Abstract
This study investigates the evolution of gaseous pollutants with prevailing meteorology using a network of Alpha sense low-cost portable air monitoring devices deployed to six selected sampling sites namely the Iron steel smelting factory (ISSF), Modomo, Eleweran, Fire service station, staff quarters, and OAU Teaching and research farm (OAUTRF). All six sampling sites were chosen along the pollution hotspot in Ile-Ife. The low-cost sensor has Optical particle counters (OPC-N2) which measured particulate matter and the gaseous pollutants were monitored with electrochemical sensors. At the sites, each sensor unit measured gaseous pollutants: carbon dioxide (CO2), ozone (O3), carbon monoxide (CO), and oxides of nitrogen (NO and NO2). Inbuilt into the monitoring device are temperature and relative humidity (RH) probes as well as a 2-D sonic anemometer to measure wind speed and directions. The directional dependence of the local source of the gaseous pollutant affecting each site was determined using the conditional probability function (CPF) model. The results showed that the highest mean values of 470.6 ppm, 7.2 ppb, and 20.6 ppb for CO2, NO, and NO2, respectively were recorded at the Modomo site while ISSF and Fire service sites were found to have the highest mean values of 67.9 ppb and 517.6 ppb, respectively for O3 and CO during the observational period. The mean values for temperature were noticed to increase in trend from OAUTRF (24.9 °C) to ISSF (26.0 °C). Also, CO2 gave a good relationship with meteorological parameters with R2 value ranging from 0.26 to 0.89 in all the sampling sites except ISSF. On contrary, weak relationships were found between the meteorological variables and other gaseous pollutants other than CO2 for both day and nighttime periods. The CPF model indicated that a larger percentage of the elevated O3, NO and NO2 were associated with pollutant transport at wind speed \(>2 m/s\) due to low event number. This study concludes that there is evidence of a significant effect of meteorological parameters on the gaseous pollutants as observed, while, the CPF plot shows a good agreement with the location of known pollution sources.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In most cities of the world, air pollution is one of the major environmental problems. Awareness and concerns about this seemingly intractable problem, the continuous deterioration of air quality, is on the increase [1, 2]. Generally, especially in developing economies, this situation of worsening air quality has been attributed to rapid industrial growth and unprecedented population explosion [3,4,5]. Population growth, urban sprawl, haphazard urban development and the proliferation of motor vehicles have contributed to traffic congestion and associated vehicular emissions with pollution considered endemic [6]. Studies have shown that the rapid rate of industrialization has not been matched with adequate urban planning for the control of environmental pollution associated with the magnitude of industrial development experienced [1, 7, 8]. Prominent anthropogenic sources include emission from industrial processes, domestic cooking and heating, iron smelting, municipal waste disposal and landfilling, transportation/vehicular emission, diesel-powered generators, offshore and onshore gas flaring, biomass burning, incineration and automobile emission while natural sources include volcanic eruption, lightning, wind erosion [9, 10]. Emissions from these sources are of great public concerns and cause significant undesirable alteration of ambient air composition; detrimentally impacting human health and environment even at low-level concentrations [11, 12]. The burden of diseases attributable to poor air quality falls heavily on the poorest members of the society specifically those with limited options for fuel and transport and consequently this intractable challenge impacts greatly on economic and social development. As such, ambient air pollution has contributed to higher mortality and significantly higher rates of Daliy Adjusted Life Years (DALY) [13,14,15]. Apart from the health risk, it contributes to changing climatic conditions. In the lower atmosphere, the effects of air pollutants include reduction in agricultural yield, loss of materials (such as marble, paints, steel and plastics), reduced visibility, alteration of Earth-Atmosphere energy budget and acid rain [16,17,18,19,20]. Atmospheric conditions play significant roles on the evolution and removal of pollutants from the ambient air. All the decisive weather processes which influence pollution spreading and dispersal take place within the mixing layer, a very thin layer of the troposphere which is the closest layer to the earth. The fate of air pollutants emitted from an industrial process such as iron smelting depends entirely on the immediate atmospheric process and conditions which ultimately affects the air quality around and downwind of such an industrial area [21]. Pollution from emission sources undergo different processes such as mixing, chemical transformation, dispersion, and transportation in the atmosphere [22]. It is therefore imperative to have adequate understanding of the dynamics of the relationship between meteorological parameters and concentration of gaseous pollutants. In addition to understanding the significant roles play by the atmosphere on the evolution and removal of pollutants from the ambient air, identifying nearby and far-off emission sources is a crucial component in managing air pollution. The conditional probability function (CPF) plot is a popular method for pinpointing the location of potential pollution nearby sources utilizing pollution data, wind speed, and direction [23, 24]. Using the CPF method, the direction of the sources that are responsible for the observed pollutant concentrations can be determined [25]. Another crucial technique for managing air quality is the identification of far-off emission sources. Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model has been shown to be an effective tool in tracking the trajectory of air mass [26, 27]. [28] coupled CPF with HYSPLIT, the result showed an enhanced performance of CPF technique providing more information on mid- to long-range sources. Previous studies on airborne pollutants carried out by [21, 29] across Ile Ife focused mostly on impact of prevailing meteorological conditions on the ambient concentration of particulate matter (PM). This study monitored gaseous pollutants with LCS, investigate the relationship between the meteorological factors and ambient concentrations of gaseous pollutants and provide both near (directional) and long distance source information of the measured pollutants using CPF and HYSPLIT respectively. [30] made an initial report of the spatial and temporal variation of pollutants however detailed relationship between ambient concentrations and meteorological factors were not made, and those analysis and the resulting information about pollutants is the focus of this work.
