1. Introduction
Atmospheric chemistry within the planetary boundary layer in urban environments is a special area of investigation of near-surface tropospheric phenomena. More interest in this subject came about from rising environmental exposure to air pollution following the industrial revolution and the Second World War. Most of the knowledge of urban air chemistry is drawn from studies in the mid-latitude, high-income countries. Comparable research in tropical regions, particularly in sub-Saharan Africa, is limited. An increasing number of atmospheric studies in Nigeria, Africa’s most populous country, have been reported since the early 2000s, for example [1,2]. These have concentrated on regional population centers around the oil-producing Niger River Delta region. The studies have reported ambient data for airborne particles and for commonly monitored gases: carbon monoxide (CO), sulfur dioxide (SO2), nitrogen oxides (NOx = NO + NO2), volatile organic compounds (VOCs), and ozone (O3). One particularly important urban center is Lagos (6.451° N; 3.388° E), located just north of the equator on the shore of the Gulf of Guinea and west of the Niger River Delta region.
The metropolitan area of Lagos is a fast-growing, high-density “mega-city” with a population of more than 20 million, making it the largest city in West Africa. It is mainly a coastal city with some small islands south of a large lagoon. The economy of the city is diverse and depends in part on its regional financial activities, local industry, and its seaport complex. The metropolitan area has a variety of manufacturing, including power generation, an automobile assembly facility, steel and other metal works, petroleum storage, and fabric production as well as a major agricultural market associated with the seaport. With its current rapid growth, the Lagos area experiences stress on its on-road transportation system. Most of the roads throughout the city are highly congested, with as long as four-hour commutes at times for a journey that typically should last less than an hour [3]. A large fraction of the vehicle fleet is older than 15 years, with poor maintenance and an uncertain effectiveness of emission controls. The economic growth also has affected energy supply; the demand for electricity exceeds the generation capacity, leading to a shortage of electricity. The unreliable electricity supply has resulted in a proliferation of diesel electric generators and other fossil-fuel-based units. The variety of transportation, seaport, alternative power generation, and manufacturing pollutant sources along with residential emissions (cooking) and open waste burning provide important sources driving atmospheric chemistry in the city and contributing to urban air pollution [3,4].
Lagos is in a tropical forest region just south of an agricultural and Savanah regime and grasslands merging into the southwestern edge of the Sahel–Sahara Desert regime. Lagos’s climate is dominated by a wet season and a cooler dry season. Its local climate is strongly dependent on a sea breeze flowing from the south and southwest. As one would expect in the tropics, the temperature is warm year-round, in the 27–30 °C range and moist, generally with a daytime relative humidity exceeding 50%. Long-range transport also can play a key role in the observed air chemistry, particularly the well-known large-scale natural “source” phenomenon of blowing dust from the desert across the west coast and across the Atlantic Ocean to North and South America [5]. Further south, wildfires in the jungles of central Africa have produced smoke plumes that have also been observed across the Atlantic Ocean [6].
This study is intended to complement and extend previous atmospheric composition measurements in southern Nigeria. The project was organized by EnvironQuest with the World Bank’s support and interest in aiding Nigeria and Lagos officials to develop a plan to manage the severe air pollution in the metropolitan area [3]. The project involved comprehensive chemical and meteorological measurements at multiple representative sites in the Lagos metropolitan area from August 2020 to July 2021 [7]. A selection of observations of trace gases and fine and coarse particles provides major insights into the air quality of Lagos, which are interpreted in the light of knowledge from mid-latitude studies and multiscale tropospheric phenomena of interest. The results provide a contemporary baseline for quantifying changes in Lagos air quality with expected future changes in emissions, meteorology, and urban demography.
2. Methods and Materials
Measurements during this study were conducted at six locations in the Lagos area (Table 1). Included in the table is a brief summary of the site characteristics.
The location of each measurement site is shown in Figure 1. The orientation of the stations is roughly north–south, with representation at the south and north extremes of the east–west siting. With a sea breeze, pollution from the Lagos seaport complex south of JAN and from sources in the city is transported inland towards IKO, EPA, and ABE.
The Lagos metropolitan area is potentially exposed to regional and larger-scale contaminants. These include: (a) the Niger River Delta for oil and gas production and refineries at Port Harcourt 290 km and Warri 170 km to the southeast; (b) the Savanah grasslands to the north for vegetation debris and wildfires, and (c) the Sahel and Sahara for rising dust storms and O3 transported over West Africa (Figure S1).
The instrumentation used at the six stations included continuous chemical species monitoring and surface meteorology and periodic sampling for organic species and particles. The instrumentation is listed in Table 2. The manufacturers’ detection limits and the estimated accuracy of the Zephyr are listed in Table S1. The gas and particle instruments were set at a height of 1.7 m; the anemometer was located at a height of 10 m. The stations were located with 25 m spacing to the surroundings, except for JAN, which had space limitations.
The gas and aerosol sampling and meteorological measurements at the sites were operated continuously by field technicians following standard operating procedures. These are given in detail in [7]. The gas measurements were recorded with the Zephyr manufacturer’s calibration. These were verified with the AQM 65 analyzer, which included an internal calibration procedure. The PM2.5 and PM10 data were calibrated using the filter-gravimetric measurements [7] (See also Section 4.3). The PM1 observations were not used in the study analysis since they were not calibrated with a “secondary reference” standard.
At each site, the field measurements included canister sampling for the laboratory characterization of VOC species and filter sampling for PM mass concentration and composition determination. Two canister samples were collected per site every month (mid-month and the last week in the month). In addition, during the last week in the month, one duplicate (for external laboratory analysis) and one blank were also collected. Hence, a total of four canisters were deployed per site per month. Overall, a total of 288 canisters (48 canisters per site) were collected at the six sites (August 2020 to July 2021), comprising 72 blanks, 72 duplicates, and 144 samples. Before sampling, the canisters were manually cleaned by filling them with clean dehumidified air, heated, and evacuated per Lagos Air Quality Monitoring and Source Apportionment (LAQMSA) Study standard operating procedure. The canisters were tested until certified free of contamination (target VOC concentration <20 pptv).
Prior to leaving the laboratory for the field, a canister’s initial vacuum pressure was verified using a gauge to ensure that the vacuum pressure was between −28 and −30 in Hg. The canister flow controller was calibrated through a pressure–time measurement to verify or adjust the canister flow rate to achieve 1.93 cc/min for a 3.2 L canister or 3.6 cc/min for a 6.1 L canister prior to field deployment. The canisters were programmed to run for 24 h, midnight to midnight. Once the canisters were located at a site, technicians verified the timer operation and documented the flow rate. Subsequent monitoring was conducted through an online portal during the 24 h sampling protocol.
Table 2Species measurements showing instrumentation and sampling averaging time.
| Measured Parameter | Monitoring Instrument | Averaging Time |
|---|---|---|
| Field Measurement Methods | ||
| CO, CO2, NO, NO2, SO2, | EarthSense Zephyr [8] (electrochemical detectors for gases; optical light scattering for particles; PID for TVOC) | Continuous |
| O3 | ||
| PM2.5, PM10 | ||
| Trace gases and particles (same as |
Aeroqual AQM65 [9] (electrochemical gas detectors; PM light scattering) | Intermittent/Continuous |
| for calib. of Zephyr; 5 min avg. | ||
| PM2.5, PM10 | ARA n-FRM Sampler | Time-integrated, 24 h |
| VOCs, CH4, N2O, CO2, CFCs, HFCs | Entech Silonite Canister | Time-integrated, 24 h |
| Precipitation | Davis Vantage Pro2 Weather Station | As detected |
| Air Temperature and Rel. Humidity | Continuous, averaged 5 min | |
| Wind Speed and Wind Direction | ||
| Laboratory Methods | ||
| PM Gravimetry for PM2.5 and PM10 | RADWAG Microbalance | Post and pre-sampling |
| Elements | RIGAKU X-ray Fluorescence (XRF) | |
| Ions | Dionex ICS-3000 | |
| Organic Source Markers | Agilent GC-MS | |
| EC and OC | ANSTO MABI Sunset Analyzer [10] | |
| GHGs | GC/FID/ECD/MS instruments | |
PM2.5 and PM10 samples were collected at the six sites on a 3-day schedule for 24 h from 0000 to 2400 h with a low-flow (16.7 L/min) ARA FRM sampler. This covered a total of 216 site-days during the study period. Of the 2928 PM2.5 and PM10 filter samples received, 2880 PM2.5 and PM10 filter samples were considered valid after level 1 data validation. Twenty-five percent of a total of 864 filters (432 PM2.5 samples and 432 PM10 samples) and 86 field blanks were submitted for comprehensive chemical analyses. Twenty-five percent of the selected samples were sent to a reference laboratory for an interlaboratory comparison of the analytical results.
PM2.5 and PM10 samples were collected on 47 mm Pall Teflon™ filters and quartz fiber filters. All filters were housed in petri slides and stored under dry conditions at 4 °C. Prior to use, the quartz filters were pre-baked in a furnace at 900 °C for 4 h to remove residual organic species (carbon level) associated with the filters as obtained from the manufacturer. Teflon™ and PTFE filters were equilibrated under constant temperature (20 ± 3 °C) and humidity (30–40%) for 24 h prior to gravimetric analysis. Teflon™ filters, both before and after sample collection, were weighed using a microbalance, and the differences in filter weight were divided by the 24 h sample volume to calculate the PM2.5 and PM10 mass concentrations (µg m−3).
The valid data rate exceeded 99%. Eight filter samples (three Teflon™ filters and five quartz filters) were flagged for meteorological conditions such as a combination of temperature, wind speed, and wind direction or relative humidity and precipitation anomalously impacting the concentrations of atmospheric pollutants.
Data validation was a critical component of the overall quality control and assurance program evaluating single- and multicomponent data consistency. The validation of the field and laboratory analysis of the collected data was part of the LAQMSA network protocols. The project used a number of standard procedures for maintaining the field operations as well as the analytical laboratory. The quality control and assurance process and results of procedures, field and laboratory audits and intercomparisons are given in detail in [7].
