrainment, can pre- Europe and Northem Africa, local continental ad from the near Atlan- sumably reflect rapid decrease of traffic density and decline of the tic Ocean (affected by pollutants emitted from the Iberian coast). : 02.2 avr) 7 so2.atn ge . ® z* grt 8 5 0. i. 4 8° ° Nee = ° Mar 12 Mar” 16Mar’ 20Mar’24Mar” 26Mar tAar” AGC aMar 12Mar 16Mar 20 Mar 24Mar”28Mar tApr Apr te no ae tw Mor 12har tee 29Mar Zetuar ZeNler Thor SAR Date Fg 3. aly avecage concentration a PM SD a NO, om March 1h Agi 2 (ite sated on March 2st in Moree) i Sa.‘Aunt Senco the Tet Erne 725 (202) 13854 5 Indeed, the effects of reducing local air pollution on PMyo concentra- tions might stil be limited due to the contribution of long-range ‘wansported particulate matter. ‘While NO, and PMjo concentrations decreased since the strict im- ‘plementation of pandemic prevention and control actions, thei rate re- ‘duction and temporal gradient in the concentrations were significantly cilerent. I would be hypothesized thatthe different amount ofresi- dence time of PMyp and NO, in the atmosphere is another factor that ‘might be implicated to understand ther trends, 5.Conclusion In conclusion, the reduction in PM, $0, and NO> concentrations in the studied area can mainly be attributed tothe drastic measures imit- inghuman movement and industrial activities uring the Covid-19 pan- demic, which resulted in a significant reduction in emissions from vehicle exhaust and industrial production. The authors believe that the PMjo, NO» and $0, concentrations will continue to decrease and keep down to minimal levels during the Covid-19 lockdown period. (RediT authorship contribution statement ‘Anas Otmani: Conceptualization, Writing - original drat, nvestiga- tion, Abdetfettah Benchrif: Conceptualization, Writing - original draft, Supervision. Mounia Tahri: Writing - review & editing, Supervision. ‘Moussa Bounakhhla: Supervision. EI Mahjoub Chakir: Supervision, Mo- ‘hammed El Bouch: Resources. M'hamed Krombi: Resources Declaration of competing interest The authors declare that they have no known competing financial, interests or personal relationships that could have appeared to influ- tence the work reported in this paper. References ea A Gul, Bout, M, Cache, Dap 8 Raha, 2018 Acros innortern Moret: input pathways ad their ema ingernit.Atmos on 174 (November 207) #40147: /dtony10 1016) amesen 2017107, ‘awas MS, chads, SD, Crm Pabars , Mujn AS 2018, Scans ‘bervains of nope ae lonley and oan th locate Pa ‘eros A Qu Re 19,1789 1768 Bret, Sti D. Brook. CakakS. Dales, Rigen M. Vint R, Dann. UGE Asotin berwcen shorten changes israel naa ces Arch Even Heh 9, 239-236 (roar, Stal M, 2017 Estimating he eat cost far polation: the ase of Mare) int O87“ 10 pda ry W233}ep.201 781006, Dram, HEC. Vices M. Constantin, DE, Georges, 201% Preicion of he NO2 ‘Conecnration Oats in an Urban Area Using Mulepl Regression ané Neronal Networks: cso 10.1063/18857255 (December Drader lp, 2005 HYSPUT (vi Sige rte Lagrangian eget Tejectery) Madel NOAA Ar Resources Laboratory, ive Spring MD sce a NORA AR EADY Wetter anos govready pple Drader air Clon Aon 200 MSA Users ld. Fe fan, Tanaka, 50208 COWD-19, Cy Lockdown, and A Plain ‘Sane am Cts Say, 0000 282000010, He. 2h S. Hu. U2. Zeng X, Ca, Yi, Ya, ML WY. 20208 On 0 "sion measirements of reactive moze compa am heavy dt ess trucks In china. "Envton, Plt 260” 114280. hp! Motor 01076) enepelatan 1460, Iss. 200 The dramatic impact formar otek ar qa asi sed ‘meh as chased 60 fa? "6 (3). 275-288, hp oiong 1022038) esmanznos01. ote, Py) Pes, WT Ye. F Elhauser, A Olsen LM Fur. C200. Temperaue, ‘i polation and haspaltion fo eardawecuar diseases ony ele oop in enver nro Heath espe 111 (10, 1312-137 ‘Kuma Mt Veh H, Peta T-Dal Mast Mz La, Reign. V-M. Bie ‘Nery, PH 2004, Forman and growth ates flatness patie reve of oberon fel S35, 19-176. LeTerue A Medina S- SunalE Fsbetg 8. MidlazP, Boumghar, A Vok JM Belin A Akio, Ry Aye, [Suny Sehwarta J asouyan, 202 Shorr eects of parla sir pollution on eardiovsclar eases inh European cies Eigemol Commit Hel 56, 773-79. anes PE Alsat Ax, 2020, He Cou Soath Mite COVD-19: Modelsof ‘Moro and Kerala Ia. Avable at SH ps/Ssmeo stat — 376 s/o 1021393967098 (Ap 3) ‘aa RS Tora, 2008 Ee a teen he atmospescncetatins wate, 3 (5) 325-35 Salar, Ps Arana 8. Quer X, Alastury, A 008 A combined analss ofc Tractors and ares enemy t haar ngage anspor pes of arate mater: the Made a asi, case say. aT Eaveon 390 (23), {35 s08 np ong 1 1OV6 ste 2007 Os “Thorpe, A. Herta RM, 2008 Sores and properties ne exh pit a Ter For fod ah: review. Se. Ttal Evite. 400, 270-282 “obs, Camere. C. Rech. C Masagu. J. Va M Mingiln, MLC Alas A ‘Que X. 200, Changes ae quay ring the on Mrclna Ssi) fe month inthe SARS-CV-2 epidemic Se. Total Environ 725 138540. tps? ‘tuo ta or6gssttenv 2020 38540 Wang. P Chen KZ 8 Wang, P Zhang 200, Severe air potion events not ‘ide by reduced anthrpogenc ates during COMID-19 ouorak Resur Gonser.” ecjek 158 (Meh). 1OASTA. tps dLar/10 1006) evens 2000108814