2 Material and methods
2.1 Site description
The study was conducted in Ile-Ife, a university town located in southwest Nigeria. Figure 1 shows the map of the study locations within Ile-Ife. A major road runs through the city, and other minor paved and unpaved roads connect the various areas of the city to ease the mobility of automobiles and passengers. The city play host to both public and private institutions as well as an iron smelting industry located at the outskirt of the city [30]. The sampling sites (Table 1) are made up of residential areas and areas where there could be heavy human exposures. Ile-Ife lies within the tropical rain forest zone with two distinct seasons (dry and wet seasons). The dry season is between November and March; a period of dry, dusty, warm and moist wind with little or no rainfall. The wet season starts from April to October each year with generally lower temperature and high relative humidity [21, 31].
2.2 Monitoring procedures and measurement of meteorological parameters
Ambient levels of five (5) gaseous pollutants—CO2, O3, CO, NO and NO2—were sampled at six locations downwind an iron smelting plant using low-cost amperometric electrochemical sensors manufactured by Alpha Sense Ltd, UK. The gaseous pollutants concentrations were monitored at sampling locations using a network of low-cost air quality (AQ) sensors (SNAQ boxes) and simultaneously operated over a period of two months (June–July, 2018). This low-cost sensor has inbuilt temperature and relative humidity sensor probes and 2-dimensional sonic anemometer for monitoring meteorological parameters. For detailed principle of operation of the low-cost sensor, see [32, 33]. In-depth comparisons of these monitors to reference monitoring systems were conducted during field tests in the United Kingdom by [33,34,35,36], while [37,38,39] have all investigated PM sensors. This (AQ) sensor sampled the concentrations of gaseous air pollutants at 20 s interval in part per billions (ppb) and part per millions (ppm). The data are stored in the USB storage device and retrieved every two weeks using the USB serial cable. Data are imported to a computer and reduced to 1 h average. Further statistical analysis was conducted using Open air, a package in R for air pollution analysis and MicroCal Origin (v7.0) software.
2.3 Conditional probability function (CPF)
The conditional probability function (CPF) majorly focuses on identifying the source locations that are most dominant; this is unlike bivariate polar plot which gives information on the directional dependence of concentrations on the wind speeds and directions as well as source dispersion. The CPF mathematical expression according to [24] is given by Eq. 1
where \({m}_{\Delta \theta }\) is the number of samples in wind sector \(\Delta \theta \) with a concentration greater than or equal to threshold.
2.4 Hybrid single-particle lagrangian integrated trajectory (HYSPLIT) model
HYSPLIT model is most widely used source-receptor model to describe the transport within the atmosphere, pollutants dispersion and deposition of dust particles as well as toxic materials using variety of simulations. The model is very effective in analyzing and interpreting the ground based in-situ together with satellite observations. More so, the HYSPLIT backward trajectory analysis is helpful in tracking the air mass origin and useful in establishing the source and receptor relationship [20, 40].
Thus, in addition to CPF in this study, we employed the back trajectory model by setting the period of dispersion model at the backward trajectory interval of 12 h and run for a 2-day (June 9 and July 20, 2018) using three different heights 500, 1000 and 1500 m above the ground level so as to determine the long range contribution air pollution to the identified sampling locations.
3 Results and discussion
3.1 Quality assurance and control (collocation)
The daily time series plots of the O3, NO, CO and PM2.5 concentrations for the co-location measurement for SNAQ boxes (S12, S22, S28, S30 and S32) during the period of 1st to 7th of November, 2018 is presented in Fig. 2a–d. The daily time series plots of the SNAQ boxes are consistent and show remarkable similar features [41]. This great similarity inspires trust in the reliability of these sensors. Also, the result of the Kruskal–Wallis non-parametric t-test (p-value > 0.05) showed that the measured pollutants by the SNAQ boxes were not statistically significant. This means that when installed individually, each SNAQ box can function independently.
3.2 Variations of pollutants’ concentrations with meteorological parameters
Table 2 gave the statistical analysis of mass concentrations of gaseous pollutants and meteorological parameters at the six sampling locations. At ISSF site, CO, NO, and NO2 data were missing due to the faulty sensor nodes. Table 2 showed that CO2 concentration ranged from 408–537 (ISSF), 380–569 (Modomo), 382–593 (Eleweran), 396–549 (Fire service), 386–536 and 393–654 ppm for staff quarters and OAUTRF, respectively. OAUTRF and Modomo sites had the highest and the lowest concentrations of CO2 of 653.8 and 379.6 ppm, respectively during the observation period. Also, nighttime CO2 mean values were higher than those of the daytime at Modomo, Eleweran, Fire Service, Staff Quarters and OAUTRF sites. This was attributed to the increase in CO2 emissions at the sites together with the inhibition of CO2 dispersion due to the fact that the atmosphere is more stable during the nighttime. The low mean values observed during the daytime suggest that CO2 is taken up in photosynthesis. On contrary, daytime mean values was higher than the nighttime mean value at ISSF. This could be attributed to the superposition of anthropogenic activities over photosynthetic activities as a result of the closeness of the site to the Iron and Steel Smelting Factory, Fasina. The mean concentrations of ozone (O3) found at all the sampling locations ranged between 57.7 and 69.0 ppb. The highest and lowest O3 mass concentrations of 67.9 and 12.2 ppb respectively were noticed at ISSF for the entire period (Table 2). More so, the daytime daily mean values are higher than the nighttime values, which are adduced to strong solar heating which greatly enhances photochemical reactions leading to O3 formation [42, 43]. The elevated O3 level observed at ISSF, Modomo and Eleweran sites may be attributed to the emission from the Iron Steel Smelting Factory, Fasina due to the close proximity of these sites to the factory. The CO values for the entire period ranged from 217 to 1541 at Modomo, 231–997 (Eleweran), 231–1433 (Fire Service), 231–846 (Staff Quarters) and 227–785 ppb, at OAUTRF. The mean (± SD) of the concentrations of CO for the entire period were 341.0 ± 118.3, 382.9 ± 101.6, 517.6 ± 200.1, 348.9 ± 66.9 and 380.6 ± 65.8 ppb at Modomo, Eleweran, Fire Service, Staff Quarters and OAUTRF, respectively.