3. Basic Resources for Air Chemistry
3.1. Emissions Sources
Emissions of gases and particles for oxidant photochemistry have been reported in part by Fakinie et al. [11]. Emissions in Lagos derive from a large variety of sources. Many of these sources are combustion-derived, including electric power generation, ground transportation, the seaport, open or municipal burning, and residential cooking. Other emissions include sources such as refining, chemical processing, other manufacturing, and commercial activity. There are also natural sources of potential concern such as wildfires, street and soil dust, and vegetative emissions. It is unclear from the literature if a complete inventory for Lagos has been assembled. However, there is one example of an annual inventory [11] that covers most combustion sources, except for the in-harbor shipping. This assessment for 2007–2016 includes solid wastes, fuel wood, gasoline and diesel fuel, aviation fuel, and saw dust. The first three are important for Lagos. Their estimates of annual combustion emissions in kilotons per year (ktpy) are: PM10 188.5; CO 5921; SO2 11.2; NOx 348; and VOCs 4.14. Kerosene use in cooking was identified as a major contributor for NOx, SO2, and CO. These estimates compare with values derived from aircraft sampling of CO (1440 ktpy); NOx (30 ktpy); and VOCs (37 ktpy) with the stated uncertainty of +250 to −60% [11]. Additional emissions data have been reported. The estimates of SO2 by Fakinie et al. [11] differ from Olatunji et al. [12]. The difference appears to be a large contribution of SO2 emissions from kerosene combustion. In any case, the two estimates suggest low SO2 emissions compared with NOx but larger emissions than VOCs from incomplete combustion.
Additional emissions estimates were reported for individual pollutants in [13] for local traffic, industry, and residential cooking for the Okobaba area adjacent to the Lagos Lagoon. This inventory provides insight into one high-population-density area of the city but is not useful to characterize the entire metropolitan area. Oketola and Osibanjo [14] summarized the emissions from generic sources for Lagos but have not organized this information in a spatial summary covering the city.
Marais et al. [15] reported emissions estimates for Nigeria using satellite data combined with literature values. Their estimates are higher than the other estimates for the listed species and show emissions are most heavily concentrated in southern Nigeria.
From the limited emissions estimates available for Lagos, it is clear that a wide range of estimates exist, which are difficult to interpret. An accurate emissions inventory is needed to support planning for the management of air pollution in Lagos. Considerable effort needs to be devoted to the preparation of a spatial-temporal inventory with a consistent methodology. In the meantime, investigators need to rely on receptor modeling to gain insight into emissions in the metropolitan area.
3.2. Meteorology
In addition to the different pollutant emission rates present in the city, ambient concentrations are affected by meteorological conditions. These range in spatial and temporal scale from the trade winds and the monsoons to mesoscale circulation and local sea breezes. The large-scale meteorology is dominated by the West African wet and dry periods; the former occurs in the warmer (June, July, and August) months, and the latter occurs in the cooler (November, December, January, and February) months. In the dry season, the southerly and southwesterly winds penetrate inland with the shift northward of the intertropical discontinuity (ITD). In summer, the ITD moves southward, allowing the dry desert winds to shift to the Gulf of Guinea coast, mixing with the southerly winds. Nigerian wind data suggest the sea–land breeze is strongest in the fall and weakest in spring [16,17].
The dispersion and advection of air pollutants with the surface winds are potentially important for chemistry conditions. The wind roses in Figure 2 show the predominant wind sector at each of the monitoring sites. The frequency of calm conditions (wind speed < 0.5 m/s) ranged from a minimum of 7.4% at IKO to a maximum of 33% at ABE. South- (~13%) and north-westerly (~11%) flow were predominant at ABE. Wind flow at JAN and EPA were similar in their distributions except for their sectors of highest occurrence. While JAN had the highest frequency of wind blowing from the south or southwest (17%), EPA was from the southwest (22%). Although both sites have significant contributions from the west, the frequency was twice as high at EPA compared with JAN. Based on the wind roses, EPA appeared not to be affected by southerly winds, likely due to local obstructions.
Overall, all the sites were characterized mostly by south-westerly to westerly winds, except for NCF, which had predominant wind sectors from the north- and south-easterly directions. The winds at NCF reflected the sea breeze condition.
Table 3 provides a summary of the meteorological data during the study (August 2020–July 2021). The mean temperature and relative humidity were characteristic of the tropics. The temperature was mild to warm with a range of <10 °C. The relative humidity was typical of tropical conditions near the ocean, with a range from ~50% during the day and in the dry season to >90% at night and in the wet season. Winds generally are light across Lagos. The diel variation in wind speed showed an afternoon peak at all sites, with a maximum temperature in the afternoon and a minimum relative humidity. The higher temperatures during the afternoon sometimes resulted in an increase in convective activity, leading to an increase in wind speed during the afternoon. The near steady temperature and elevated humidity provide for “ideal” photochemical and nocturnal chemistry conditions in the Lagos atmosphere, modulated by the wet and dry season sunlight.
Across the sites, the observed maximum wind speed decreased in the order among the sites: UNI > NCF > IKO > JAN, and EPA > ABE. The two highest observed wind speeds of 4.1 and 3.9 m s−1 were detected at UNI and NCF, respectively, while the mean wind speed observed at all sites was ≥1 m s−1, except ABE, which was slightly less than the other sites.
The number of sunlight hours per day is a useful measure of the potential for photochemical processing. The climatological averages for Lagos [18] were as follows: in the dry season (November–March), 6.2 h/d, sunlight 55% of the time during daylight; wet season (June–September), 3.8 h/d, sunlight 27% of the time in daylight.
4. Results and Discussion
The gas and particle measurements in Lagos give a picture of the urban air chemistry currently occurring in a tropical mega-city. The continuous and semi-continuous data from this study provide insight into the major chemical reaction pathways associated with air pollution. Described below are the atmospheric characteristics across Lagos.
4.1. Gases
4.1.1. Carbon Monoxide
Carbon monoxide originates from incomplete carbon fuel combustion. In the troposphere, it participates in the photochemical oxidant cycle in relatively slow reactions. In cities, the photochemical and associated reactions unrelated to CO are sufficiently rapid that CO is essentially inert, and ambient concentrations in urban areas depend solely on emissions and meteorology.
The CO measurements at two sites across Lagos, as shown in Figure 3, illustrate the average diel, seasonal, and weekday variation, as they represent an “inert” gaseous emission and dispersion from city-wide sources including transportation, electricity generation, the seaport, industry (including sawmills), residential–commercial cooking, sawdust, and open burning. The diel variation in CO concentrations was similar northward inland with the southerly sea breeze, with both NCF and EPA (like the other sites) showing morning peaks (after 0600 h local time) characterizing the onset of daily human activity, including morning traffic. At midday, a minimum occurred, with increased atmospheric dispersion, followed by an evening increase to a peak at 2200 h local time, with traffic, followed by a decline in residential activity and an ease of the sea breeze with the onset of a nocturnal inversion. This diel variation is characteristic of virtually all cities studied across the world.
Seasonally, average CO concentrations tended to be steady through the months, with elevated values in in the dry period of winter and lower concentrations during the wet months of summer. The weekday average concentrations were variable but tended to show minima on the weekends that were associated with reductions in commercial and industrial activity and traffic.
The average CO concentrations are typical of contemporary concentrations seen in mid-latitude North America [19]. Earlier studies in Lagos prior to 2016 reported higher values of CO, e.g., Njoku et al. [20] (21 source-related sites, unspecified time) 11–46 mg m−3; Azeez et al. [1] (3 sites, 3 months) 16.1 mg m−3; and Olajire et al. [21] 19.3 mg m−3 near a roadway for 72 days. More recently, two short-term campaigns reported CO concentrations. The first involved one month of sampling in 2017 and distinguished between residential and transport (roadway) influences. For residential, CO was 3.4 ± 1.6 mg m−3, while transport CO was 3.6 ±1.9 mg m−3 [22]. The second study took place over 40 days with sampling at five sites. Their mean CO concentration was reported as 3.1 mg m−3 [23].
4.1.2. Sulfur Dioxide
Sulfur dioxide is reactive in the troposphere. It is oxidized through aqueous phase processes and in the gas phase via the photochemical oxidant cycle to eventually form particulate sulfate (see also Section 4.2.2). The gas in cities generally comes from the oxidation of sulfur from fuel combustion, although it has some natural contributors, including wetlands or marine S gases such as hydrogen sulfide or dimethyl sulfide. The SO2 measurements in Lagos in 2020–2021 were mostly below the instrument detection limit of 8 µg m−3. The Lagos emission estimates for SO2 discussed above indicate that they are far less than other gases and appear to be associated mostly with the combustion of refined petroleum, including gasoline, diesel, and kerosene.
Some insight into the characteristics of higher SO2 concentrations can be gained from the outlier data in the uppermost 5% range. At the shoreline (NCF), measurements indicated outliers from 10 to 50 µg/m3, mostly during the day, with one value of 100 µg m−3. Observing SO2 outliers at the shore does not necessarily mean local sources. The high peak values may reflect SE winds (Figure 2) potentially blowing traces of the gas from as far away as the petroleum production and refining region in the Niger River Delta. For JAN, daytime outliers (15–200 µg m−3) exceeded the night values of 10–90 µg m−3. At UNI, afternoon and evening outliers were the lowest (15–200 µg m−3), while morning and night outliers were 10–200 µg m−3. EPA showed outlier values from 10 to 200 µg m−3; ABE values ranged from 10 to 200 µg m−3. IKO had the fewest outliers, except in the afternoon, ranging from 10 to 200 µg m−3. The outliers suggest that occasional high SO2 concentrations occur at all sites, even at the shoreline. The origins of these outliers are uncertain but could be associated with (a) occasional local use of imported fuel of high S content or upsets in local industrial processing, or (b) less likely, from regional transport.
Prior to 2016, SO2 concentrations reported around Lagos included maxima of 260 µg m−3 on the roadside [21] and 1400 µg m−3 [1]. Raheem et al. [2] reported site average SO2 diel concentrations from five Lagos sites from 2003 to 2006. Their results indicated that in the dry season SO2 concentrations peaked at 38 µg m−3 in the morning at 1015 h and in the afternoon and evening after 1400 h at ~42 µg m−3. In contrast, in the wet season, the SO2 morning peak dropped to about half that of the dry period at 16.4 µg m−3 and ~21 µg m−3 for morning and evening, respectively. Obanya et al. [22] reported, in 2017, mean values of residential SO2 concentrations of 155 ± 80 µg m−3 and mean vehicle transport corridor SO2 concentrations of 310 ± 235 µg m−3. These concentrations were higher, on average, than those seen in 2020–2021 and are seemingly highly variable. One reason for this could be the elimination of “dirty” imported fuels [24] from the market as part of the Nigerian government instituting low S requirements for diesel, marine, and residential fuels. The increasing demand for petroleum fuels in Nigeria continues to be a major source of high S and methanol fuel imports [25], making it a challenge to interpret the ambient SO2 levels.
