OCTOBER 1977HUSAR ET AL1089Three-Dimensional Distribution of Air Pollutants in the Los Angeles Basin RUDOLF B. HUSAR AND DAVID E. PATTERSON Washington University, St. Louis, Ma. 63130DONALD L. BLUMENTHAL, WARREN H. WHITE' AND THEODORE B. SMITH Meteorology Research, Inc., Altadena, Calif. 91001 (Manuscript received 28 May 1976, in revised form 18 August 1977)Air Pollution Research Laboratory, Department of Mechanical Engineering,ABSTRACTData from a three-dimensional pollutant mapping program, conducted in the Los Angeles basin, wereanalyzed to obtain "grand average" vertical profiles sampled on 24 summer days in 1973. Morning andafternoon profiles at four locations show an erosion of the nighttime radiation inversion, increased tempera-tures, more intense mixing in the inland areas, and a semi-permanent subsidence inversion at higher levels.High values of primary pollutant parameters (NO, and condensation nuclei) are seen in the western part ofthe basin at Hawthorne. Secondary pollutant parameters (Os and light scattering coefficient) were domina-ting at the inland receptor site, Riverside. Ozone concentrations in the morning were consistently higheraloft. The deficit near the surface is attributed to ozone scavenging by primary emissions.1. IntroductionThe experience of observing the smog layer froma descending airplane is probably the most convincingdemonstration of the three-dimensional nature of theLos Angeles smog. In the summer months of 1972and 1973, the Three Dimensional Pollutant GradientStudy was conducted to map the vertical and hori-zontal pollutant structure, and to study the transportand transformation processes in the Los Angeles basin.Two aircraft made soundings, in the form of verticalspirals, from the top of the polluted layer to thesurface at 17 locations with an average of about40 spirals for each location. Three meteorologicalparameters, temperature, relative humidity and tur-bulence (energy dissipation rate), as well as fivepollutants, 03, NO,, CO, light scattering coefficientand condensational nuclei, were monitored continu-ously along with the necessary coding and positionparameters. A list of the instruments used and theircharacteristics is given in Table 1.The three-dimensional pollutant mapping programimproved substantially our understanding of the pol-lutant transport and transformation processes in theLos Angeles basin. The data indicate complex andvarying patterns of these processes from one time orlocation to another. Detailed inspection of the datarevealed, however, that certain phenomena are soPresent affiliation: 1180 North Chester, Pasadena, Calif.91104.characteristic for a given location or time period thatthey survive extensive averaging.In this paper, we present and discuss the "grandaverage" vertical profiles measured over the airportsof Hawthorne, El Monte, Ontario and Riverside.Their respective locations in the basin are shown inFig. 1. All 165 soundings, conducted on 24 days in1973 at these locations, were used to calculate themean and standard deviation of all measured parame-ters for each location, time period (A.M. or P.M. PDT)and hundred-foot altitude increment.The average profiles presented in this section arecalculated from a sample population which was biasedtoward photochemical smog conditions as mappingwas intentionally restricted to days with above averagesmog potential. Although they cannot be taken asrepresentative on an overall basis, the mean profilesshould be fairly typical of conditions during smoggysummer and early fall days since they are drawnfrom data on 24 days of this type. It is hoped thatthe findings presented here will have utility in theassessment of average transport and transformationprocesses and in providing information for the testingof three-dimensional