ebook img

Ozone pollution over the Arabian Gulf PDF

31 Pages·2012·6.81 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Ozone pollution over the Arabian Gulf

Atmos.Chem.Phys.Discuss.,12,6331–6361,2012 Atmospheric Dis www.atmos-chem-phys-discuss.net/12/6331/2012/ Chemistry c ACPD u doi:10.5194/acpd-12-6331-2012 and Physics ss ©Author(s)2012.CCAttribution3.0License. Discussions io 12,6331–6361,2012 n P a p Thisdiscussionpaperis/hasbeenunderreviewforthejournalAtmosphericChemistry e Ozone pollution over r andPhysics(ACP).PleaserefertothecorrespondingfinalpaperinACPifavailable. the Arabian Gulf | D L.Smoydzinetal. is c u s s Ozone pollution over the Arabian Gulf – io n TitlePage P a role of meteorological conditions p Abstract Introduction e r Conclusions References | 1 2 1,2,3 L. Smoydzin , M. Fnais , and J. Lelieveld D Tables Figures is 1Max-Planck-InstituteforChemistry,DepartmentofAtmosphericChemistry,P.O.Box3060, cu 55020Mainz,Germany ss J I io 2KingSaudUniversity,Riyadh,SaudiArabia n 3TheCyprusInstitute,CentreforEnergy,EnvironmentandWaterResearch,Nicosia,Cyprus Pa J I p e Received: 31January2012–Accepted: 18February2012–Published: 29February2012 r Back Close | Correspondenceto: L.Smoydzin([email protected]) FullScreen/Esc D PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion. is c Printer-friendlyVersion u s s io InteractiveDiscussion n P a p e r 6331 | Abstract D is c ACPD u The Middle East and particularly the Arabian Gulf region are characterised by highly s s favourable conditions for O3 formation in summer. We investigated the role of mete- ion 12,6331–6361,2012 orological conditions in O formation using the WRF-chem model. The dispersion of P 3 a p 5 air pollutants strongly depends on local wind patterns, in particular the persistent low- er Ozone pollution over levelnorth-westerlyflowknownasthesummerShamal,andrecurrentland-seabreeze the Arabian Gulf | circulation systems. A general finding from our simulations is that extreme pollution events, with O mix- D L.Smoydzinetal. 3 is −1 c ing ratios exceeding 150nmolmol , can occur regularly over the Arabian Gulf, how- u s ever, their location and magnitude can vary widely. O mixing ratios are highest when s 10 3 io the outflow of the regions with major anthropogenic emissions along the coast is ad- n TitlePage P vectedovertheGulf,wherepollutionplumesarecapturedintheshallowandstablema- a p Abstract Introduction e rine boundary layer allowing little ventilation. The sea-breeze circulation often causes r onshore advection of the pollution in the afternoon, affecting the densely populated Conclusions References | coastal regions along the western shoreline of the Gulf. 15 D Tables Figures When the pollution is transported deeper over land, O3 mixing ratios are generally isc lower due to rapid dilution of precursor gases in the very deep convective boundary u ss J I layer over the desert. io n P J I a p 1 Introduction er Back Close | Enhancedlevelsofozone(O )pollutioncanbeamajorproblembydegradingairqual- FullScreen/Esc 20 3 ity in (sub)urban and industrial areas, especially in summer when its photochemical D is formation by the oxidation of volatile organic compounds (VOC) and catalysed by ni- c Printer-friendlyVersion u trogen oxides (NO ), is most efficient. From satellite observations it is known that NO ss (NO+NO ) concenxtrations in the Arabian Gulf area are exceptionally high (Stavrakoux ion InteractiveDiscussion 2 P 25 et al., 2008; van der A et al., 2008; Lelieveld et al., 2009). Mean concentrations of tro- ap pospheric NO columns exceed 10×1015moleccm−2 along the northern and western e 2 r 6332 | coast of the Arabian Gulf (Beirle et al., 2011). Analysis of data retrieved from the TES D is satelliteinstrumentrevealthatO concentrationsareelevatedinthelowerandmidtro- c ACPD 3 u s posphereovertheentireMiddleEast(Liuetal.,2009). Modelcalculationsindicatethat s io 12,6331–6361,2012 