4.1.3. Nitrogen Oxides
Nitrogen oxides, NOx (NO + NO2), coupled with VOCs, are key reactants for photochemical oxidant formation. The sources of NOx are principally from fossil fuel combustion or biomass burning. NOx combustion is dominated by NO with a lesser fraction of NO2 [26]. The diel pattern for ambient NOx concentrations shows their reaction sequence (Figure 4). The Lagos sites displayed a characteristic concentration variation that is well-known in studies of NOx-O3 chemistry. The gases peak in the morning and evening in a similar way to CO, consistent with combustion sources in the city. Upwind, at NCF, the variation in the average NO and NO2 showed a peak early in the morning, followed by a second peak in the evening. At the other sites, the early morning peak of NO concentration was narrower. The NO2 concentrations showed a strong morning peak, paralleling NO, and a secondary peak in the evening. The peak of NO2 concentration paralleling the NO peak is ascribed partly to the reaction of NO and residual O3 to form NO2. The morning peak concentration of NO at NCF of 58 µg m−3 was lower than that at EPA at 120 µg m−3. At the other sites, the morning peaks ranged from 80 to 95 µg m−3. The morning peaks for NO2 ranged from ~60 µg m−3 at NCF to 80 µg m−3 at EPA. The evening peaks for NO2 concentration ranged from less than 50 µg m−3 at NCF to 90 µg m−3 at EPA, near 1600 h. The other sites range from 55 to 95 µg m−3.
The monthly average NO and NO2 concentrations tended to have opposite maxima and minima. NO tended to be highest in the dry period, while NO2 concentrations tend to be highest in the wet period. The weekday averages both indicated the weekend minima with reductions in urban activities.
Raheem et al. [2] reported site average morning and evening NOx peaks of 11 µg m−3 and, after a minimum, ~12 µg m−3, respectively. A comparison during the wet season showed a morning peak of 6.5 µg m−3, and the afternoon peak was ~7 µg m−3. Obanya et al. [22] found zero concentrations or “non-detect” of NO2 at both residential and transport-related sites. These levels were lower than those seen in this study.
The ambient concentration ratio of NOx to CO is of interest in relation to the emission ratio. The 2016 average emission ratio for the combustion sources from Fakanie et al. [11] was 0.058, similar to the value of 0.044 of Marais et al. [15]. If we take the morning peak estimate of NO, representing a minimum reaction of ambient species (NO), the NOx/CO ratios by site were 0.007–0.0144, with a site average of 0.012. This variability in the “combustion index” can be ascribed to differences in the methods to estimate emissions, combined with a lack of knowledge of the combustion processes forming NOx and CO that are relevant to Lagos activities.
4.1.4. Volatile Organic Compounds (VOCs)
VOCs represent the second key reactant in the photochemical cycle; they are composed of a wide variety of hydrocarbons and oxygenated organic compounds. The Marais et al. [15] inventory indicates inordinately high VOC levels compared with other cities.
The ratio of the morning peak VOC to NOx is of a coarse measure of O3 formation sensitivity. The distinction between NOx and VOC sensitivity is sometimes taken as a ratio of 8–10 [27]. The results of a study by Njoku et al. [20] for pre-2016 conditions suggested VOC sensitivity across the city. If compared with an example of a motor-vehicle-dominated VOC environment of a subtropical city, Atlanta, GA, USA [28], the ratio is 9–10.
The large number of VOCs present in the urban air has created difficulties for interpreting simple measures of total VOC concentrations since the 1970s. Most observational studies have concentrated on identifying and quantifying the VOC species to seek similarities and differences in VOC composition. A sense of the species in VOCs is obtained from a GC-MS analysis of the canister samples. The average speciation from these samples is shown in Figure 5. The mixture was a combination of mostly light (<C6) alkanes, alkenes, and aromatics according to this assay. These have been reported in virtually every sample of VOCs obtained previously for city air and are associated with petroleum products, including gasoline, kerosene, and diesel fuel [29].
The species distribution by site indicates, with the exception of isoprene, VOC species are lowest at NCF near the Gulf. JAN, also near the Gulf, shows low concentrations for all except ethane, ethylene, and acetylene. Further inland, the VOCs accumulate from the emissions of transportation and industrial and residential cooking sources.
An unusually high concentration of dichloromethane was seen at IKO and EPA. This chlorocarbon is a solvent used in industrial manufacturing and degreasing facilities. Its presence in high concentrations at these two sites implies local non-transportation sources.
Isoprene is a hydrocarbon predominantly from vegetation emissions. In many locations, it is identified with the ozone formation cycle and with secondary aerosol production. The average isoprene concentrations were low compared with midlatitude values, which tend to be above 0.10 to 1.0 ppbv (e.g., [28]). The isoprene concentration at IKO appeared to be an exception, with its location closest to the tropical forest boundary to the north. At the median levels seen in the Lagos data, this hydrocarbon level suggests a less important biogenic contribution to O3 and aerosol formation than in other cities.
A useful relation to identify a motor vehicle source can be examined using a toluene-to-benzene ratio, which is characteristic of gasoline blends. This ratio, from Figure 6, ranged from 2.5 to 6.3 in the canister data. This compares with the ratio found in the 1990s to mid-2000s in Atlanta and North Carolina for non-methane organic carbon --NMOC (a part of VOCs), as in ppbC, ranging from 1.7 to 2.3 [28]. Aircraft observations over Lagos indicated a ratio of approximately 1.5 [15]. The relatively large variation in the toluene–benzene ratio was identified mainly at ABE and IKO, suggesting multiple sources of aromatic compounds at these sites.
The toluene/benzene ratio from this study differs from that Olajire and Azeez [29] reported for earlier 2014 sampling along Oba-Akran Road (ratio 0.45–0.75). Drozd et al. [30] measured the ratio of 1.4 in vehicle exhaust in California. The differences can be attributed, in part, to the source of the gasoline blends, which vary somewhat by refinery. Hypothetically, the ratio may be affected by fuel from unofficial artisanal refining in the region. Other examples of characteristic source ratios include the ethane-to-propane ratio. It ranged from 0.2 to 0.7 for the Lagos sites vs. 0.15–0.6 in the 1999–2007 Atlanta samples. The predominant alkane concentrations, presence of aromatics, and the consistency of the ethane–propane ratio affirms the importance of petroleum-based emissions to O3 precursors.
4.1.5. Ozone
Ozone is an important indicator for the oxidation products of the photochemical NOx-VOC reaction sequence. Local production of O3 follows the solar radiation cycle, resulting in a maximum concentration by midday. Solar radiation and ambient temperature near the equator are annually more constant than in the mid-latitudes. Unlike in the mid-latitudes, seasonally averaged O3 concentrations in Lagos follow the dry period cycle. The maximum average and extreme concentrations are observed in the dry months, while lower average concentrations are seen in the wet months [7]. The potential for increased O3 formation during the dry season is partly due to the added hours of sunshine compared with the wet season (for example, see Section 3). It also linked with the abundance of NO2 in the dry season when clouds and precipitation tend to scavenge and deplete the NO2 concentration.
The average diel concentration profiles for the sites show similar time-dependent characteristics (Figure 7 and Figures S2–S5). The examples in Figure 6 indicate the maximum O3 concentrations occur at ~1300 h, with nearly equal values (NCF, 50 µg m−3 and EPA, 52 µg m−3). Interestingly, the nocturnal O3 minimum for EPA was <10 µg m−3, while at NCF it was 27 µg m−3. This variation was present despite the nocturnal NO levels being uniformly ~60 µg m−3. The influence of NO-O3 titration that these results suggest is readily illustrated in Figure S6. The variable residual O3 concentrations at night with NO present also suggest a nocturnal chemistry involving N2O5 and/or NO3. An offshore component seen at NCF may represent a land breeze component. The combination of the temporal and spatial characteristics of the average diel NOx, speciated VOC, and O3 concentrations exemplifies the locally dominant chemistry within the metropolitan area, with the peaking of O3 near noon, for example. If this were not the case and regional transport were frequently involved, the diel elevated O3 would tend to extend well into the afternoon and possibly into the evening [31].
The monthly variation in O3 concentration showed the maximum in the dry season and the minimum tendency in the wet season. The weekday variation near the Gulf shore at NCF was variable, while EPA showed a distinct weekend increase. At both sites, there was a depletion of NOx on the weekend. The four other sites showed a weekend maximum for O3 with depleted NOx (Figure 6 and Figures S2–S6).
For comparison, Abulade et al. [23] reported the mean O3 concentrations for their sites as 32.5 ± 38.7 µg m−3, which was similar to those found in this study. Raheem et al. [2] reported average O3 concentrations for the 2003–2006 period. During the dry period, the six-site average maximum O3 concentration occurred at 12:45 and was 155 µg m−3. In the wet period, the average maximum occurred at 1400 h and was 32 µg m−3. A comparison of the wet season maximum with the study averages indicates that the Raheem et al. wet season results were consistent with this study’s overall average. However, their dry season result was higher. This difference suggests that the maximum O3 in the dry season concentrations in earlier years may have been generated by more active NOx-VOC chemistry than in 2020–2021.
Theregional character to the O3 concentrations with NOx. CO, formaldehyde, and glyoxal can be assessed from satellite and aircraft flights within 500 km of Lagos [15]. This analysis notes that extremely high VOC levels are inferred. The analysis indicates that modeled emissions distributions include a large regional component of elevated reactant concentrations, and O3 was present in Lagos’s surroundings. The satellite retrievals for O3 suggested O3 levels of 118–126 µg m−3 at ~300 m height over the Lagos–Niger River Delta region. This result indicates that with vertical mixing a substantial potential exists for a regional component of urban chemistry in Lagos.
As noted above, the seasonal O3 concentration distributions showed increased O3 concentrations with NO2 in the dry months, but significant O3 concentrations were found in the wet period as well. Both periods have sunny days that would be conducive to O3 formation. The dry period experiences increases in large-scale westerly to southwesterly Harmattan winds from inland (e.g., Figure 2 in Marais et al. [15]). CO and O3 concentrations have been observed at lower altitudes close to the city, potentially influencing local O3 concentrations. An example of vertical profiles in the wet season is found in Figure 7 [32]. With high concentrations of O3 aloft, the potential exists for downward mixing into the Lagos air, causing extreme cases of O3 exposure in the city. Possible evidence for such an occurrence could be inferred from this study in the occasional incidents of high concentrations of in the range of 100–250 ppb, above the 95% level [7]. Another incident was reported from O3 sounding above Cotonou, west of Lagos, in which an O3 layer exceeding 295 ppb concentration was observed at just below 2 km of altitude. The high O3 aloft in the first case is attributed to mesoscale air flow from the southeast over wildfires. O3 precursors from the fires are lifted up into the lower troposphere and carried by a convective baroclinic flow associated with the ITD. The latter case was associated with the northeasterly flow over fires passing over Lagos on 20 December 2005. Minga et al. [33] used chemical transport modeling to show that the injection aloft of large quantities of VOCs from a petrochemical plant or refinery accident was required to achieve the high level of O3 concentration. Neither the frequency of occurrence of these regional incidents nor their influence on ground-level O3 has been established yet. However, one case study in the wet season of 2014 [34] indicated the surface impact of regional wildfire smoke from convective mixing is estimated as: 0.15 mg m−3 CO, 10–20 µg m−3 O3, and 5 µg m−3 PM10.