diffusion-transformation models.2. Meteorological parametersa. TemperatureA primary factor in the production of high pol-lutant concentrations in Los Angeles air is the fre-1090JOURNAL OF APPLIED METEOROLOGYVOLUME 16TABLE 1. Aircraft instrumentation.Time responseInstrument Ranges available (90%, for usual ranges)1. MRI Integrating Nephelometer2. Environment One Condensation3. REM 612 Ozone Monitor (chemiluminescent)4. REM 642 NO-NO, Monitor (chemiluminescent)5. Andros 7000 CO Monitor6. MRI Airborne Instrument Package(dual isotope fluorescence)TemperatureHumidityTurbulenceAltitudeIndicated air speedConverter (uses aircraft radio)7. Metrodata M/8 VOR Analog8. Metrodata 620 Data Logger (20 channels)lo*, 40, 1OOX 10-4 m-11, 3, 10, 30, loo*, 300*lOKXlOl CN cm-3SO*, 200 pphmOS*, 2*, 10 ppm20, 50*, 100, 200 ppm-5 to +45"C0 to 100%0 to 10 cmf sec-10 to 10 000 ft50 to 150 mph1s5s5s5s5s5s30 s3 s (to 60%)1s6 1s1s48 channels per second* Range normally used.quent occurrence of temperature inversions whichrestrict the vertical dispersion of pollutants. Thisphenomenon is evident in Fig. 2, which shows themean morning and afternoon temperature profiles atHawthorne, El Monte, Ontario and Riverside. In themean morning profiles at all four locations, the tem-perature of the air at 3000 ft MSL is higher thanthe temperature of the air near the surface, witha negative lapse rate from 3000 ft down almost tothe surface. As the day progresses, the surfaces ofinland areas of the basin are heated by the sun,warming the air near the ground and eroding theinversion from below (Edinger, 1973). By afternoon,the mean profiles for all three inland locations showa well-defined layer at the surface in which the lapserate is adiabatic.The effect of surface heating is seen from anotherperspective in Fig. 3a, which displays the meanafternoon temperature profiles for the four samplinglocations. At Hawthorne the afternoon profile is littlechanged from the morning profile due to the stabilizinginfluence of the nearby ocean. As air moves inlandwith the usual afternoon flow, however, temperaturesnear the surface increase, eroding the inversion. TheFIG. 1. Map of the Los Angeles air basin indicating the locations where the "grand average" profileswere calculated. Streamlines show most frequent afternoon surface winds during July (after Blumenthalet al., 1974).OCTOBER 1977 HUSARcU5 5-4-3-2-I-ET AL.- ONTARIOIII1091ohk-3hk+0EL MONTEIO 15u20 25_J30'10 15 20 25 30TEMPERATURE PC )FIG. 2. Mean morning and afternoon temperature profiles.unstable layer at the surface deepens until, at Ontarioand Riverside, it may "break through" the inversion.b. TurbulenceThe mean profiles of small-scale turbulence inten-sity (Fig. 4), as indicated by the energy dissipationcoefficient E* (MacCready, 1964), are compatible withthe mean thermal structure. In the morning, whenair is stable, turbulence is confined to a shallow layerat the surface where it is generated by mechanical effectsand surface heating. In the afternoon, turbulence ex-tends about 1000 ft above the morning mixing layer.In all profiles, the intensity of the turbulence decreaseswith height until it reaches a "background" value ofabout =OS cmg s-l, reflecting the decay of small-scale turbulence with increasing height.c. Relative humidityMean afternoon profiles of relative humidity areshown in Fig. 3b. Mean morning and afternoon rela-tive humidities are generally in the range 30-50%,except in the mixing layers at Hawthorne andRiverside.3. Reactive gasesa. Nitrogen oxidesThe first two oxides of nitrogen are importantparticipants in the photochemistry of the Los Angelesatmosphere. Photodissociation of NO2 (N02+hv+N0+0) starts the chain of reactions leading to thebuildup of ozone. Oxidation of NO(NO+03+N02+02) limits the rate of this buildup in its early stages.5-4- LAPSE- RATE2I-w TEMPERATURE ('C)00 0.1 0.2 03NO, (ppm)2OO 1 I_I_20-I70RELATIVE HUMIDITY (%I(dl HAWTHORNE EL MONTE ONTARIO..... . . . . . . . . . RlVERSlDE-4:. .