Also, the highest mean concentrations (517.6 ppb) of CO for the entire period was observed at the Fire Service and this may be linked to vehicular emission and incomplete combustion of fossil fuel from diesel-powered generating set from banks in around the sampling location. Daytime CO mean values of 350.9, 381.5, 510.8, 374.4 and 394.3 were recorded while nighttime CO mean values were found to be 338.1, 412.9, 562.2, 356.8 and 390.5 ppb at Modomo, Eleweran, Fire Service, Staff Quarters and OAUTRF, respectively. As presented in Table 2, the mean (± SD) of the concentrations of NO were 7.2 ± 7.5, 6.5 ± 6.2, 7.0 ± 7.3, 6.9 ± 6.5 and 4.4 ± 5.3 ppb at Modomo, Eleweran, Fire service, staff quarters and OAUTRF, respectively. During the daytime, the value of mean concentrations recorded are 9.3, 6.8, 6.9, 6.2 and 7.0 ppb while nighttime mean concentrations were 5.7, 6.3, 6.7, 3.7 and 2.8 ppb at Modomo, Eleweran, Fire service, staff quarters and OAUTRF, respectively (Table 3). As observed, the diurnal patterns of NO in Modomo, Fire service, Staff quarters and OAUTRF sites showed two peaks corresponding to morning and evening rush hours because of the variations in traffic flow and atmospheric dispersion [44]. Ambient concentration of NO was observed to be minimum during afternoon hours due to increased advective and convective dilution arising from the reactions with the O3 which is created during this period [44]. For NO2, Modomo site had the highest mean values (20.6 ppb) for the entire period due to the close proximity of the site to Iron and Steel Smelting factory followed by Fire service, Eleweran, OAUTRF and Staff quarters with concentrations of NO2 of 18.6, 16.3, 14.8 and 14.6 ppb, respectively. The highest wind speed (4.3 m/s) and the lowest one (0.1 m/s) were observed at ISSF and Modomo site, respectively (Table 2). On average, the highest and lowest temperature values of 26 and 23.5 OC were recorded at ISSF and staff quarters, respectively. Relative humidity is high in ISSF, Eleweran, fire service and observed to be low in Modomo, staff quarters and OAUTRF, respectively. The day hourly time variation plots of the pollutants at the measurement sites are shown in Figs. 3 and 4.
3.3 Correlation matrix plot of gaseous pollutants and meteorological parameters
The correlation matrix of the gaseous pollutants and meteorological parameters at all the study sites are presented in Fig. 5. Among the meteorological parameters, temperature had relatively moderate correlation (0.45–0.57) with wind speed in all the six sites. Also, O3 concentrations had moderate correlation with wind speed in all the six sites with the highest correlation (r = 0.54) observed at OAUTRF site. Temperatures were positively correlated with O3 concentrations in all the sites except for Modomo that show a moderate negative correlation (Fig. 5). It was observed that the O3 concentrations gave a relatively weak negative correlation with relative humidity in all the study sites and this is attributable to depression of O3 through process of NOX titration. Wind speed and temperature showed a very strong negative correlation with the CO2 concentrations except for ISSF site which recorded − 0.03 and − 0.04 for correlation of CO2 with wind speed and temperature, respectively. Correlation between CO2 concentrations and relative humidity were seen to be very high in Modomo (0.93), Eleweran (0.83), Fire service (0.66), Staff quarters (0.83) and OAUTRF (0.83) while ISSF site had relatively weak correlation (0.02).
The correlation matrix plot in Fig. 4 for CO concentrations with wind speed and temperature gave a relatively weak correlation in the entire sites except for Fire service with positive correlation of 0.22 and 0.25 for wind speed and temperature, respectively. It was noted that there is a poor positive correlation (0.02 and 0.23) between CO concentrations and relative humidity except Fire service site which behaved differently with correlation value of − 0.28. For NO concentrations, strong positive correlation with wind speed and temperature were recorded except for some sites (Eleweran and Staff quarters) which had low correlation value. Correlation between NO and relative humidity was observed to be negative and very strong in all the sites during observation period. As presented in Fig. 4, NO2 concentrations are negatively correlated with wind speed and temperature except for Eleweran (0.01) and Modomo (0.88) which indicated a very weak and strong positive correlation with wind speed and temperature respectively. NO2 was positively correlated with relative humidity but the strength is weaker in all the study sites except Modomo site with correlation value of \(0.53\).
Generally, it was observed that O3, CO2, CO and NO2 gave a relatively weak negative correlation with wind direction in all the sites while NO mass concentrations had a weak positive correlation with wind direction. This suggests that the source of NO was predominantly local to the sites.