Figure 7Example of vertical profile for O3 and CO concentrations for 3 July 2003, showing a layer of O3 and CO aloft above 2 km from measurements from an aircraft descending to Lagos airport at approximately 1100 h. Ll and L2 designate layers for the authors’ modeling study. Note that the instrument for O3 concentration apparently stopped functioning at ~800 m of altitude. Figure from Global Commons, European Geoscience Union [32].
[Figure omitted. See PDF]
4.2. Aerosol Particles in Lagos Air
In addition to gas-phase chemistry, urban air chemistry involves a complex of suspended particle (aerosol) chemistry and airborne particle interactions, many of which remain poorly understood. In this section, we describe the particulate measurements from the Lagos study [6]. The results are interpreted in the light of the city emissions and links with air chemistry. The former is discussed in detail with several different receptor modeling calculations [35,36,37,38].
Urban aerosols originate from two classes of sources: “primary” from ground-level activity from natural debris, such as soil dust and detritus, to anthropogenic sources, including transportation, industry, road dust, and waste burning, and “secondary” from chemical reactions occurring in the atmosphere involving the oxidation of gases such as SO2, NOx, and VOCs. Most of these reactions are linked with the photochemical oxidant cycle; they include both homogeneous gas to particle formation as well as particle growth by aqueous or heterogeneous reactions [39].
4.2.1. Mass Concentrations
Mass concentrations of particles are essential for tracking conventional air quality. Importantly, for aerosols, they provide a means to test composition measurements of particles to achieve a material balance of chemical constituents. The measurement of mass concentration is measured in terms of (fine) particles less than 2.5 µm diameter vs. fine plus coarse particles less than 10 µm in diameter. The data in Figure 8 show an annual and seasonal character of PMx mass concentrations in Lagos.
The annual mean 24 h PM2.5 concentrations ranged from 40 (NCF) to 135 (IKO) µg m−3. The annual 24 h mean PM10 concentrations ranged from to 87 (NCF) to 189 (IKO) µg m−3. The range of PM2.5/PM10 mass ratio was 0.46 (NCF) to 0.8 (IKO). Seasonally, both PM measures tended to peak in the dry season, suggesting either a rise in local urban dust or a contribution from regional dust sources associated with the Harmattan winds.
One review of particulate matter in several Nigerian locations between 1985 and 2015 [35] indicated, for Lagos, a mean of 56.5 µg m−3 for PM2.5. The mass concentrations reported by other investigators included those of Owoade et al. [38]. Nine months of 12 h sampling in 2010 at four sites similar to those of the Owoade et al. study yielded mean values of PM2.5 mass concentration ranging from 28.0 (residential) to 30.9 µg m−3 (heavy traffic). Their values for PM2.5–10 ranged from 26.1 (marine) to 64.5 µg m−3 (residential). The mean PM2.5/PM10 ratio was determined to be 0.41 ± 0.15. In 2017, Obanya et al. [22] found mean values for PM2.5 of 34.9 ± 12.1 µg m−3 at residential sites and transport-related concentrations of 69.6 ± 335 µg m−3. For PM10, they reported 73.4 ± 27.5 µg m−3 at residential sites and transport-related values of 144 ± 76.6 µg m−3. These values give mass concentration ratios of 0.467 for residential and 0.495 for transport.
A local influence is observed seasonally at IKO. The PM2.5 concentration at this location is systematically higher than at other sites year-round, particularly during the dry season. For PM10, the IKO elevated mass concentration was seen in the monthly data, except for the dry period, December-February. During the dry period, PM10 was much higher at all sites than at other times, even at NCF. The results also showed that the near-shore location (NCF) was low compared with the inland sites, except for the PM10 in the dry season.
4.2.2. Particle Composition
The species compositions for aerosol particles are the principal chemical characteristics of interest. In the following section, the inorganic and organic speciation from our study is described and interpreted in light of PM2.5 particle sources. Corresponding PM10 data are included in the Supplementary Materials.
Figure 9 presents the PM2.5 abundance of elements, ions, elemental carbon (EC), and organic carbon (OC) across the six sites. These species make up the bulk of the composition of the gravimetric mass of PM2.5 and PM10. A comparison of Figure 9 indicates that Na, Si, SO42−, NO3−, NH4+, OC, and EC were the most abundant quantified species found in PM2.5. These species were also the most abundant in PM10, along with Al (See Figure S8), across the sites. Two major “quasi-natural” particle sources common to many locations are soil dust and sea salt (near marine areas). The abundances shown in Figure 9 indicate that there was a surplus of Na, accounting for sea salt; the surplus is probably mainly associated with soil dust in the city as sodium aluminum silicates. Relative to other sites, IKO PM2.5 tended to have higher concentrations of all species. This is associated with the presence of industry and traffic locally. Of the higher atomic weight metals, Fe and Zn tend to dominate, especially in IKO where two steel plants are located (e.g., Figure S7).
The PM2.5 and PM10 compositions were evaluated relative to the measured gravimetric mass by converting individual species into materials that estimated their chemical composition in the air. For example, including oxygen with trace elements and oxygen, nitrogen, and other elements with organic carbon. Here, the mass balance was determined using a chemical mass balance approach, specifically, the US Interagency Monitoring of Protected Visual Environments (IMPROVE) model [40]. The percentages of the gravimetrically measured PM2.5 mass as ammonium nitrate, ammonium sulfate, sea salt (NaCl), trace elements, elemental carbon, and organic material (assumed OM ~1.6 × OC [41]) are included. The identified components of PM2.5 for each of the sites are shown in Figure 10.
An estimated distribution of major sources to PM2.5 is shown in Figure 10 for our Lagos sites. The primary particle attribution indicates that the combination of soil dust and carbonaceous material accounted for the majority of PM2.5 mass, as was the case for PM10. The OC-EC component at the sites was large, representing approximately 30% of the PM2.5. In PM2.5, SO42− and NO3− were minor contributors to fine particle mass, unlike most cities in northern mid-latitude locations.
For particles with primary sources, most sites had a large percentage of Si and Al identified with fine soil dust, likely due to the re-entrainment of dust from Lagos streets and walkways and the impact of Harmattan dust. For individual sites, ABE was dominated by fine soil, organic matter, and elemental carbon, which comprised approximately 74% of the PM2.5 mass. At IKO, fine soil, trace elements, and organic matter comprised 66% of the PM2.5 mass, while at JAN approximately 72% of PM2.5 was composed of organic material, fine soil, and elemental carbon. Part of OM was identified with the oxidation of VOCs. A large fraction of the PM2.5 mass, equal to about 74%, 75%, and 65%, was composed of fine soil, EC, and organic material at EPA, NCF, and UNI, respectively. IKO had a relatively high abundance of trace elements, especially Pb, Zn, Fe, and Cl, which were associated with local industry within a few km of the site (see Figure S8). A comparison of the sea salt in NCF (an onshore site) and IKO (an inland site) indicates that sea salt as NaCl at NCF was nearly half that of IKO.
A study of Leck et al. [42] adds some insight into the background of fine particles using data for <PM1 obtained from ship data offshore of the coast of the Gulf of Guinea during the dry period. Data from this 1993 expedition are listed in Table S4. The species listed in the table indicate that between non-sea salt (nss) SO42− (identified with fuel combustion) and methane sulfonic acid (from oxidation of marine dimethyl sulfide), a significant fraction of the SO42− equivalent can be accounted for by marine air. The Leck et al. data also suggest from Si and EC concentrations that dust and carbon from inland biomass burning are present in a regional background aerosol. Particulate vanadium in this Gulf region was not detected, indicating, by this tracer, that residual oil combustion was regionally unimportant.
Assuming that the offshore observations of Leck et al. represent marine conditions over the Gulf of Guinea, we note that, like O3, a regional influence of aerosol dust is present, at least during the Harmattan period [43,44].
Secondary particles are identified with SO42− and NO3− as NH4+ salts and a part of the OC. The low average concentrations of SO2 found in Lagos (<20 µg m−3) and the uniform levels of SO42− at ~2 µg m−3 do not support a SO42− spatial gradient that would form from a local rapid oxidation source for SO42− production. A more likely interpretation for SO42− is a ubiquitous regional presence ranging up to 5 µg m−3 in addition to 1.7 µg m−3 SO42− in local sources, as suggested by the chemical mass balance [45]. The regional SO42− contribution is stated in reviews of PMx ([35,37]) and multiscale reviews of SO42− in the mid-latitude conditions [46]. Regional sources are diverse but probably include a range of petroleum refining and combustion, agricultural sulfur, and marine or mineral dust. Examples of the marine component offshore are shown in Table S4. A non-sea salt SO42− contribution from the onset of sea breeze could be as high as 0.5 µg m−3. However, if methyl sulfonic acid (MSA) in aerosol is oxidized to SO42−, its contribution could be as high as half of the Lagos PM2.5 concentration.
The NO3− concentrations were similar to SO42−. This result is unusual for warm tropical conditions since the equilibrium chemistry for ammonium nitrate formation is preferential in cold air because of the volatility of the salt. The filter sampling approach of this study suggests that nitrate concentrations are likely to reflect the capture of HNO3 vapor on the filter substrate as well as particulate nitrate [47,48] and the loss of NO3− from the Teflon filter by volatility. The latter favors an underestimate of NO3−-, but it adds uncertainty to the reported value.
The fraction of OC that is secondarily derived from the photochemically related oxidation depends on the VOC carbon number. From atmospheric observations and smog chamber experiments, this nominal constraint is ~>C5 [39]. An estimation of the secondary contribution (SOC) can be made from the OC/EC ratio if this ratio is known for primary sources [49]. The OC/EC fraction in this study averaged 1.9 with a range of 1.6–2.1. For many cities, the principal primary sources of carbon are fossil or biomass fuel combustion. In the US, for example, up to the early 2000s, the fresh domestic internal combustion and diesel vehicles had an OC/EC ratio of ~1 [50] (the ratio relevant to vehicle emissions in Lagos is unknown). In contrast, biomass combustion is highly variable, depending on the biomass resource and the burning conditions. The ratio can vary from ~1 to >5, e.g., [51,52]. A ratio of <~1.9 could be interpreted as the primary aerosol associated with motor vehicle and biomass burning emissions alone. Based on experience in US cities, an OC/EC ratio of <2 is considered fresh primary carbon, while >2 is identified with “aged” conditions with increased secondary OC present superimposed on oxidized OC from primary sources [53].