\iCONDENSATION NUCLEI (cm-')FIG. 3. Mean afternoon profiles of temperature, humidity, NO, and condensation nuclei at the four sampling sites.Although these two reactions affect the relative con-centrations of NO and NO2 in the air, they leaveinvariant their sum, the concentration of NO,. NO,is formed almost exclusively by combustion sourcessuch as motor vehicles and power plants and is lost0246802460-E it\, ONTARIO ,~~~lD~3\ I\\ '.\ '\.2 '\I02 4 6 802 4 6 8TURBULENCE(~m"~sec-9FIG. 4. Mean morning and afternoon turbulence profiles.1092 JOURNAL OF APPLIED METEOROLOGY VOLUME 166 5L HAWTHORNE `i EL MONTE5In reacts very rapidly with NO, the principal con-stituent of NO, emissions, so that fresh emissions ofNO, tend to lower, rather than raise, 03 concentra-tions. These characteristics are clearly demonstratedin the mean profiles of 03 (Fig. 6). Note that ozone concentrations do not drop to .,. background levels (-0.04 ppm) above the mixing layer. At Hawthorne, for example, the mean tem-0 01 02 03 04 0 01 02 03 04 perature, turbulence and NO, profiles all indicate a mixing layer strongly confined below 2000 ft MSL, yet the mean 03 concentration at 3000 ft MSL is 0.1 ppm, greater than the federal standard. Mecha- nisms by which polluted air occurs above the mixing layer are discussed elsewhere (Blumenthal et al., 1974). What is important in the present context is that the O3 at 3000 ft over Hawthorne is part of a polluted air mass which has aged above the mixing layer.0 01 0.2 03 04 0 01 02 03 04 In the absence of scavenging by fresh NO, emissions, which are confined within the mixing layer, reactions have proceeded almost to completion, producing rela- tively large ozone concentrations from small concen-trations of NO, and hydrocarbons.- a E :k,; 2 :, ;::, ,f) I I ,sp c -.._ :-RIVERSIDE:[,I 3 , ,;I,,, I , ,`\-.\,INO,, (ppm)FIG. 5. Mean morning and afternoon NO, profiles.through reactions with surfaces and by the formationof nitrates.Mean profiles of NO, concentration are shown inFigs. 5 and 3c. Substantial concentration gradientsappear near the surface in the morning profiles due tothe generally poor mixing prevailing at this time.By afternoon, concentrations become fairly uniformthrough the unstable mixing layer. In both morningand afternoon profiles, mean concentrations drop toabout 0.03 ppm above the mixing layer.The smallest values for mean NO, concentrationsand the integral of these concentrations with respectto height are seen in the afternoon profile for Riverside.The latter half of a typical afternoon trajectory toRiverside passes over a predominantly rural area,where emissions of NO, are apparently not sufficientto balance losses by aerosol formation and dry depo-sition. Peroxyacetyl nitrate (PAN) concentrations atRiverside can reach 0.05 pprn on smoggy days(Lundgren, 1970). Particulate nitrate concentrations>lo0 pg m-3, equivalent to over 0.04 ppm of NO,,were measured there during the 1973 Aerosol Charac-terization Study (Hidy et al., 1974). The measuredhigh concentrations of PAN and particulate nitratesindicate that a substantial fraction of the NO, emittedin the western portion of the air basin is convertedto nitrates. It is also likely that a further, at present,unknown fraction of NO, is lost to vegetation andground (Hill, 1971) enroute to Riverside.b. OzoneUnlike NO,, ozone is not emitted directly, but isformed in the atmosphere through the sequence ofreactions initiated by the photodissociation of NOz.A second feature bf the mean profiles which isunique to 03 is the occurrence of deficits near thesurface. The most striking example of this is themorning profile