3.4 Relationship between day- and nighttimes concentration of gaseous pollutants and meteorological parameters
The results of the relationship between CO2 concentrations and wind speed for the entire sampling sites are presented in Table 4. The relationship between the CO2 and wind speed showed an exponential distribution curve (decay curve). During the daytime, coefficient of determination (R2) values ranged from \(0.03\) to \(0.68\) while the nighttime values ranged from 0.03 to \(0.54\) for the entire period (Supp.Fig. 1a and b). As observed, CO2 had poor relationship with wind speed at Modomo. Table 4 indicates that nighttime relationship between CO2 and wind speed had a better model fit than the daytime in some stations such as Modomo, Fire service, Staff quarters and OAUTRF. Also, the decay curve suggests that there was an increase in CO2 concentrations at low wind speed and this was evidence that there was an inhibition of the dispersion of CO2 (Supp.Fig. 1a and b). At high wind speed, CO2 concentration decreased significantly due to an enhanced dispersion of CO2. Elevated CO2 level might be attributed to the release of CO2 by respiration from both plants and with the inhibition of CO2 dispersion during the nighttime. CO2 decreases during the daytime as it is taken up in photosynthesis [45].
The results in Table 4 shows that temperature with CO2 had an exponential decay curve while relative humidity had an exponential growth curve with CO2. However, the decay curve fit of the relationship between CO2 and temperature gave coefficient of determination (R2) values ranging from \(0.01\) to \(0.88\) during the daytime while nighttime had coefficient of determination (R2) values range between \(0.02\) and \(0.83\) (Supp.Fig. 2 a and b). As observed, the decay curve indicates a strong and better model fit of the relationship between CO2 and temperature as larger percentage of CO2 fall along the fit line during the day time and nighttime periods in all the stations except ISSF which was very low for both day and nighttime. More so, the decay curve fit at ISSF was observed to be better during the daytime than the nighttime as more than \(98 \%\) of the variance in CO2 is explained by unidentified variables other than the temperature during the nighttime. The coefficient of determination (R2) of temperature versus CO2 concentration at ISSF site suggests that there was superposition of the anthropogenic activities over the photosynthetic activities. For relative humidity versus CO2 concentration, daytime had a growth curve with coefficient of determination (R2) values ranges from \(0.01\) to \(0.90\) while nighttime periods had coefficient of determination (R2) values from \(0.03\) to \(0.84\) (Supp.Fig. 3a and b, Table 4).
During the daytime, it is observed that the coefficient of determination (R2) relative humidity versus CO2 concentration was higher above \(0.75\) in trend from ISSF to OAUTRF except for Fire service station (\(0.45\)) which did not follow the trend due to the humid condition in the site (Supp.Fig. 3a and b). Generally, the decay curves of the relationship between CO2 and temperature show that CO2 concentration was high when temperature was low and vice versa, an indication that plants consume more CO2 as a result of the increasing photosynthetic rates as well as carbon cycle of the atmosphere due to high temperature during the daytime which lower the CO2 mass concentrations [45]. The relationship between O3 concentrations and wind speed are presented in Table 4. There was a positive linear relationship between the concentrations of O3 and wind speed in all the sampling sites during the daytime and nighttime. The data for Staff Quarters was not included in the analysis because the electrochemical sensor node for O3 was faulty. The coefficient of determination (R2) ranges between \(0.01-0.24\) and \(0.03-0.40\) for day and night times, respectively (Supp.Fig. 1c and d). Temperature was also observed to have positive linear relationship with O3 during the day and night times in most of the sites except Eleweran and OAUTRF, sites which shows negative relationship with temperature (Supp.Fig. 2c and d). In contrast, relative humidity had negative relationship with O3 during the day and night times in all the sites except Eleweran site (Supp.Fig. 3c and d).
The relationship between O3 concentration with temperature and relative humidity during the nighttime showed a better correlation than the one observed for the daytime. In both cases, variation of O3 with temperature and relative humidity were poorly correlated with coefficient of determination (R2) ranging from \(0.01-0.06\) and \(0.06-0.24\) (for temperature) and from \(<0.01-0.10\) and \(0.07-0.29\) (for relative humidity) during the day and night times, respectively (Supp.Fig. 3c and d, Table 4).
The positive association of temperature with O3 concentrations during the daytime which suggest that higher temperature is accompanied with intermittent lower values of relative humidity which eventually favor the accumulation of O3 concentrations. The higher concentration of O3 during the daytime is attributed to strong solar heating which greatly promotes photochemical reactions between the abundant built up of air pollutants [42, 46, 47].
Wind speed and temperature had negative relationship with CO concentration except for Fire service and Staff quarters with coefficient of determination (R2) ranged from \(0.02-0.08\) and \(0.01-0.10\) (for wind speed) and \(0.01-0.08\) and \(0.01-0.04\) (for temperature) during the day and nighttimes, respectively for the entire period (Supp.Fig. 1e and f and Supp.Fig. 2e and f). Also, relative humidity (RH) had positive correlation with CO in most of the site during the daytime except for Fire service while Modomo and Eleweran sites were positively correlated with RH and in contrast, Fire Service, staff quarters and OAUTRF had negative correlation with CO during the nighttime (Supp.Fig. 3e and f, Table 4).
Generally, it was observed that coefficient of determination (R2) for the relationship of CO with wind speed (\(0.01-0.08\)), temperature (\(0.01-0.10\)) and relative humidity (\(0.01-0.14\)) was very low which indicated that CO concentration had little or no relationship with the meteorological parameters.