There have been a number of other studies reporting elemental analyses of filter samples for PM2.5 and PM10 [35,37]. Investigators have applied receptor modeling, either chemical mass balance or positive matrix factorization, to these data and have found essentially the same particle sources. These include soil dust, marine- and traffic-related sources, biomass burning, and industrial sectors.
4.2.3. OM Speciation
A GC-MS analysis of the organic components of particulate matter was undertaken using the quartz fiber filter samples from the study. The average concentration profiles by site for the PM2.5 samples are given in Figure 11. The results for PM2.5 show a major concentration grouping of alkanes and relatively high-molecular-weight alkyl-chain compounds along with a variety of other species. The PM10 samples showed a similar OC speciation pattern, as noted in Figure S9. Those classes of compounds are listed below in Table 4 along with the probable source identification for primary particle emissions.
Some of the carbon species were detected with carbon numbers >20. Biogenic emissions tend to produce waxes with more odd-numbered species than even. Figure S10 indicates the carbon number associated with the species found in the GC-MS analysis. The prevalence of odd carbon numbers is apparent, with a much stronger signal in the dry season than in the wet season. This is interpreted as the importance of biogenic species dominance of high-molecular-weight organics in the aerosol. These species could come from direct emissions during a warmer portion of the year, or they could be the result of wildfire emissions in the interior of West Africa during the dry season.
Table 4Categories of organic species observed by GC-MS analysis in Lagos PM2.5 samples.
| Class of Compound | No. of Species | Range (ng m−3); Site Order | Source/Indicator | References for Source Identification |
|---|---|---|---|---|
| Polycyclic aromatic hydrocarbons (PAHs) | 10 | 15 (NCF)-33(IKO) | Carbon fuel combustion | [54] |
| n-Alkanes | 23 | 23(NCF)-159 (EPA) | Petroleum; vegetation | [54,55,56,57] |
| Iso/Anteiso-alkanes | 10 | 1.3(NCF)-13(JAN) | Waxes; leaves | [58] |
| Hopanes | 18 | 2(NCF)-19(EPA) | Petrol prod., lub. oil; motor vehicle exhaust | [59,60] |
| Steranes | D13 | 4(NCF)-28(EPA) | Motor vehicle exhaust | [59,60] |
| Branched alkanes | 2 | 2–100 | Diesel fuel; lub. oil | [57,60] |
| Alkyl cyclohexanes | 5 | <0.1–20 | Diesel fuel; lub. oil | [57,60] |
| Alkenes | 2 | 0.1–2 | Squalene a; octadecene b identified | [61] |
| Monocarboxylic acids | 1 | (UNI)-(IKO) | Nonanoic acid; cooking; fuel combustion; biogenic | [55,62,63,64] |
| Alkenoic acids | 4 | (NCF)-(IKO) | Oleic acid; biomass burning | [64] |
| Resin acids | 4 | (NCF)-(IKO) | Biomass burning | [64] |
| Dicarboxylic acids | 2 | (NCF)-(IKO) | Vehicle exhaust, Biomass burning; Meat cooking. Oxalic acid possible marker for SOC | [60,64,65,66] |
| n-Alkanols | 18 | (NCF)-(IKO) | 1-octa;1-tetra;1-hexa-cosanol; charbroiling | [66,67] |
| Anhydro-sugars | 1 | 6(NCF)-80(IKO) | Levoglucosan; biomass burning | [68] |
| Biomarkers | 11 | Stigmasterol—biomass burning; char-broiling | [69,70] |
a Biogenic triterpene (C30H50) sterol intermediate from flora and fauna. b Suspected source pyrolysis of plastic wastes.
Essentially all of the high-molecular-weight classes of OC can be interpreted in terms of primary sources. Nevertheless, there is the potential for SOC formation, not only from isoprene and substituted aromatic compounds but also from intermediate volatility alkanes (e.g., [71,72]). From the evidence from the OC/EC ratios and the ambiguous speciation from the GC-MS analysis, we infer that secondary OC was minimally present in the Lagos samples.
4.3. Uncertainties and Limitations
There are important limitations in field studies that involve representative site selection, instrument or measurement method selection, sampling frequency and duration, and the quality control or assurance practice. The details addressing these aspects of the study are included in [7]. In particular, the procedures adopted for documenting the continuous and laboratory measurements were reported in this reference. The results of the study are constrained by the six urban sites and use of the compact multicomponent Zephyr units. Field comparisons using the Aeroqual AQM 65 instrument as a secondary standard were in used to constrain measurement uncertainties. Recording meteorological observations was performed with conventional units. The laboratory analyses relied on conventional sampling using filters and canisters.
The uncertainty levels for laboratory-related samples (VOC speciation) and PM chemistry have been reviewed extensively, for example, in Fehsenfeld et al. [73]. Their review also included comments on established uncertainties in meteorological observations. The discussion and precision and accuracy results are summarized in Fehsenfeld et al. Tables 5.2 and 5.3 and are assumed to apply to this study. Their summary indicates that accuracies range from ±5 to 10%, with a precision of ±10–30%, depending on the specific VOC analyte. Measurements of organic species in aerosols are less well established or perhaps rigorously unknown [73]; uncertainties depend on species and methods used in GC/MS analysis, e.g., [74]. For specific compounds, intercomparison studies have been attempted [75], and reference material has been proposed [76]. A typical generic value for organic species is estimated as ±20% accuracy [77]. We note that the organic component is highly complex, and typically about 25% or less up to 50% of the total in atmospheric aerosols have been identified. Furthermore, additional uncertainties derive from sampling and the extraction of samples, including the issue of gas-particle partitioning and atmospheric oxidation. Major recent improvements [78] in instrumentation are facilitating substantial advancements in the identification and quantification of organic species. One example of an approach to the quantitation of multiple condensed and vapor phase species derives from studies in Baltimore, Maryland [79].
The gas observations using the inexpensive, multi-species Zephyr and Aeroqual AQM 65 (with internal calibration) do not have the credibility that supports the reference instrument historically cited, for example, in [74]. The application of these electrochemical-based instruments to long-term measurements is limited but illustrates the potential (Table S1) for these units to be used for ambient air measurements [80]. One example of thorough laboratory and CO and NO2 detector calibration and performance was reported [81]; another is discussed by Mead et al. [82]. Relevant to this study, The USEPA Community Air Sensor Project (CAIRSENSE) [83] compared, for an 8-month period, several electrochemical-based instruments with each other and federal reference instruments established by independent studies. The test included an Aeroqual O3 detector. An unpublished Aeroqual AQM 65 comparison with an EPA monitor was also reported [84]. Their results indicated reliable quantitative gas measurements over the periods studied. In our Lagos study, the Zephyr gas measurements were collocated with the Aeroqual AQM 65 instrument over six months to assess the field-based precision and accuracy estimates and establish a species concentration re-calibration to account for bias and gain. The methods used to follow this procedure are described in the EnvironQuest final report [7].
The Zephyr and Aeroqual instruments are designed for air pollution monitoring of relatively high concentrations, especially those near or exceeding health-related thresholds [78]. Their use in research studies such as our Lagos study is limited in this respect. In the Lagos study, we were challenged, particularly with the interpretation of data near or below the detection limits of the Zephyr unit. For the purposes of investigating the character of the diel, weekly, and monthly averages for gas concentrations, we assumed that the linear relationship with the gas concentration applied below the detection limit. Following the Analytical Methods Committee guidelines [85], we used such low concentration observations to at least semi-quantitatively complete our knowledge of the time series averages and avoid excessive biasing of the summary statistics through left censoring. We note that, on average, conditions near or below the instrument detection limits should be considered to have unknown accuracy.
The Zephyr and Aeroqual AQM 65 also measure PMx by light scattering or the ARA PM sampler. We did not examine these data for this paper. A PMx comparison with gravimetric results is planned for a separate analysis. Some performance observations were obtained in the CAIRSENSE study [83].
5. Discussion and Conclusions
The 2020–2021 campaign characterized air-pollution-related chemistry measurements for gases and airborne particles at six sites in the Lagos metropolitan area. The observations enhanced the understanding of results from earlier studies by showing the presence of urban oxidant chemistry consistent with known precursor–ozone product relationships. Paralleling the gas-phase processes was the urban chemistry of aerosols containing both particles from local primary sources and secondary particles produced in the atmosphere.
The continuous measurements were obtained with inexpensive compact instruments at six geographically distinct surface sites. These were complemented with conventional filter and canister sampling for fine and fine plus coarse particle chemistry and VOC species. The sites also included surface meteorological observations.
The interpretation of the measurements for air chemistry was constrained by the available spatially and temporally limited gas and particle emissions data and meteorological data.
Some important conclusions are:
Unlike in the mid-latitudes, gas and particle chemistry increases with the onset of the dry season during the mid-latitude winter months; chemical processes decrease during the summer wet months.
The diel variation in NO and NO2 concentrations are maximal before the midday ozone maxima, suggesting a pattern dominated by local processes; a well-known effect is seen with ozone increasing on the weekends with decreased NO emissions from reduced human activity, likely reduced traffic.
VOC precursor species are observed to reflect fossil fuel and biomass components typical of large cities with heavy vehicle traffic supplemented with natural emissions and industrial or domestic sources.
Fine particles and fine plus coarse particle mass concentrations are seasonally variable with maxima in the dry season.
The particle composition contains a variety of urban-related species dominated by elemental and organic carbon with sulfate produced from SO2 oxidation in the air. Particles also contain significant amounts of nitrate, which may be underestimated from the sampling method used.
The organic carbon fraction can be attributed to primary sources; there was no direct evidence of significant amounts of organic carbon produced by the oxidation of VOCs.
The exploration of the organic species in particles indicates the presence of high-molecular-weight material from anthropogenic sources supplemented by biomass contributions from open burning and waxy material from vegetation.
There is indirect evidence of the occasional influence of regional air mass transport of ozone and dust, especially during the Harmattan wind period in the dry season.
A.O.-O., project principal investigator; P.H. (Pierre Herckes) and M.F., analytical chemistry; P.H. (Philip Hopke), data analysis; J.O., meteorological assessment; P.A.S., quality assurance and coordination; O.P., critical review of operations and analysis; G.M.H., data interpretation and manuscript preparation. All authors have read and agreed to the published version of the manuscript.