the Arabian Gulf region is strongly affected by photochemical smog due to high back- n P groundlevelsofozonebylong-distancetransport,highlyfavourableweatherconditions a 5 p for ozone formation and strong local pollution emissions (Lelieveld et al., 2009) e Ozone pollution over r Ingeneral,climaticconditionsovertheArabianPeninsulainsummeraredryandhot the Arabian Gulf | ◦ withtemperaturesregularlyexceeding35 C(ElagibandAbdu,2010). Alongthecoast D L.Smoydzinetal. of the Gulf, humidity is often high in the afternoon hours due to the land-sea-breeze is c circulation (Hubert et al., 1983). u 10 s A frequent and persistent weather phenomenon over the Arabian Gulf in winter as sio n TitlePage wellassummerareShamalwinds(Membery,1983;Barth,2001;Raoetal.,2003;Shi P et al., 2004). The Shamal is a low level, thermally driven northerly to northwesterly a p Abstract Introduction −1 e flow, regularly reaching a speed of up to 30ms . During the night, the Shamal often r weakens near the surface but strengthens at elevated levels (300–700m) (Rao et al., Conclusions References 15 | 2003;GiannakopoulouandToumi,2011). Shamalsarethepredominantcausefordust D Tables Figures outbreaks over Iraq and northern Saudi Arabia which impact visibility, air quality and is c the radiation budget throughout the Arabian Gulf region. uss J I During winter, Shamals are known to be initiated after the southeastward passage io n 20 of cold fronts (Perrone, 1979; Walters and Sjoberg, 1988). In summer, the semi- P J I a permanent high pressure system over the Eastern Mediterranean, which extends into p e northern Saudi Arabia, and a semi-permanent low pressure system over Iran lead r Back Close to a north-westerly flow over the Arabian Peninsula and the Gulf (Rao et al., 2003). | FullScreen/Esc Rao et al. (2003) further hypothesised that conditions under which the monsoon low D 25 over India and Pakistan extends to the lee of the Zagros mountains and the south- isc Printer-friendlyVersion u eastern part of the Arabian Peninsula, are favourable for the outbreak of Shamals as s s the east-west pressure gradient is increased. The Zagros mountains to the east of the io InteractiveDiscussion n Gulf, which can exceed 2000m altitude, strongly contribute to the channelling of the P a near surface air flow. Additionally they modulate the diurnal evolution of the Shamal pe r 6333 | through orographic flows (Giannakopoulou and Toumi, 2011). Furthermore, the land- D is sea-breezecanmodulatetheShamalbychangesinthegeostrophicwindcomponent. c ACPD u s Particularly in summer a sea breeze regularly develops over the Arabian Gulf, s io 12,6331–6361,2012 however over the northern coast landward penetration of marine air masses is rare, n P whereastheycanbeadvectedmorethan200kminlandalongthesouth/south-western a 5 p coast (Zhu and Atkinson, 2004; Eager et al., 2008). The vertical extent of the sea e Ozone pollution over r breeze is difficult to determine and in summer may reach an altitude of about 1.5km the Arabian Gulf | (Zhu and Atkinson, 2004; Eager et al., 2008) with a return current above the onshore D L.Smoydzinetal. flow. At night the land-sea breeze over the Gulf is associated with uplift and conver- is c gence whereas during daytime it is associated with subsidence and divergence (Zhu u 10 s and Atkinson, 2004). The sea breeze is initiated rather late during the day (13:00– sio n TitlePage 14:00 LT) compared to other coastal areas (Eager et al., 2008). During strong Shamal P conditions the sea breeze can be fully inhibited (Zhu and Atkinson, 2004). a p Abstract Introduction e The boundary layer over the Gulf is often very shallow and stable (Brooks and r Rogers,2000;AtkinsonandZhu,2005;Brooksetal.,1999)ashotanddryairmasses Conclusions References 15 | from the surrounding deserts are advected over the cooler sea resulting in a strong D Tables Figures temperature inversion and a very shallow