Ambient levels of NO, had positive relationship with wind speed and temperature but negative correlation with relative humidity during the day and night times for the entire period. It was also observed that NO concentrations had better correlation with wind speed with coefficient of determination, R2, ranging between \(0.02\) and \(0.29\) (Supp.Fig. 1 g and h). More so, the coefficient of determination R2 observed for the correlation between CO with temperature and relative humidity indicated that it has a range of \(0.32-0.61\) and \(0.36-0.66\) during day and night times for temperature and relative humidity, respectively (Supp.Fig. 2g and h and Supp.Fig. 3g and h).
NO2 concentrations had positive correlation with the wind speed and temperature while the relative humidity had negative relationship with NO2 concentrations during the day and night times except for Modomo and OAUTRF sites. The obtained coefficient of determination (R2) ranged from \(0.01-0.04\), \(0.03-0.12\) and \(0.01-0.20\) during the daytime while nighttime R2 value ranged from \(0.01-0.26\), \(0.01-0.45\) and \(0.03-0.45\) for wind speed, temperature and relative humidity, respectively (Supp. Figure 1i and j, Supp. Figure 2i and j, and Supp. Figure 3i and j).
3.5 CPF plots
Figures 6–14 shows the CPF plots at 80th percentiles for \(C{O}_{2}, {O}_{3}, CO, NO and N{O}_{2}\) at ISSF, Modomo, Eleweran, Fire Service, Staff Quarters and OAUTRF, respectively. The CPF plots revealed that the probability was low at ISSF site with about 25% of the \(C{O}_{2}\) concentrations \(>471 ppmv\) been transported from the Iron-Steel smelting factory (South west) during the daytime while nighttime clearly identified the North and west as the dominant source directions for \(C{O}_{2}\) concentrations \(>463 ppmv\) (with probability of 25%) Fig. 6. As observed at Modomo, the daytime plot showed that the main sources of \(C{O}_{2}\) concentrations are primarily transported from the north-east of the site with probability of up to 90% of \(C{O}_{2}\) concentrations \(>478 ppmv\). During the nighttime, probability of \(\sim 60 \%\) that \(C{O}_{2}\) concentrations \(>510 ppmv\) are transported at low wind speed \((<1{ms}^{-1})\) from North–East and North–West sectors of the site were recorded. The CPF analysis at Eleweran site indicated that the probability of up to 35% of \(C{O}_{2}\) concentrations \((>466 ppmv\)) and 50% of \(C{O}_{2}\) concentrations \((>503 ppmv)\) came from west and north-west during daytime and nighttime, respectively. At fire service, there is \(\sim 35 \%\) (daytime) and \(\sim 45 \%\) probabilities that \(C{O}_{2}\) concentrations \(>426 and>452 ppmv\), respectively are transported at low wind speeds from the west.
Also, CPF analysis for \(C{O}_{2}\) concentrations in staff quarters showed that there is \(\sim 60\%\) probability that concentrations \(>429 ppmv\) (80 percentile) occur under calm conditions \((< 1{ms}^{-1})\) for daytime and an \(\sim 50\%\) probability that concentrations \(>454 ppmv\) was transported from the north-west for both time periods Fig. 7. At OAUTRF during daytime, highest concentrations of \(C{O}_{2}\) \((>460 ppmv)\) are estimated to originate from North-East with a probability of 90% while nighttime CPF plot indicate that the \(C{O}_{2}\) pollutant sources are from the north with a probability of \(60 \%.\)
As observed in Fig. 8, the contributing sources for ozone concentrations > 80th percentile of their respective values are identified. At ISSF site, there is \(\sim 25\%\) probability that \({O}_{3}\) concentrations \(\left(>89 ppb\right)\) occur during the daytime from the South–West and South–East. Nighttime showed high probability of up to 100% that \({O}_{3}\) concentrations \(>76 ppb\) are transported from the South–West and South–East due the emissions from the industrial activities at Iron-steel smelting factory. At Modomo, from 80% to about 100% account for the probability that \({O}_{3}\) concentrations \(>81 ppb\) and \(>70 ppb\) occuring during the daytime and nighttime, respectively and are transported from North–East and South–West directions. However, the main source contributing to the high \({O}_{3}\) concentrations from the Southwest of the site during nighttime is in the direction relative to Iron-steel smelting factory, Fasina. The CPF plot for Eleweran site indicated that the probability varied from 80% up to 100% for \({O}_{3}\) concentrations \(>70 ppb\) (for daytime) and \(>68 ppb\) (for nighttime) originating from the South–East for both time periods.
Also, south-west direction was observed during the nighttime as one of the dominant sources of \({O}_{3}\) concentrations and little trace of \({O}_{3}\) was equally noticed from northwest direction. CPF plot at Eleweran further showed that there is a higher probability (from 80% up to 100%) for \({O}_{3}\) concentrations from South–East which is an indication that the high \({O}_{3}\) generated from this wind direction corridor may be attributed to emission from Ede road (high way connecting Ile-Ife to neighboring towns) located in this direction. As observed at OAU Staff quarters, daytime CPF plot indicated that the main sources of \({O}_{3}\) are primarily transported from the north-east for \({O}_{3}\) concentrations \(>69 ppb\) and from south for \({O}_{3}\) concentrations \(>65 ppb\) during the daytime and nighttime, respectively Fig. 9. At OAUTRF, there is probability \((>80 \%)\) for \({O}_{3}\) concentrations higher than the 80th percentile \((69 ppb)\) to occur under low wind speed (1 m/s) from the south-east and north-east during the daytime. At nighttime, \(>80 \%\) and up to 100% probability of \({O}_{3}\) concentrations \(>65 ppb\) was transported from the south-west and south-east wind direction corresponding to the direction of emission from diesel powered generator used at OAUTRF to power silos. The high probabilities of ozone levels recorded in all the sites may be attributed to the rising photochemical reactions of nitrogen oxides in the presence of sunlight.