The data and final report [
This project was sponsored partly by the World Bank as part of its environmental assessment activities.
The authors have no conflict of interest in the project.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Enlarge this image.Figure 1. Map of Lagos metropolitan area with monitoring station locations (airport shown as reference location). The central body of water is the Lagos lagoon. The entry from the Gulf of Guinea is located at the bottom center of the map. The seaport and lagoon entry are bottom center. A map of the Niger Delta region with Lagos is included in Figure S1.
Enlarge this image.Figure 2. Surface (10 m height) wind roses for the Lagos monitoring sites. Color indicates wind speed range in m/s).
Enlarge this image.Figure 3. Example of CO variation for near Gulf-of-Guinea site (NCF) and inland site (EPA), including diel average in local time (left), monthly average (center), and weekday average (right). Graphs (a) are for NCF and (b) for EPA sites. The vertical bands show variability (pink) ±in values, e.g., 5–95% range. Graphs for the remaining sites are in the Supplementary Materials, Figures S2–S5.
Enlarge this image.Figure 4. Example average concentrations at NCF and EPA for NO and NO2 for the duration of the study. Graphs in (a) are NCF and (b) are EPA sites. The remaining site data are located in Figures S2–S5.
Enlarge this image.Figure 5. Box and whisker mixing ratio plots of canister sample analysis for VOC species between 2020 and 2021.
Enlarge this image.Figure 6. Annual average diel O3 concentrations for sites NCF (a) and EPA (b). Ozone concentrations for the other sites are in Figures S2–S5.
Enlarge this image.Figure 8. Seasonal variation in particle mass concentration for 24 h samples of fine ≤ 2.5 micrometer diameter (PM2.5) and fine + coarse particles ≤ 10 micrometer diameter (PM10).
Enlarge this image.Figure 9. Average plots of abundance of various PM2.5 species in 24 h filter samples from this study. (Top) Selected primary ionic (ion chromatograph assayed) species and secondary species with relatively high concentrations. (Bottom) Abundance of elements found in samples from X-ray fluorescence analysis.
Enlarge this image.Figure 10. Pie charts for PM2.5 showing percentage of components by apparent source to total mass concentration from CMB analysis.
Enlarge this image.Figure 11. Organic species in PM2.5 quartz fiber filter samples identified in canister samples by GC-MS. Top panel includes hydrocarbon species. Bottom panel includes oxygenates. OM speciation for PM10 is included in Figure S9.
Monitoring station locations and characteristics.
| No. | Key: Name and District | Latitude | Longitude | Site Characteristics |
|---|---|---|---|---|
| 1. | University of Lagos (UNI), Akoka | 6.515500° N | 3.390450° E | Urban scale: Business/residential/institutional |
| 2. | Nigerian Conservation Foundation (NCF), Lekki | 6.432200° N | 3.535917° E | Regional scale: low-density residential; near coast |
| 3. | Abesan Estate (ABE), Ipaia | 6.609062° N | 3.267706° E | Neighborhood scale: High-density residential, school environment |
| 4. | King Ado School (JAN), Jankara | 6.459800° N | 3.391100° E | Commercial scale near seaport: High vehicular traffic and high-density residential |
| 5. | Community Secondary School (IKO), Ikorodu | 6.672360° N | 3.534030° E | Neighborhood scale: Industrial low-density residential |
| 6. | Lagos State EPA (EPA), Alausa | 6.612873° N | 3.360596° E | Urban scale: Institutional, business district with moderate traffic |
Summary of diel changes in local Lagos meteorology.
| Location | Air Temperature (°C) | Relative Humidity (%) | Wind Speed (ms−1) | |||
|---|---|---|---|---|---|---|
| Range | Mean | Range | Mean | Range | Mean | |
| ABE | 22.9–29.6 | 26.9 | 51.0–95.0 | 87.7 | 0.0–1.8 | 0.8 |
| IKO | 21.9–31.9 | 26.7 | 56.0–96.0 | 87.6 | 0.3–3.7 | 1.7 |
| JAN | 23.4–31.4 | 27.9 | 49.0–91.0 | 83.6 | 0.1–3.4 | 1.2 |
| EPA | 22.8–35.0 | 27.5 | 50.0–95.0 | 84.8 | 0.0–3.4 | 1.4 |
| NCF | 23.4–29.8 | 27.3 | 69.0–96.0 | 88.3 | 0.1–3.9 | 1.7 |
| UNI | 23.1–30.0 | 27.1 | 54.0–94.0 | 86.6 | 0.1–4.1 | 1.5 |
Supplementary Materials
The following supporting information can be downloaded at:
References
1. Azeez, L.; Oyedeji, A.; Adewuyl, S.; Tijni, K.; Adebisi, S.; Olanaogun, N. Precursors Influencing Tropospheric Ozone Formation and Apportionment in Three Districts of Ilupeju Industrial Estate. Am. J. Chem.; 2016; 6, pp. 65-73.
2. Raheem, A.; Adekola, F.; Obioh, L. The Seasonal Variation of the Concentration of Ozone, Sulfur Dioxide, and Nitrogen Oxides in Two Nigerian Cities. Environ. Model Assess.; 2008; 14, pp. 497-509. [DOI: https://dx.doi.org/10.1007/s10666-008-9142-x]
3. Kemper, K.; Chaudhui, S. Air Pollution: A Silent Killer; The World Bank: Washington, DC, USA, 2022.
4. Tawari, C.; Abowei, J. Air Pollution in the Niger Delta Area of Nigeria. Int. J. Fish. Aquat. Sci.; 2012; 1, pp. 94-117.
5. Prospero, J. Long term measurements of transport of African mineral dust to the southeastern United States: Implications for regional air quality. J. Geophys. Res.; 1999; 104, pp. 15917-15927. [DOI: https://dx.doi.org/10.1029/1999JD900072]
6. Ansmann, A.; Baars, H.; Tesche, M.; Müller, D.; Althausen, D.; Engelmann, R.; Pauliquevis, T.; Artaxo, P. Dust and smoke transport from Africa to South America: Lidar profiling over Cape Verde and the Amazon rain forest. Geophys. Res. Lett.; 2009; 36, L111802. [DOI: https://dx.doi.org/10.1029/2009GL037923]
7. EnvironQuest Ltd. Lagos Air Quality and PM Source Apportionment Study. Final 12 Month Summary Report; EnvironQuest: Lagos, Nigeria, 2022.
8. EarthSense. Zephyr Air Quality Sensor (Specifications). 2021; Available online: www.earthsens.cp.uk (accessed on 29 May 2022).
9. Aeroqual. AQM 65. Available online: www.aeroqual.com/prioducts/aqm-stations/aqm-65-air-quality-monitoring-station (accessed on 20 February 2022).
10. Sunset Laboratory, Inc. Cutting Edge Carbon Aerosol Particulate Analysis Instrument. Available online: www.sunla.com (accessed on 20 February 2022).
11. Fakinie, B.; Odekanie, E.; Olalekan, A.; Ije, H.; Oke, D.; Sonibare, J. Air pollutant emissions by anthropogenic combustion processes in Lagos, Nigeria. Cogent Eng.; 2020; 7, 1808285. [DOI: https://dx.doi.org/10.1080/23311916.2020.1808285]
12. Olatunji, S.; Fakanie, B.; Jimoda, L.; Adeniran, J.; Adesanmi, J. Air Emissions of Sulfur Dioxide from Gasoline and Diesel Consumption in the Southwestern States of Nigeria. Pet. Sci. Technol.; 2015; 33, pp. 678-685. [DOI: https://dx.doi.org/10.1080/10916466.2014.1002929]
13. Jimoda, L.; Sonivare, J.; Akeredolu, F.A. Emission Inventory of Anthropogenic Activities at Southeastern Lagos. Ife J. Technol.; 2009; 18, pp. 35-541.
14. Oketola, A.; Osibanjo, O. Assessment of Industrial Pollution Load in Lagos, Nigeria by Industrial Pollution Projection System (IPPS) versus Effluent Analysis; Bronicewcz, E. Intech Open Book Series; Chap/10 Available online: www.Intechopen.com/books/213 (accessed on 10 February 2022).
15. Marais, E.; Jacob, D.; Wecht, K.; Lerot, C.; Zhang, L.; Yu, K.; Kurosu, T.P.; Chance, K.; Sauvage, B. Anthropogenic emissions in Nigeria and implications for atmospheric ozone pollution: A view from space. Atmos. Environ.; 2014; 99, pp. 22-40. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2014.09.055]
16. Amadiou, C.; Omotosho, B.; Coulialy, A.; Ballo, A. Characteristics of land and sea breezes along the Guinea Coast of West Africa. Theor. Appl. Climatol.; 2019; 188, pp. 953-971.
17. Gbambie, A.; Steyn, G. Sea breezes at Cotonou and their interaction with the West African Monsoon. Intl. J. Clim.; 2012; 33, pp. 2889-2899. [DOI: https://dx.doi.org/10.1002/joc.3637]
18. Climate Data for Lagos, Nigeria. Available online: www.lagos.climatemps.com/sunlight.php (accessed on 20 February 2022).
19. US Environmental Protection Agency. Carbon Monoxide Trends. Available online: www.epa.gov/air-trends/carbon-monoxide-trends (accessed on 11 February 2022).
20. Njoku, K.; Rumide, T.; Akinola, M.; Adesuyi, A.; Jolaoso, A. Ambient Air Quality Monitoring in Metropolitan City of Lagos, Nigeria. J. Appl. Sci. Environ. Manag.; 2016; 20, pp. 178-185. [DOI: https://dx.doi.org/10.4314/jasem.v20i1.21]
21. Olajire, A.; Azeez, L.; Oluyemi, E. Exposure to hazardous air pollutants along Oba Akran Road, Lagos-Nigeria. Chemosphere; 2011; 84, pp. 1044-1051. [DOI: https://dx.doi.org/10.1016/j.chemosphere.2011.04.074]
22. Obanya, H.; Ameeze, N.; Otitoloju, A. Air Pollution Monitoring Around Residential and Transportation Sector Locations in Lagos, Mainland. J. Health Poll.; 2018; 8, 9180903. [DOI: https://dx.doi.org/10.5696/2156-9614-8.19.180903]
23. Abulude, F.; Damodharen, U.; Acha, S.; Adamu, A.; Arifalo, K. Preliminary Assessment of Air Pollution Quality Levels of Lagos. Aerosol Sci. Eng.; 2021; 5, pp. 275-284. [DOI: https://dx.doi.org/10.1007/s41810-021-00099-1]
24. George, L. Nigeria to Cut Sulfur in Fuels a Year after U.N. Deadline. 2018; Available online: www.reuters.com/article/us-nigeria-idUSKNGP1HQ (accessed on 15 February 2022).