stable layer. Since this near-surface layer is c canformwithinanexisting,deeperboundarylayeritrepresentsan“internal”boundary uss J I layer(Garratt,1990;Angevine,2008). IftheShamalisnottoostrong,thusallowingthe io n 20 seabreezetodevelop,theonshoreadvectionofrelativelycoolairfromtheGulftoward P J I a the desert may also be associated with an internal boundary layer, possibly trapping p e air pollution near the surface in the coastal region. Highly stable marine boundary r Back Close layerssimilarlyoccure.g. overtheGulfofMaineaffectingthetransportandconcentra- | FullScreen/Esc tions of trace gases as observed during the International Consortium for Atmospheric D 25 Research in Transport and Transformation (ICARTT)/New England Air Quality Study isc Printer-friendlyVersion u (Angevine et al., 2006; Fehsenfeld et al., 2006). s s The objective of this study is to assess the influence of meteorological conditions io InteractiveDiscussion n (windregime,hightemperaturesinsummer,humidityconditions)onO formationover P 3 a the Arabian Gulf region. Further, we analyse the rate of photochemical O3 formation pe r 6334 | in view of uncertainties in the emission inventories of anthropogenic O precursor D 3 is sources. c ACPD u s s io 12,6331–6361,2012 n 2 Case study description P a p e Ozone pollution over For our case studies, we have selected a period of five days in mid July (14–18) in the r the Arabian Gulf years2009,2010and2011. Ingeneralthepotentialforabuildupofhighconcentrations | 5 ofO3ishighestinsummer(Lelieveldetal.,2009). Duringapproximatelyhalfofthetotal D L.Smoydzinetal. of 15 days of simulation the wind direction in the boundary layer is northerly to north- is c u westerly which is the predominant flow in July based on climatological data. However, s s the initial meteorological conditions on 14 July in all three years are rather distinct io n TitlePage (Figs.1and2)thusprovidingtheopportunitytoinvestigateO formationunderdifferent P 10 3 a ambient conditions, particularly focusing on (i) days and regions with and without the p Abstract Introduction e r development of land-sea breeze systems and (ii) periods with and without Shamal Conclusions References winds. | D Tables Figures 2.1 Synoptic conditions isc u ss J I 15 2.1.1 July 2009 ion P J I a Following the work of Rao et al. (2003) we have selected one simulation period p e favourable for a persistent Shamal over the Gulf. An upper level trough was located r Back Close over Turkey accompanied by a ridge to the west across the Eastern Mediterranean | FullScreen/Esc and a ridge to the east over western Iran. The cyclone remained very stable over D Turkey during the entire simulation period and propagated only slowly eastward. The is 20 c Printer-friendlyVersion u surfacepressuredistributionshowsasignificantnorth-southgradientovertheArabian s s Gulf area as the Mediterranean high was strong with a ridge extending over central io InteractiveDiscussion n Saudi Arabia (Fig. 1). Such pressure conditions typically enhance the north-westerly P a flow near the surface (Rao et al., 2003). Compared to 2010 and 2011, wind speeds p e r 6335 | −1 D overtheGulfwerehigher(maximumupto20ms ). Furthermore,thetemperatureon is ◦ c ACPD the simulation days in 2009 was lower (regionally more than ≈5 C) than in the years u s 2010 and 2011 (Fig. 2). The temperature differences between air masses over land sio 12,6331–6361,2012 n and water were smaller leading to a higher boundary layer over water (compared to P the years 2010 and 2011). Furthermore, meteorological conditions lead to a less effi- a 5 p e Ozone pollution over cientorinhibitedseabreeze. AtypicalphenomenonovertheGulfisthedevelopmentof r the Arabian Gulf a shallow inversion only a few hundred metres above the sea surface (≈950hPa) cap- | ping a stable and moist marine boundary layer. At approximately 500hPa a very dry D L.Smoydzinetal. subtropical