The CPF plots of CO concentrations in Figs. 10 and 11 indicated that all the sampling stations recorded high level CO concentrations at low wind speed which is an indication that most of CO concentrations were emitted from the local sources. During the daytime at Modomo, high probability (60%) of CO concentrations \(>366 ppb\) were transported from east wind direction corridor while nighttime sources is related to south-east directions with probability up to 70% for CO concentrations \(>402 ppb\). The impact on this Modomo site from the eastern directions during the daytime is strongly associated with the vehicular emissions from the Ife-Ibadan express road. At Eleweran, CO was dominated from the north with probability of up to 60% for CO concentrations \(>444 ppb\) while nighttime clearly identified northeast as the dominant source with probability of about 35% for CO concentrations \(>466 ppb\). As observed during the daytime in CPF plot for Eleweran, the main source of CO is primarily located along Ibadan-Ife-Akure expressway (North). During the daytime at Fire service, the CPF analysis indicate that the probability (40%) was observed for CO concentrations \(>802 ppb\) which were most likely to be generated from vehicle emissions from the campus major road which is along south-west direction of the measurement location. The low probability (20%) observed during nighttime for CO concentrations \(>550 ppb\) were transported from southern wind direction corridor. More so, the CPF plots for OAU staff quarters during the daytime indicate the presence of a major source from the south-east with probability of 50% for CO concentrations \(>408 ppb\). At nighttime, there was clear indication that 40% probabilities of CO concentrations \(>413 ppb\) was transported from south-west direction which may be linked to the emissions from the Iron-steel smelting factory, Fasina. At OAUTRF, it was noticed that major source of CO concentrations was traceable to the southern directions with probability of 18% \(( for CO>424 ppb)\) and 16% \(( for CO>454 ppb)\) during the daytime and nighttime, respectively.
Figures 12 and 13 present CPF plots of NO concentrations at the measurement sites. As observed at Modomo, the highest probability (70%) for NO concentrations \(>16 ppb\) came from West and North–West directions during the daytime. At nighttime, the highest probability 50% for NO concentrations \(>4.1 ppb\) source contributions originated from the North–West and North–East directions under the wind speed > 1.5 m/s. The CPF plot of NO concentration at Eleweran site showed that the probability (60% for NO \(>14 ppb\)) was higher during the daytime than the nighttime (30% for NO \(>4.7 ppb\)). The high probability during the daytime were found to associated with the South eastern sources while North western sources contributing highly in the nighttime at Eleweran site. At Fire service station, high probabilities of 25% was observed for NO \(>18 ppb\) from the west and southeastern directions during daytime, while nighttime showed that there is probabilities of 70% that NO concentrations \(>22 ppb\) was transported from southwest direction owing to the incomplete combustion of fuels of the generator which is positioned in the direction. At staff quarters, the potential source is of NO \(>16 ppb\) (with probabilities 20%) presented in the CPF plot was attributed to southwest direction during the daytime. During the nighttime, northwest direction was observed to be the transportation source and there is a low probability of up to 20% for NO concentrations \(>2 ppb\). The CPF plots for OAUTRF in Fig. 13 during daytime indicated that there is an \(\sim 80 \%\) probability that NO concentrations \(>11 ppb\) are transported at low wind speeds in the range > 1 m/s to < 3 m/s from northeast and southeast corridors of the site location. During nighttime, the CPF analysis for NO concentrations \(>2 ppb\) with probability up to 60% was strongly associated with the transport from the Northeast and Northwest directions. In Fig. 14, the CPF analysis for NO2 concentration at Modomo site show that the probability ranges from 80% up to 100% for NO2 concentrations \(>24 ppb\) and it was transported from the northeast direction and with little trace from northwest during the daytime.
The nighttime indicates that Southwest and Southeast directions dominate the NO2 emissions for concentrations \(>28 ppb\) with probabilities of occurrence ranging from 80% up to 100%. During the daytime at Eleweran site, the NO2 emission is dominantly from Southeast direction for concentrations \(>22 ppb\) (\(\sim 100 \%\) probability) while nighttime occurrence show that NO2 concentrations \(>23 ppb\) are influenced by emission source from the Southeast and Northeast. During the daytime at Fire service the high probabilities of NO2 concentrations \(>27 ppb\) from 80 upto 100% are closely associated with the transport-related activities around the OAU campus banking area while nighttime CPF analysis for NO2 concentrations \(>23 ppb\) with probability up to 100% was transported from the southeast direction. The CPF plot for staff quarters in Fig. 14 indicates that northeast sources are associated with NO2 concentrations \(>22 ppb\) with probability of ranging 80% up to 100% during the daytime. As observed during the nighttime, the main source of NO2 concentrations \(>16 ppb\) are primarily located in southwest and partly in the northeast direction with probabilities of about 70%. The probability of up to 100% for NO2 concentrations \(>16 ppb\) coming from the southwest direction is observed during the daytime at OAUTRF site. On the other hand, the results of CPF analysis during nighttime clearly identified southwest and northwest as the dominant source directions for NO2 concentrations \(>18 ppb\).