25. Ogundairo, F. Nigeria: Petrol, Diesel Imported from Europe Dirtier Than Fuel from Illegal Refineries—Report. 2020; Available online: http://allafrica.com/stories/202007030199.html (accessed on 20 February 2022).
26. Yao, X.; Lau, N.; Chan, C.; Fang, M. The use of tunnel concentration profile data to determine the ratio of NO2/NOx directly emitted from vehicles. Atmos. Chem. Phys. Discuss.; 2005; 5, pp. 12373-12740.
27. Sillman, S.; He, D.; Cardelino, C.; Imhoff, R.E. The use of photochemical indicators to evaluate ozone-NOx-hydrocarbon sensitivity: Case studies from Atlanta, New York, and Los Angeles. J. Air Waste Manag. Assoc.; 1997; 47, pp. 1030-1040. [DOI: https://dx.doi.org/10.1080/10962247.1997.11877500]
28. Blanchard, C.; Hidy, G.; Tanenbaum, S.; Rasmussen, R.; Watkins, R.; Edgerton, E. NMOC, ozone, and organic aerosol in the southeastern United States 1999–2007: 1. Spatial and temporal variations of NMOC concentrations and composition in Atlanta, Georgia. Atmos. Environ.; 2010; 44, pp. 4840-4849. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2010.07.030]
29. Olajire, A.; Azeez, L. Source apportionment and ozone formation potential of volatile organic compounds in Lagos, Nigeria. Chem. Ecol.; 2014; 30, pp. 156-168. [DOI: https://dx.doi.org/10.1080/02757540.2013.844796]
30. Drozd, G.T.; Zhao, Y.; Saliba, G.; Frodin, B.; Maddox, C.; Weber, R.J.; Chang, M.-C.O.; Maldonado, H.; Sardar, S.; Robinson, A.L. et al. Time resolved measurements of speciated tail pipe emissions from motor vehicles: Trends with emission control technology, cold start effects, speciation. Environ. Sci. Technol.; 2016; 50, pp. 13592-13599. [DOI: https://dx.doi.org/10.1021/acs.est.6b04513]
31. Hidy, G.M.; Blanchard, C.L.; Baumann, K.; Edgerton, E.; Tanenbaum, S.; Shaw, S.; Knipping, E.; Tombach, I.; Jansen, J.; Walters, J. Chemical climatology of the southeastern United States, 1999–2013. Atmos. Chem. Phys.; 2014; 14, pp. 11823-11914. [DOI: https://dx.doi.org/10.5194/acp-14-11893-2014]
32. Sauvage, B.; Gheusi, F.; Thouret, V.; Cammas, J.-P.; Duron, J.; Escobar, J.; Mari, C.; Mascart, P.; Pont, V. Medium-range mid-tropospheric transport of ozone and precursors over Africa: Two numerical case studies in dry and wet seasons. Atmos. Chem. Phys.; 2007; 7, pp. 53757-53770. [DOI: https://dx.doi.org/10.5194/acp-7-5357-2007]
33. Minga, A.; Thouet, V.; Saunola, M.; Dlon, C.; Serca, D.; Mari, C.; Sauvage, B.; Mariscal, A.; Leriche, M.; Cros, B. What caused extreme ozone concentrations over Coutonou in December 2005?. Atmos. Chem. Phys.; 2010; 10, pp. 895-907. [DOI: https://dx.doi.org/10.5194/acp-10-895-2010]
34. Menut, L.; Flamant, C.; Turquety, S.; Derioubaix, A.; Chazette, P.; Meynadier, R. Impact of biomass burning on pollutant surface concentrations in mega-cities of the Gulf of Guinea. Atmos. Chem. Phys.; 2018; 18, pp. 2687-2707. [DOI: https://dx.doi.org/10.5194/acp-18-2687-2018]
35. Offor, H.; Adie, G.; Ana, G. Review of Particulate Matter and Elemental Composition of Aerosols at Selected Locations in Nigeria from 1985–2015. J. Health Pollut.; 2016; 6, pp. 1-18. [DOI: https://dx.doi.org/10.5696/2156-9614-6-10.1]
36. Ezeh, G.; Obioh, I.; Asubiojo, O. Trace Matals and Source Identification of Air-Borne Particulate Matter Pollution in a Nigerian Megacity. J. Environ. Toxicol.; 2017; 7, 3. [DOI: https://dx.doi.org/10.4172/2161-0525.1000463]
37. Taiwo, A.; Arowolo, T.; Abdullah, K.; Taiwo, O. Particulate Matter Pollution in Nigeria. Proceedings of the 14th Conference on the Environmental Influence of Atmospheric Science and Technology; Rhodes, Greece, 3–5 September 2015; Available online: https://cest2015.gnest.org/papers/cest2015_poster_paper.pdf (accessed on 25 January 2022).
38. Owoade, O.K.; Fawole, O.G.; Olise, F.S.; Ogundele, L.T.; Olaniyi, H.B.; Almeida, M.S.; Ho, M.D.; Hopke, P.K. Characterization and source identification of airborne particulate loadings at receptor site-classes of Lagos Mega-City, Nigeria. J. Air Waste Manag. Assoc.; 2013; 63, pp. 1026-1035. [DOI: https://dx.doi.org/10.1080/10962247.2013.793627]
39. Seinfeld, J.H.; Pandis, S.N. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; John Wiley & Sons: New York, NY, USA, 2006.
40. Interagency Monitoring of Protected Visual Environments (IMPROVE). Available online: http://vista.cira.colostate.edu/Improve/ (accessed on 10 February 2022).
41. Landis, M.; Lewis, C.; Stevens, R.; Keeler, R.; Dvonch, J.; Trembly, R.T. Ft. McHenry tunnel study: Source profiles and mercury emissions from diesel and gasoline powered vehicles. Atmos. Environ.; 2007; 41, pp. 8711-8724. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2007.07.028]
42. Leck, C.; Engardt, M.; Heintzenburg, J. A meridional profile of the chemical composition of submicrometre particles over the East Atlantic Ocean: Regional and hemispheric variabilities. Tellus; 2002; 54B, pp. 377-394. [DOI: https://dx.doi.org/10.3402/tellusb.v54i4.16672]
43. Oluleye, A.; Jimoh, O. Influence of atmospheric circulation patterns on dust transport during the Harmattan Period in West Africa. Pollution; 2018; 4, pp. 9-27.
44. Chukwuemeka, M.; Adenola, E. Variability of Harmattan Dust Haze Over Northern Nigeria. J. Pollut.; 2018; 1, 1000107.
45. Oluyemi, E.; Asubojo, O. Ambient Air Particulate Matter in Lagos, Nigeria: A Study Using Receptor Modeling with X-ray Fluorescent Analysis. Bull. Chem. Soc. Ethiop.; 2001; 15, pp. 97-108.
46. Hidy, G.M. Worldwide aerosol chemistry: From hemispheric distributions to megacity sources. J. Air Waste Manag. Assoc.; 2009; 59, pp. 770-789. [DOI: https://dx.doi.org/10.3155/1047-3289.59.7.770]
47. Shaw, R., Jr.; Stevens, R.; Bowermaster, J.; Tesch, J.; Tew, E. Measurements of atmospheric nitrate and nitric acid: The Denuder Difference Experiment. Atmos. Environ.; 1982; 16, pp. 845-855. [DOI: https://dx.doi.org/10.1016/0004-6981(82)90403-6]
48. Ashbaugh, L.; Eldred, R. Loss of Particle Nitrate from Teflon Sampling filters. Effects on Measured Gravimetric mass in California and in the IMPROVE network. J. Air Waste Manag. Assoc.; 2002; 54, pp. 93-104. [DOI: https://dx.doi.org/10.1080/10473289.2004.10470878]
49. Lim, H.; Turpin, B. Origins of primary and secondary organic aerosol in Atlanta: Results of time-resolved measurements during the Atlanta supersite experiment. Environ. Sci. Technol.; 2002; 36, pp. 4489-4490. [DOI: https://dx.doi.org/10.1021/es0206487]
50. Chow, J.C.; Watson, J.G.; Kuhns, H.; Etyemezian, V.; Lowenthal, D.H.; Crow, D.; Kohl, S.D.; Engelbrecht, J.P.; Green, M.C. Source profiles for industrial, mobile, and additional sources in Big Bend Regional Visibility and Observation Study. Chemosphere; 2014; 554, pp. 185-208.