subsidence inversion associated with the downward motion of the Hadley is c 10 circulationistypicalformostdaysinsummer(Fig.2)(AckermannandCox,1982;Reid us s et al., 2008). io n TitlePage P 2.1.2 July 2010 ap Abstract Introduction e r On 14 July 2010 the 500hPa flow over the eastern Mediterranean, Iraq and Iran Conclusions References | was almost meridional. On 15 July a weak upper level trough formed over south- D Tables Figures 15 ern Turkey which slightly intensified in the following days while it slowly propagated is c north-eastward. The surface pressure chart is significantly different compared to the u ss J I simulationperiodin2009. AweakeningoftheIranianlow,withasimultaneousnegative io n pressure tendency over central Saudi Arabia and the northern Gulf was unfavourable P J I a for the development of a Shamal (Fig. 1). p e r Back Close 2.1.3 July 2011 20 | FullScreen/Esc On 14 July a cyclone was located over the Black Sea and propagated slowly south- D eastward during the following days. Generally the surface pressure distribution was isc Printer-friendlyVersion u similar as during the Shamal period in 2009 with a distinctive surface low over the s s southernGulfandasouth-eastwardpenetratingMediterraneanhigh. However,aposi- io InteractiveDiscussion n tivepressuretendencyovercentralIran,QatarandtheUAEweakenedorinhibitedthe P 25 a p Shamal. e r 6336 | 3 Model setup D is c ACPD u The fully coupled chemistry version of the Weather Research and Forecasting model s s (WRF-chem) (Grell et al., 2005) version 3.3.1 was used for this study. The treatment io 12,6331–6361,2012 n of aerosols and aerosol-cloud interactions in WRF-chem is described in detail by Fast P a et al. (2006) and Chapman et al. (2009) and only a brief overview of the features most p 5 e Ozone pollution over r relevant in this study is given below. the Arabian Gulf Gas phase chemistry is calculated using the CBMZ mechanism (Zaveri and Peters, | 1999). Direct and indirect aerosol effects are calculated in the model as described D L.Smoydzinetal. is by Chapman et al. (2009). The representation of aerosol activation is based on the c u parametrisation by Abdul-Razzak et al. (1998) and its implementation into WRF-chem s 10 s io follows the description by Chapman et al. (2009) and references therein. Cloud micro- n TitlePage physical processes are calculated using the scheme by Lin et al. (1983) with modifi- P a p Abstract Introduction cations in order to treat cloud drop numbers prognostically. For nests with horizontal e r resolutionsrequiringtheusageofaconvectionparametrisationweusetheschemeby Conclusions References | Grell and Devenyi (2002). For calculating aerosol thermodynamics and microphysics 15 the Module for Simulating Aerosol Interactions and Chemistry (MOSAIC) is used (Za- D Tables Figures is veri et al., 2008). Boundary layer physics is parametrised using the Yonsei University c u (YSU) PBL scheme (Hong et al., 2006). ss J I io Initialandboundaryconditionsforgasphasespeciesforoursimulationsarederived n P J I from MATCH-MPIC (Model of Atmospheric Transport and Chemistry, von Kuhlmann a 20 p etal.,2003)whichisranoperationallyasaglobalchemicalweatherforecastingmodel er Back Close at the Max Planck Institute for Chemistry. Initial and boundary conditions for aerosol | FullScreen/Esc constituents are taken from simulations performed with the global chemistry-climate D model EMAC (Pringle et al., 2010). For anthropogenic emissions we use the high is resolution(0.1×0.1◦)EDGAR-CIRCEemissioninventory(Doeringetal.,2009a)which cu Printer-friendlyVersion 25 s hasbeenpreparedintheframeworkoftheCIRCEProject(No. 036961)bytheEDGAR sio InteractiveDiscussion n group(EmissionDatabaseforGlobalAtmosphericResearch)oftheEC-JointResearch P Center Ispra (Italy), Climate Change Unit and has been evaluated by Doering et al. ap e (2009b). r 6337 | The setup of our simulations comprises three nests