3.6 Air mass trajectories over the sampling locations
The pathways of the air mass transport were analyzed using the backward trajectory analyses for the sampling locations as shown in Figs. 15 and 16. The results showed that the trajectories for different heights (500, 1000 and 1500 m) emanated from the same potential source region as observed in all the sampling sites. Also, HYSPLIT model indicated that the dominant source can be attributed to the maritime flow emanating from the Gulf of Guinea. Furthermore, on July 20, 2018, the HYSPLIT simulation results showed that the pollutants arriving at 500 m and 1000 m heights traveled at long distances at the study sites. This represents faster trajectory relative to the short trajectory pathway arriving at 1500 m height.
However, the transport from ground-level is likely to be the major transport pathway for the gaseous pollutants at few kilometers in distance as indicated in the HYSPLIT back trajectory model.
4 Conclusion
The gaseous air pollutants mass concentrations and its relationship with meteorological parameters were observed at six sampling locations in Ile-Ife area between June and July, 2018 using low cost electrochemical sensors. The study found CO2 had good exponential relationship with meteorological parameters in all the sampling sites except ISSF and Modomo. On contrary, other gaseous pollutants such as O3, CO and NO2 showed that there was weak relationship between gaseous pollutants and meteorological variables for both day and nighttime periods. Also, the relationship between CO2 and wind speed and temperature gave an exponential distribution (decay) curve while the relationship of other gases with wind speed and temperature were observed to be linear. The CPF model results showed that larger percentage of high concentration of gaseous pollutants such as O3, NO and NO2 were associated with pollutant transport at wind speed \(>2 m/s\) due to low event number. Also, the CPF model identified the meteorological parameters importance in determining the sources of pollutants concentrations peaks. HYSPLIT model indicated that the dominant source of pollutants arriving at the study sites originated from maritime flow from the Gulf of Guinea while pollutant transmission from the ground level is most likely the main transport pathway. Local sources such as traffic, burning, and the activities of iron smelting factory to the southwest of the research region, were likely key sources. The area is a rapidly rising semi-urban hamlet with increasing urbanization. As a result, mitigating techniques to reduce local pollutant emissions are required. Furthermore, reducing emission sources to the southwest of the study region, such as the scrap processing facility, could reduce the burden of pollution across the area under investigation.
Data availability
Experimental data for this study is available by contacting the corresponding author.
References
Chan CK, Yao X. Air pollution in mega cities in China. Atmos Environ. 2008;42(1):1–42.
Qiu G, Song R, He S. The aggravation of urban air quality deterioration due to urbanization, transportation and economic development-panel models with marginal effect analyses across China. Sci Total Environ. 2019;651:1114–25.
Fang M, Chan CK, Yao X. Managing air quality in a rapidly developing nation: China. Atmos Environ. 2009;43(1):79–86.
Ipeaiyeda A, et al. Gaseous air pollutants emissions from ota industrial estate in Ogun state, Nigeria. Ife J Sci. 2018;20(1):145–54.
Rajper SA, Ullah S, Li Z. Exposure to air pollution and self-reported effects on Chinese students: a case study of 13 megacities. PLoS ONE. 2018;13(3): e0194364.
Amegah AK, Agyei-Mensah S. Urban air pollution in Sub-Saharan Africa: time for action. Environ Pollut. 2017;220:738–43.
Orogade SA, et al. Source apportionment of fine and coarse particulate matter in industrial areas of Kaduna, Northern Nigeria. Aerosol Air Qual Res. 2016;16(5):1179–90.
Owoade OK, et al. Characterization and source identification of airborne particulate loadings at receptor site-classes of Lagos Mega-city, Nigeria. J Air Waste Manag Assoc. 2013;63(9):1026–35.
Fawole OG, Cai X-M, MacKenzie A. Gas flaring and resultant air pollution: A review focusing on black carbon. Environ Pollut. 2016;216:182–97.
Tiwary A, Williams I. Air pollution: measurement, modelling and mitigation. Boca Raton: CRC Press; 2018.
Zmirou D, et al. Traffic related air pollution and incidence of childhood asthma: results of the Vesta case-control study. J Epidemiol Community Health. 2004;58(1):18–23.
Bagieński Z. Traffic air quality index. Sci Total Environ. 2015;505:606–14.
Lim J-M, et al. Source apportionment of PM10 at a small industrial area using positive matrix factorization. Atmos Res. 2010;95(1):88–100.
Dominici F, et al. Revised analyses of the national morbidity, mortality, and air pollution study: mortality among residents of 90 cities. J Toxicol Environ Health A. 2005;68(13–14):1071–92.
Babatola SS. Global burden of diseases attributable to air pollution. J Public Health Afr. 2018;9(3):813.
Pillai PS, Babu SS, Moorthy KK. A study of PM, PM10 and PM2.5 concentration at a tropical coastal station. Atmos Res. 2002;61(2):149–67.
Schwartz J, Dockery DW, Neas LM. Is daily mortality associated specifically with fine particles? J Air Waste Manag Assoc. 1996;46(10):927–39.
Heimann I, et al. Source attribution of air pollution by spatial scale separation using high spatial density networks of low cost air quality sensors. Atmos Environ. 2015;113:10–9.
Zeng S, Zhang Y. The effect of meteorological elements on continuing heavy air pollution: a case study in the Chengdu area during the 2014 spring festival. Atmosphere. 2017;8(4):71.