51. Atiku, F.; Mitchell, E.; Le-Langton, A.; Jones, J.; Williams, A.; Bartle, K. The Impact of Fuel Properties on the Composition of Soot Produced by the Combustion of Residential Solid Fuels in a Domestic Stove. Fuel Process. Technol.; 2016; 151, pp. 117-125. [DOI: https://dx.doi.org/10.1016/j.fuproc.2016.05.032]
52. Sirignano, C.; Riccio, A.; Chianese, E.; Ni, H.; Zenker, K.; D’Onofrio, A.; Meijer, H.A.J.; Dusek, U. High Contribution of Biomass Combustion to PM2.5 in the City Centre of Naples (Italy). Atmosphere; 2019; 10, 451. [DOI: https://dx.doi.org/10.3390/atmos10080451]
53. Robinson, A.L.; Donahue, N.M.; Shrivastava, M.K.; Weitkamp, E.A.; Sage, A.M.; Grieshop, A.P.; Lane, T.E.; Pierce, J.R.; Pandis, S.N. Rethinking organic aerosols: Semi-volatile emissions and photochemical aging. Science; 2007; 315, pp. 1259-1262. [DOI: https://dx.doi.org/10.1126/science.1133061]
54. Wang, M.; Huang, R.-J.; Cao, J.; Dai, W.; Zhou, J.; Lin, C.; Ni, H.; Duan, J.; Wang, T.; Chen, Y. et al. Determination of n-alkanes, polycyclic aromatic hydrocarbons and hopanes in atmospheric aerosol: Evaluation and comparison of thermal desorption GC-MS approaches. Atmos. Meas. Tech.; 2019; 12, pp. 4779-4789. [DOI: https://dx.doi.org/10.5194/amt-12-4779-2019]
55. Brown, S.G.; Herckes, P.; Ashbaugh, L.; Hannigan, M.P.; Kreidenweis, S.M.; Collett, J.L., Jr. Characterization of organic aerosol in Big Bend National Park, Texas. Atmos. Environ.; 2002; 36, pp. 5807-5813. [DOI: https://dx.doi.org/10.1016/S1352-2310(02)00578-2]
56. Simoneit, B.R.T. Characterization of organic constituents in aerosols in relation to their origin and transport: A review. Int. J. Environ. Analyt. Chem.; 1986; 23, pp. 207-237. [DOI: https://dx.doi.org/10.1080/03067318608076446]
57. Li, J.; Li, K.; Li, H.; Wang, X.; Wang, W.; Wang, K.; Ge, M. Long-chain alkanes in the atmosphere: A review. J. Environ. Sci.; 2022; 114, pp. 37-52. [DOI: https://dx.doi.org/10.1016/j.jes.2021.07.021]
58. He, D.; Simoneit, B.; Jara, B.; Jaffe, R. Composition and isotopic differences of iso- and anteiso-alkanes in black mangroves (Avicennia germinans) across a salinity gradient in a subtropical estuary. Environ. Chem.; 2015; 13, 6230630. [DOI: https://dx.doi.org/10.1071/EN15128]
59. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; Simoneit, B.R.T. Measurement of emissions from air pollution sources. 5. C1–C32 organic compounds from gasoline-powered motor vehicles. Environ. Sci. Technol.; 2002; 36, pp. 1169-1180. [DOI: https://dx.doi.org/10.1021/es0108077]
60. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; Simoneit, B.R.T. Measurement of emissions from air pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks. Environ. Sci. Technol.; 1999; 33, pp. 1578-1587. [DOI: https://dx.doi.org/10.1021/es980081n]
61. Popa, O.; Babeano, N.; Popa, I.; Nita, S.; Dinu-Parvu, C. Methods for obtaining and determination of squalene from natural sources. Biomed. Res. Int.; 2015; 2015, 367202. [DOI: https://dx.doi.org/10.1155/2015/367202]
62. Budsaereechai, S.; Hunt, A.; Ngernyen, Y. Catalytic pyrolysis of plastic waste for the production of liquid fuels in engines. RSC Adv.; 2019; 9, pp. 5844-5857. [DOI: https://dx.doi.org/10.1039/C8RA10058F]
63. Rogge, W.; Hildemann, L.; Mazurek, M.; Cass, G.; Simoneit, B. Sources of fine organic aerosol. 1. Charbroilers and meat cooking operations. Environ. Sci. Technol.; 1991; 25, pp. 1112-1125. [DOI: https://dx.doi.org/10.1021/es00018a015]
64. Schauer, J.J.; Kleeman, M.J.; Cass, G.R.; Simoneit, B.R.T. Measurement of emissions from air pollution sources. 3. C1–C29 organic compounds from fireplace combustion of wood. Environ. Sci. Technol.; 2001; 35, pp. 1716-1728. [DOI: https://dx.doi.org/10.1021/es001331e]
65. Kawamura, K.; Bikkina, S. A review of dicarboxylic acids and related compounds in atmospheric aerosols: Molecular distributions, sources and transformation. Atmos. Res.; 2016; 170, pp. 140-160. [DOI: https://dx.doi.org/10.1016/j.atmosres.2015.11.018]
66. Wang, Q.; Shao, M.; Zhang, Y.; Wei, Y.; Hu, M.; Guo, S. Source apportionment of fine organic aerosols in Beijing. Atmos. Chem. Phys.; 2009; 9, pp. 8573-8585. [DOI: https://dx.doi.org/10.5194/acp-9-8573-2009]
67. Tao, S.; Yin, X.; Jiao, L.; Zhao, S.; Chen, L. Temporal Variability of Source-Specific Solvent-Extractable Organic Compounds in Coastal Aerosols over Xiamen, China. Atmosphere; 2017; 8, 33. [DOI: https://dx.doi.org/10.3390/atmos8020033]
68. Simoneit, B.; Schauer, J.; Nolte, C.; Oros, D.; Elias, V.; Fraser, M.; Rogge, W.; Cass, G. Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles. Atmos. Environ.; 1999; 33, pp. 173-182. [DOI: https://dx.doi.org/10.1016/S1352-2310(98)00145-9]
69. Kemmo, S.; Ollilainen, V.; Lamopi, A.-M.; Piiromem, V. Determination of sigmasterol and cholesterol using atmospheric pressure chemical ionization liquid chromatography/mass spectroscopy. Food Chem.; 2007; 10, pp. 1438-1445. [DOI: https://dx.doi.org/10.1016/j.foodchem.2006.03.052]
70. Lee, A.; Lau, A.; Cheng, J.; Fang, M.; Chan, C. Source identification analysis for the airborne bacteria and fungi using a biomarker approach. Atmos. Environ.; 2007; 41, pp. 2831-2843. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2006.11.047]
71. Lim, Y.; Ziemann, P. Chemistry of Secondary Organic Aerosol Formation from OH Radical-initiated Reactions of Linear, Branched, and Cyclic alkanes in the presence of NO. Aerosol Sci. Technol.; 2009; 43, pp. 604-619. [DOI: https://dx.doi.org/10.1080/02786820902802567]
72. Tkacik, D.; Presto, A.; Donohue, N.; Robinson, A. Secondary organic aerosol formation from intermediate-volatility organic compounds: Cyclic, linear and branched alkanes. Environ. Sci. Technol.; 2012; 46, pp. 8773-8781. [DOI: https://dx.doi.org/10.1021/es301112c]
73. Fehsenfeld, F.; Hastie, D.; Chow, J.; Solomon, P. Particle and Gas Measurements. Particulate Matter Science for Policy Makers; McMurry, P.; Shepherd, M.; Vickery, J. Cambridge University Press: Cambridge, UK, 2004.
74. Jiminez, O.; Perez-Pastor, R.; Garcia-Alonso, S. Assessment uncertainty associated to the analysis of organic compounds linked to particulate matter of atmospheric aerosols. Talanta; 2010; 80, pp. 1121-1128. [DOI: https://dx.doi.org/10.1016/j.talanta.2009.08.036]
75. Yttri, K.E.; Schnelle-Kreis, J.; Maenhaut, W.; Abbaszade, G.; Alves, C.; Bjerke, A.; Bonnier, N.; Bossi, R.; Claeys, M.; Dye, C. et al. An intercomparison study of analytical methods used for quantification of levoglucosan in ambient filter samplers. Atmos. Meas. Tech.; 2015; 8, pp. 125-147. [DOI: https://dx.doi.org/10.5194/amt-8-125-2015]
76. Schantz, M.; Cleveland, D.; Heckert, N.; Kucklick, J.; Leigh, S.; Long, S.; Lynch, J.; Murphy, K.; Olfaz, R.; Pintar, A. et al. Development of two particulate matte standard reference materials (<4 µm and <10 µm) for the determination of organic and inorganic constituents. Anal. Bioanal. Chem.; 2016; 408, pp. 4257-4266. [DOI: https://dx.doi.org/10.1007/s00216-016-9519-7]
77. Mancilla, Y.; Medina, G.; Gonzalez, L.; Herckes, P.; Fraser, M.; Mendoza, A. Determination and Similarity Analysis of PM2.5 Emission Source Profiles Based on Organic Markers for Monterrey, Mexico. Atmosphere; 2021; 12, 554. [DOI: https://dx.doi.org/10.3390/atmos12050554]
78. Parshintsev, J.; Hyotylainen, T. Methods for characterization of organic compounds in atmospheric aerosol particles. Anal. Bioanal. Chem.; 2015; 407, pp. 5877-5897. [DOI: https://dx.doi.org/10.1007/s00216-014-8394-3] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25542579]
79. Rogge, W.; Ondov, J.; Bernardo-Bricker, A.; Sevimogu, O. Baltimore PM2.5 Supersite: Highly time-resolved organic compounds-sampling duration and phase distribution-implications for health effects studies. Anal. Bioanal. Chem.; 2011; 401, pp. 3069-3082. [DOI: https://dx.doi.org/10.1007/s00216-011-5454-9] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22009047]
80. Cretescu, I.; Lutic, D.; Manea, L. Electrochemical Sensors for Monitoring of Indoor and Outdoor Air Pollution. Electrochemical Sensors Technology; Muzibu, R.; Asiri, A. Intechopin: London, UK, 2017; [DOI: https://dx.doi.org/10.5772/intechopen.68512]
81. Popoola, O. Studies of Urban Air Quality Using Electrochemical Based Sensor Instruments. Ph.D. Dissertation; Queens College, University of Cambridge: Cambridge, UK, 2012.
82. Mead, M.I.; Popoola, O.A.M.; Stewart, G.B.; Landshoff, P.; Calleja, M.; Hayes, M.; Baldovi, J.J.; McLeod, M.W.; Hodgson, T.F.; Dicks, J. et al. The use of electrochemical sensors for monitoring urban air quality in low-cost, high density networks. Atmos. Environ.; 2013; 70, pp. 186-203. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2012.11.060]
83. Jiao, W.; Hagler, G.; Williams, R.; Sharpe, R.; Brown, R.; Garver, D.; Judge, R.; Caudill, M.; Rickard, J.; Davis, M. et al. Community Air Sensor Network (CAIRSENSE) project: Evaluation of low-cost sensor performance in a suburban environment in the southeastern United States. Atmos. Meas. Tech.; 2016; 9, pp. 5281-5292. [DOI: https://dx.doi.org/10.5194/amt-9-5281-2016]
84. Taylor, J.; Henshaw, G. Aeroqual, Aukland, New Zealand. AQM 65 Colocation Field Study Hartford, CT, USA; Unpublished Report 2017.
85. Analytical Methods Committee. What Should Be Done with Results below the Detection Limit? Mentioning the Unmentionable; Tech. Brief No. 5 Royal Society of Chemistry: London, UK, 2001.
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Fraser, Matthew 3 ; Hopke, Philip 4
; Ondov, John 5 ; Solomon, Paul A 6 ; Popoola, Olalekan 7 ; Hidy, George M 8 1 EnvironQuest, Lagos 102273, Nigeria;
2 School of Molecular Sciences, Arizona State University, Tempe, AZ 85287-1604, USA;
3 School of Sustainable Engineering and the Built Environment, Arizona State University, Tempe, AZ 85287-3005, USA;
4 Department of Public Health, University of Rochester, Rochester, NY 14642, USA;
5 Chemistry Department, University of Maryland, College Park, MD 20742, USA;
6 Independent Researcher, Henderson, NV 89052, USA;
7 Yusuf Hamied Department of Chemistry, Cambridge University, Cambridge CB2 1EW, UK;
8 Envair/Aerochem, Placitas, NM 87043, USA