at Hawthorne, where mean 03 con-centrations near the surface are under 0.02 ppm, lessthan the 0.04 ppm background levels observed inclear air. In fact, all four morning profiles as well asthe afternoon profile at Hawthorne, show ozone con-centrations which are lower within the mixing layerthan they are immediately above it. These 03 deficitsare due principally to scavenging by fresh emissionsof NO, which are trapped within the shallow, well-defined mixing layers exhibited in these profiles.80W3aL3 ,I:: ,;I2I I'/'0 005 OK) 015 020 03t2l IONTARIO?? ,-'4:j 32l I\--.'005 010 0.15 0.20IRIVERSIDE-;j,`hLII1 I0 0.05 010 015 Ox) 0 0.05 010 015 Ox)OZONE (ppm)FIG. 6. Mean morning and afternoon 03 profiles.OCTOBER 1977 HUSARET AL10934. Aerosol parametersa. Condensation nuclei countSince small particles are much more numerous thanlarge particles, the condensation nuclei count (CN)is primarily an index of the small-particle fraction ofthe aerosol. Particle size distribution measurements(Whitby et al., 1972) have shown that particles under0.1 pm in diameter account for nearly all of the CN.Particles in this size range are produced by com-bustion sources and (under conditions thought to berare in ambient Los Angeles air) by homogeneousnucleation (Husar et al., 1972; Whitby et al., 1972).These particles coagulate rapidly with larger particlesand with each other, so that the half-life of CN undertypical Los Angeles smog conditions is on the orderof an hour (Husar and Whitby, 1973; Husar et aZ.,1972; Davies, 1974).Mean profiles of CN are shown in Figs. 7 and 3d.Substantial CN gradients are observed near the sur-face in all four morning profiles and in the afternoonprofile at Hawthorne. Above the mixing layer, CNcounts drop to about 5X103 ~m-~, less than one-tenthof the counts within the mixing layer. Counts withinthe mixing layer are lowest at Riverside.b. Light scattering coeficientThe light scattering coefficient (bscat) of the atmo-sphere is determined largely by the concentration ofparticles with diameters in the range 0.1-1.0 pm(Ensor et al., 1972), since these are the most efficientscatterers of visible radiation. Electron micrographs(Husar et al., 1976) and chemical analyses (White,1976) indicate that, under photochemical smog condi-651 HAWTHORNEPM.3/ CNTARIO ~ 1 RIVERSIDE4uu0 50 100 150 0 50 100 150CONDENSATION NUCLEI (1000/cm3~FIG. 7. Mean morning and afternoon condensation nucleiu L--0 2 4 6 802 4 6 8b,,,, ( m-' X IO-' 1FIG. 8. Mean morning and afternoon bacst profiles.tions, much of the aerosol in this size range is com-posed of secondary sulfates, nitrates and organics(Hidy et al., 1974; Gartrell and Friedlander, 1975)produced from the gas phase. Some of this materialis hygroscopic and bscat is affected by high ambientrelative humidities (Covert et al., 1972). Once formed,the light scattering fraction of the aerosol is a fairlystable component of the atmosphere, with low surfaceloss rates (Chamberlain, 1967) and low coagulationrates (Husar et al., 1972; Davies, 1974).Of the four air quality parameters studied in thiswork, bscat is the most stable and least sensitive tofresh emissions. These qualities are shown in the meanafternoon profiles of bScat (Fig. 8). In all four profilesa RIVERSIDE0 HAWTHORNE1NO, CN 03 bscatPRIMARY SECONDARYFIG. 9. Mean afternoon contaminant characteristics in firstprofiles.200 ft above ground at Hawthorne and Riverside.1094 JOURNAL OF APPLIED METEOROLOGY VOLUME 16bscat is fairly uniform through the mixing layer anddrops to low levels above the mixing layer.5. Pollutant agingFig. 