with horizontal resolutions of D is 48km, 16km and 4km and with a vertical resolution of 53 layers with narrowly spaced c ACPD u s model levels in the lower troposphere. If not mentioned otherwise we refer in our s io 12,6331–6361,2012 analysis to the 16km grid. n P a p 5 4 Results er Ozone pollution over the Arabian Gulf 4.1 O distribution and variability | 3 D L.Smoydzinetal. ForouranalysiswefocusonthetimeofdaywhenO mixingratiosreachamaximum, is 3 c i.e. at 12:00 UTC = 15:00 Local Time (Fig. 3). A general finding from our simulations us s encompassingatotaltimeperiodof15daysisthatO hotspotsdevelopregularlyover io 3 n TitlePage the Arabian Gulf, however the location and the intensity of the O pollution can vary P 10 3 a widely. p Abstract Introduction e r The boundary layer over the Gulf is shallow under all conditions, being relatively Conclusions References deepest for the simulation period in 2009 (≈200–400m compared to ≈50–100m in | 2010 and 2011 around 12:00 UTC). It thus appears that the formation of O hot spots D Tables Figures 3 is over the Gulf is facilitated by the inhibition of vertical transport in regions with strong c 15 u pollutionemissions,sothatbothprecursorgasesandtheresultingO3 accumulatenear ssio J I the surface. Especially in the year 2009 the boundary layer is relatively low over the n P J I north-western part of the Gulf, which is in the region with highest VOC emissions. a p On the first day (14 July, Fig. 3, first row) of the simulation period, high O values e 3 r Back Close are simulated during all three years over the northern Gulf region with a distinctive hot 20 | spot along the coast south of Kuwait. In 2010 and 2011 O mixing ratios also exceed FullScreen/Esc 3 −1 D 80nmolmol further inland in the late afternoon/early evening hours (especially in is the area of Dammam/Al Jubayl) when the sea breeze transports polluted air over the cu Printer-friendlyVersion s coastal regions. Particularly on 14, 15 and 16 July 2010 the sea breeze is strong and s io InteractiveDiscussion O mixingratiosareelevated(>60nmolmol−1)upto50–100kmwestofthecoastline. n 25 3 P On15July2010,theweakbackgroundflowfacilitatestheearlierandstrongerinitiation a p e oftheseabreeze(about08:00UTC)comparedto14July2010,leadingtothetransport r 6338 | of marine air masses over the coast during almost the entire day. At 12:00 UTC (15 D is July 2010) O mixing ratios over the Gulf are very low, however, almost everywhere c ACPD 3 u along the western coast they exceed 60nmolmol−1 (over land) with local maxima of ss io 12,6331–6361,2012 more than 80nmolmol−1 (around 28◦N). Especially in the area of Al Jubayl O mixing n 3 P ratios are higher on 15 July than on 14 July because (i) the sea breeze developed a 5 p earlier,(ii)theboundarylayeralongthecoast(upto≈40kminland)isloweron15July er Ozone pollution over than on 14 July, thus trapping the locally formed pollution near the surface. In contrast the Arabian Gulf | to 2010 and 2011, on 14 July 2009 the pollution plume remains located over the sea D L.Smoydzinetal. during the whole day because the strong Shamal winds prevent the development of a is c sea breeze. u 10 s s Ontheseconddayofthesimulation(15July2009,Fig.3,secondrow)highlevelsof io n TitlePage O buildupagainoverthewesternGulfwherethepersistentnorth-westerliestransport 3 P a the pollution emitted in Kuwait over the Gulf which contributes to the photochemical p Abstract Introduction e transformation of local emissions further South, leading to the formation of O mixing r 3 −1 Conclusions References ratiosofmorethan100nmolmol . Asaconsequenceoftherareeventofaseabreeze | 15 developing over the northern part of the Gulf, on 15 July 2010 the highest local mixing D Tables Figures ratiosofO (≈120–150nmolmol−1)aresimulatedalongthecoastofKuwait. Intheyear is 3 c u 2011,O3 mixingratiosarealsoelevatedoverthenorthernGulf(≈80nmolmol−1)butin ss J I −1 io contrast to