Li X, et al. Overviewing the air quality models on air pollution in Sichuan Basin, China. Chemosphere. 2021;271: 129502.
Omokungbe OR, et al. Analysis of the variability of airborne particulate matter with prevailing meteorological conditions across a semi-urban environment using a network of low-cost air quality sensors. Heliyon. 2020;6(6): e04207.
Sumesh R, et al. Particulate matter concentrations in the southern tip of India: temporal variation, meteorological influences, and source identification. Earth Syst Environ. 2017;1(2):1–18.
Uria-Tellaetxe I, Carslaw DC. Conditional bivariate probability function for source identification. Environ Model Softw. 2014;59:1–9.
Carslaw DC, Beevers SD. Characterising and understanding emission sources using bivariate polar plots and k-means clustering. Environ Model Softw. 2013;40:325–9.
Vedantham R, Norris G, Duvall R. Receptor and hybrid modeling tools. Wiley StatsRef Statistics Ref Online. 2014. https://doi.org/10.1002/9781118445112.stat07801.
Draxler RR, Rolph GD. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) model access via NOAA ARL READY website (http://www. arl. noaa. gov/ready/hysplit4. html). Silver Spring: NOAA Air Resources Laboratory; 2003.
Stein A, et al. NOAA’s HYSPLIT atmospheric transport and dispersion modeling system. Bull Am Meteor Soc. 2015;96(12):2059–77.
Bae M-S, et al. Identifying pollutant source directions using multiple analysis methods at a rural location in New York. Atmos Environ. 2011;45(15):2531–40.
Owoade O, et al. Correlation between particulate matter concentrations and meteorological parameters at a site in Ile-Ife, Nigeria. Ife J Sci. 2012;14(1):83–93.
Owoade OK, et al. Spatial-temporal variation and local source identification of air pollutants in a semi-urban settlement in Nigeria using low-cost sensors. Aerosol Air Quality Res. 2021;21(10): 200598.
Jegede OO, et al. The Nigerian micrometeorological experiment (NIMEX-1): an overview. IFE J Sci. 2004;6(2):191–202.
Popoola O et al. A portable low-cost high density sensor network for air quality at London Heathrow airport. EGU Gen Assem. 2013. p. 7–12.
Popoola OAM, et al. Use of networks of low cost air quality sensors to quantify air quality in urban settings. Atmos Environ. 2018;194:58–70.
Chatzidiakou L, Krause A, Popoola OA, Di Antonio A, Kellaway M, Han Y, Squires FA, Wang T, Zhang H, Wang Q, Fan Y. Characterising low-cost sensors in highly portable platforms to quantify personal exposure in diverse environments. Atmos Meas Tech. 2019;12(8):4643–57.
Munir S, et al. Analysing the performance of low-cost air quality sensors, their drivers, relative benefits and calibration in cities-a case study in Sheffield. Environ Monit Assess. 2019;191(2):94.
Sousan S, et al. Evaluation of the Alphasense optical particle counter (OPC-N2) and the Grimm portable aerosol spectrometer (PAS-1.108). Aerosol Sci Technol. 2016;50(12):1352–65.
Li J, et al. Evaluation of nine low-cost-sensor-based particulate matter monitors. Aerosol Air Quality Res. 2020;20(2):254–70.
Crilley LR, et al. Evaluation of a low-cost optical particle counter (Alphasense OPC-N2) for ambient air monitoring. Atmos Meas Techn. 2018;11(2):709–20.
Abiye OE, et al. Chemical footprints of harmattan dust and traffic corridor particulates monitored at two environmentally distinct geopolitical zones in Nigeria. Environ Sci Pollut Res. 2020;27(31):39317–29.
Olufemi AP et al. Spatiotemporal distribution of pollutants and impact of local meteorology on source influence on pollutants’ level in a traffic airshed in Lagos megacity, Nigeria. Environ Monit Assess. 2023 (in press).
Olajire AA, Azeez L, Oluyemi EA. Exposure to hazardous air pollutants along Oba Akran road, Lagos-Nigeria. Chemosphere. 2011;84(8):1044–51.
Gerasopoulos E, et al. Photochemical ozone production in the Eastern Mediterranean. Atmos Environ. 2006;40(17):3057–69.
Colls J. Air pollution: measurement, modelling and mitigation. Milton park: CRC Press; 2002.
Ajao A, Jegede O. Diurnal variation of carbon dioxide flux above growing cowpea (Vigna unguiculata [L] Walp.) at an agricultural site in Ile-Ife, Nigeria. Ife J Sci. 2019;21(2):345–56.
Seinfeld JH, Pandis SN. From air pollution to climate change. In: Seinfeld JH, editor. Atmospheric chemistry and physics. Hoboken: Wiley; 1998. p. 1326.
Nair PJ, et al. Ozone trends derived from the total column and vertical profiles at a northern mid-latitude station. 2013. Atmos Chem Phys. https://doi.org/10.5194/acp-13-10373-2013.
Author information
Authors and Affiliations
Contributions
The research was conceived by: ORO, APO. Writing manuscript—original: ORO, APO. Methodology and formal analysis: ABT, POA. Writing manuscript—review and editing: APO, POA.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Omokungbe, O.R., Olufemi, A.P., Toyeje, A.B. et al. Investigating the evolution of gaseous air pollutants with prevailing meteorology across selected sites within a pollution hotspot in Ile-Ife, Southwestern Nigeria. Discov Environ 1, 5 (2023). https://doi.org/10.1007/s44274-023-00006-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s44274-023-00006-0