1 shows that the characteristic air massessampled over Hawthorne and Riverside in the afternoonhave different histories. Hawthorne is only about30 min of air parcel travel time downwind from theocean. During those 30 min over land en route toHawthorne, air passes over the San Diego Freewayand State Highway 1 and 107, as well as the heavilyindustrialized region around El Segundo. The afternoonsoundings at Hawthorne thus sample predominantlyfresh emissions, superimposed on the marine back-ground (which itself may contain well aged anthro-pogenic contaminants).Riverside, on the other hand, is from 3-6 h down-wind of the ocean. As marine air moves inland, itfirst passes over the urban and industrialized areasof Los Angeles and Orange counties, then over therelatively rural region from the Chino Hills and SantaTEMPER AT U R E ("C )000-Y61W03 5-l-ka 4-3-2-I-Ana mountains eastward to Riverside. As a result,most of the contaminants sampled in the afternoonsoundings at Riverside have had 3-6 h to age enroute.The afternoon contaminant characteristics of thetwo locations are summarized and compared in Fig. 9which shows the afternoon means for NO,, 03, CNand bscat in the first 200 feet above ground level.It is apparent from Fig. 9 that the afternoon con-taminant mixtures at Hawthorne and Riverside reflectthe different histories of the air sampled, with sig-nificantly more NO, and CN at Hawthorne and sig-nificantly more 03 and bacat at Riverside. Direct emis-sions account for most NO, and CN, at whichconcentration can only decay in the atmosphere. Highvalues of these parameters are thus generally indicesof fresh emissions. Ozone, and a large fraction of thelight scattering aerosol, are not emitted directly, butare produced in the atmosphere as secondary pol-lutants. High values of these parameters are generallyindices of pollutant aging.7OZONE (ppm) bSca, (m-' xFIG. 10. The mean and standard deviation of temperature, turbulence, ozoneand bsoat for the afternoon soundings at El Monte; solid line is the mean, dashedlines represent the mean fl standard deviation.OCTOBER 1977 HUSAR ET AL. 1095TABLE 2. Vertical integrals of the profiles and their standard deviations.Meteorological parameters Hawthorne El Monte Ontario Riversideand pollutants Mean =!CU Mean *U Mean fU Mean faTemperature A.M. 30180 4816 30850 2621 32310 4572 32310 4389("C-m) P.M. 32740 4511 33470 5242 36700 3901 37060 4206Turbulence A.M. 2170 1622 1658 902 2682 1341 2219 95 1(cd s-1-m) P.M. 2268 1402 3097 1256 3999 1609 3901 1244Condensation nuclei A.M. 19.51 10.5 15.12 7.5 34.62 15.9 20.48 9.0(1012 m") P.M. 31.70 13.9 36.58 14.9 53.16 22.4 36.09 20.2Ozone* A.M. 100 40 142 61 141 63 135 51(PPm-4 P.M. 140 69 185 79 199 86 210 77NO, A.M. 76.8 38.0 79.2 51.0 132.9 35.0 108.5 46.0(PPm-m) P.M. 64.8 38.0 101.2 27.0 74.4 41.0 53.6 24.0beat A.M. 0.295 0.166 0.490 0.239 0.671 0.269 0.507 0.254(dimensionless) P.M. 0.254 0.135 0.541 0.219 0.349 0.202 0.476 0.312* The units of the vertical integrals were obtained as ppm-meters (rather than ppm-ft)6. Variability of the profilesThe profiles presented in the previous sections areaverages over 24 days of sampling. In some instances,the variability about the mean is also of interest.For illustration purposes, the mean and the standarddeviation of temperature, turbulence, ozone and bscatfor the afternoon soundings at El Monte are shownin Fig. 10. The standard deviations typically rangefrom 15% of the mean to over 400/, of the mean;the pollutant parameters vary more than temperatureand turbulence.Since it is not convenient to display the profileswith their standard deviations for all sites for eachparameter for both morning and afternoon, we chooseto represent each profile with a single number-itsvertical integral. The integrations were carried outfrom ground level to 5200 ft MSL, and the resultsare listed in Table 2. The vertical integral of thelight-scattering coefficient, Le., the average opticaldepth of the summer smog layer, is between 0.25 atHawthorne and 0.54 at El Monte. The integrals