most other days extreme values of 100nmolmol are not exceeded. The n P J I 20 generalexplanationforthedistincthorizontaldistributionofsurfaceO3on15July2009, a p 2010and2011isthechangingwinddirection. In2009thewinddirectionispersistently e r Back Close north-west,whereasitissoutherlyorsouth-easterlyin2010duringthenightandinthe | morning. In 2011 the nighttime wind direction is also north-westerly, from Kuwait over FullScreen/Esc D the coastal area into northern Saudi Arabia. During the time of photochemical activity is 25 the outflow direction from Kuwait is southward over land. Mixing ratios of O3 near the cus Printer-friendlyVersion surfaceinthatlocationarenotashighasovertheseamainlyduetothedeepboundary s io InteractiveDiscussion layer(≈4.5kmintheoutflowregionsouthofKuwait)leadingtotheefficientmixingand n P dilution of the O plume. The same argument holds for the lack of O hot spots over a 3 3 p the Gulf on 16 and 17 July in 2009 and 2010. er 6339 | ThehorizontaldistributionofO ,especiallyintheoutflowregionofKuwaitappearsto D 3 is be generally very similar on 16 July, (Fig. 3 third row) in 2009 and 2011 though mixing c ACPD u −1 −1 s ratiosare significantlyhigher inthelatter year(≈90nmolmol versus ≈60nmolmol s io 12,6331–6361,2012 in2009)whiletheboundarylayerheightisverysimilarinbothyears. However,in2011 n P astableresiduallayerdevelopsinthenightbetween15and16Julyleadingtoelevated a 5 p O mixing ratios exceeding 60nmolmol−1 at 2000m altitude throughout the night over e Ozone pollution over 3 r north-eastern Saudi Arabia, facilitating long-distance transport from the north. When the Arabian Gulf | the boundary layer develops in the morning of 16 July the elevated O mixing ratios 3 D L.Smoydzinetal. from the lower free troposphere significantly contribute to the high near surface values is c (Fig. 4, right column). u 10 s Asindicatedabove,theoutflowofKuwaitcancontributestronglytotheO3 formation sio n TitlePage over the Arabian Gulf. On 15 July in 2011 pollution from Kuwait is transported mainly P overSaudiArabiaornorthwardoverIraq. DespitethelackingtransportofO precursor a 3 p Abstract Introduction e gases from Kuwait (or more generally polluted coastal areas) over the sea, O mixing r 3 −1 Conclusions References 15 ratios over the northern Gulf exceed 80nmolmol at 12:00 UTC. In contrast, on 16 | −1 July in 2009 O mixing ratios over the Gulf are relatively low (≈40nmolmol ) even 3 D Tables Figures though the wind regime is similar as on 15 July in 2011. From Fig. 5a we can identify is c u the regions with high VOC emissions over the northern Gulf region. The O3 formation ss J I rate between 10:00 and 12:00 UTC on 16 July 2009 in the circled region in Fig. 5a, io n b is approximately 0.5–5 pmol(mol−1s) whereas it reaches values twice as high on P J I 20 a 15 July 2011 (Fig. 5c), even though the emission strength of VOC’s in the area does p e not differ between the two years. Nevertheless, in 2011 surface VOC mixing ratios are r Back Close almost a factor of two higher in this region (200–400nmolmol−1) compared to 2009 | FullScreen/Esc (50–100nmolmol−1). Thedifferenceisagainaconsequenceofthesubstantiallylower D is 25 boundary layer over the Gulf in the nights between 14 and 15 July 2011 and 15 and cu Printer-friendlyVersion 16 July 2009 (≈100m in2011 compared to 500min 2009). In addition to less efficient s s verticalmixingon15July2011,thewindspeedislowerthanon16July2009,andthe ion InteractiveDiscussion reduced venting allows the buildup of high levels of O precursor gases. P 3 a p e r 6340 |

Description:
12, 6331–6361, 2012. Ozone pollution over the Arabian Gulf. L. Smoydzin et al. Title Page. Abstract. Introduction. Conclusions. References. Tables. Figures. ◁. ▷. ◁. ▷. Back. Close . At night the land-sea breeze over the Gulf is associated with uplift and conver- gence whereas during dayt
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.