ofcondensation nuclei, ozone and NO, do not havea simple physical meaning, but they may be usefulin matching photochemical smog models with ob-served data. Also, the standard deviations of theintegrals is a measure of the variability of the profilesthemselves.7. Summary and conclusionsMorning and afternoon "grand average" verticalprofiles for Hawthorne, El Monte, Ontario and River-side show an erosion of the nighttime radiation in-version during the day and increased temperaturesand mixing due to surface heating in the inland areas.The semi-permanent subsidence inversion is also seenat higher level over the four sampling stations.Average contaminant profiles indicate high valuesof primary pollutants (NO, and condensation nuclei)in the source area near the coast at Hawthorne. Onthe other hand, the receptor site, Riverside (severalhours of transport time inland), is characterized byhigh values of secondary pollutants, 03 and bscat, andlow NO, and CN. Morning profiles of O3 at all loca-tions indicate higher concentrations aloft than at thesurface. This deficit is attributed to the scavengingof ozone by primary emissions.Acknowledgments. This work was supported by theCalifornia Air Resources Branch. The authors ap-preciate the support and guidance provided by thelate Dr. D. G. Hutchinson, Mr. Harris Samuels andDr. Jack Suder.REFERENCESBlumenthal, D. L., T. B. Smith, W. H. White, S. L. Marsh,D. S. Ensor, R. B. Husar, R. B. McMurry, S. L. Heisler and P. Owens, 1974 : Three-Dimensional Pollutant Gradient Study-1972-1973 program. Tech. Rep. MR174FR-1261,prepared for the California Air Resources Board by Mete-orology Research, Inc. [NTIS PB 241982/66].Chamberlain, A. C., 1967 : Radioactive aerosols and vapours.Conteip. Phys., 8, 561-581.Covert, D. S., R. J. Charlson and N. C. Ahlquist, 1972 : A studyof the relationship of chemical composition and humidityto light scattering by aerosols. J. Appl. Meteor., 11, 968-976.Davies, C. N., 1974: Size distribution of atmospheric particles.Aerosol Sci., 5, 239-300.Edinger, J. G., 1973 : Vertical distribution of photochemicalsmog in the Los Angeles basin. Environ. Sci. Tech., 7,Ensor, D. S., R. J. Charlson, N. C. Ahlquist, K. T. Whitby, R. B. Husar and B. Y. H. Liu, 1972: Multiwavelengthnephelometer measurements in Los Angeles smog aerosol.J. Colloid Interface Sci., 39, 242-251.Gartrell, G., and S. K. Friedlander, 1975: Relating particulatepollution to sources: The 1972 California aerosol charac-terization study. Atinos. Environ., 9, 279-299.Hidy, G. M., et al., 1974: Characterization of aerosols in Cali-fornia. Final Report to California Air Resources Boardunder Contract 358. [NTIS: Vol. 2, PB248799/66; Vol. 4,PB247947/661.Hill, A. C., 1971: Vegetation: A sink for atmospheric pollutants.J. Air Pollut. Control Assoc., 21, 341-346.247-252.1096 JOURNAL OF APPLIED METEOROLOGY VOLUME 16Husar, R. B., and K. T. Whitby, 1973: Growth mechanismsand size spectra of photochemical aerosols. Environ. Sci. Tech.. 7. 241-247.Lundgren, D. A., 1970: Atmospheric aerosol composition andconcentration as a function of particle size and of time.J. Air Pollut. Control Assoc., 20, 603-608.> I--- -~~ ..-, K. T.and B. Y. H. Liu, 1972: PhysicalMacCready, P. B. Jr., 1964: Standardization of gustiness valuesfrom aircraft. J. Appl. Meteor., 3, 439449.Whitby, K. T., R. B. Husar and B. Y. H. Liu, 1972 : The aerosolsize distribution of Los Angeles smog. J. Colloid Interface Sci.. 39. 177-204.nisms governing the dynamics of Los Angeles smog aerosol.J. Colloid Interface Sci., 39, 211-224., W. H. White and D. L. Blumenthal, 1976: Direct evidence->-. - ~of heterogeneous aerosol formation in the Los Angeles smog.Environ. Sci. Tech., 10, 490-491.White,. W. H., 1976: Reduction of visibility by sulphates inphotochemical smog. Nature, 264, 735-736.