Introduction

Sub-Saharan Africa is one of the least urbanized regions in the world, but the urban population of Africa is growing very rapidly (Potts 2009). Poverty-stricken citizens of Africa, particularly in the urban population, are most negatively impacted by poor air. Sources of air pollution include those from natural and anthropogenic sources (Asubiojo 2016). It is estimated that over one billion people are exposed to open-air air pollution yearly (Zhang and Day 2015). Urban air pollution is responsible for over one million premature deaths each year in India and China (Dockery and Evans 2017). In spite of this, some mysterious and annual air pollution episodes in Africa have received less attention.

The knowledge of ambient air quality is relevant to assessing the exposure levels and providing mitigation measures. (Bose and Diette 2016; Kim et al. 2013; Shi et al. 2016). The ambient air quality depends directly on the initial concentrations of specific contaminants, the emission sources and atmospheric conditions. Air quality sampling and monitoring can be used to establish the baseline quality of regions (Seguí et al. 2016). This includes identification of specific air pollution parameters expected to have significant impacts and assessing their existing levels in ambient air within the impact zone of the study area metropolis. Among the major sources of particulate pollution are the annual Saharan dust episodes.

The prevalence of Saharan dust in the atmosphere is a common feature of the Harmattan season in most parts of sub-Saharan West Africa. Fayiga et al. (2017) posited that the largest source of ambient PM in the world is that from the Sahel-Sahara desert. The volume of Saharan dust mobilized to the world’s atmosphere was estimated to range between 300 and 800 t each year (Garrison et al. 2014). The effect of this dust is most felt during the Harmattan season which normally spans between November and January. West African region is the most affected region because of its proximity to the Sahara desert. Recent studies have linked morbidity and mortality to exposure to PM emanating from Saharan dust (Adhvaryu et al. 2016; Goudie 2014; Jusot et al. 2017; Middleton 2017; Zhang et al. 2016). Inhaled dusts have a series of health effects as they affect the human respiratory and cardiovascular systems (Aubert et al. 2016; Chandrappa and Kulshrestha 2016; Okeahialam 2016). Ma et al. (2017) found a positive correlation between cases of measles and the incidence of elevated particulate (PMPM and TSP) concentration levels during dust events in western China. It was observed that excess measle cases were recorded during the storm periods as compared to non-dust storm periods. Adhvaryu et al. (2016) analysed data from 12 low-income countries in West Africa over two decades and posited that dust carried by Harmattan trade wind increased infant and maternal mortality.

Aside from the health effects of Harmattan dust, it is also associated with environmental impacts such as visibility impairment, soiling of photovoltaic systems and other solar energy radiating systems (Jiang et al. 2016; Sayyah et al. 2014). Energy loss is mostly prevalent in photovoltaic generating systems as dusty modules have significantly lower operation and voltage generation (Sayyah et al. 2014). Adhvaryu et al. (2016) observed that most of the plane crash incidents in Nigeria occurred during the Harmattan season. The study attributed it to poor visibility occasioned by the presence of high concentration of dust in the atmosphere. However due to scarcity of air pollution and health statistics in most part of Nigeria, assessment of health outcomes and statistics resulting from the annual haze episodes is limited. Monitoring of PM concentration levels is important to assist relevant stakeholders and concerned government agencies to ensure reduction in exposure and make necessary prevention strategies to curb future deleterious impact on human health. For the purpose of better understanding of the potential risk to health due to exposure to elevated levels of PM in the atmosphere, this study was conducted in Ilorin, a major urban settlement in North Central Nigeria.

Research methodology

Description of the study area

The study was carried out in Ilorin, the capital of Kwara State and one of the largest cities in Nigeria (Fig. 1). The city of Ilorin lies on longitude 8° 30′ N 4° 33′ E and latitude 8.500° N 4.550° E with daily average temperature ranges of 25 and 35 °C. The climate is equatorial with distinct wet and dry periods. With dry period (October–March) and wet period (April–September) each year. As of 2006 census, it had a population of 777,667, making it the sixth largest city in Nigeria by population. Land use patterns—residential, industrial and commercial—were considered in choosing the sampling locations in the metropolis (Table 1).

Fig. 1
figure 1

Map of Nigeria showing Ilorin with the study area inset

Full size image
Table 1 Sampling location and coordinates
Full size table

Meteorological data of the area were obtained from the University of Ilorin, Nigeria, and Lower Niger Basin Development Authority Meteorological stations.

Air sampling for PM

Heavy haze measurements were conducted on 5, 15, 18, 26 and 28 December 2016, while the light haze episode measurements were conducted on 3, 10, 11, 14, 17, 24, 30 and 31 December 2016. The visibility range considered for the light haze episodic days is 5 less than or equal to 10 km, while that of heavy haze episodic days is 0.5 less than or equal to 5 km. Particulate matter (PM) was measured with Aerocet 531S particle counter/dust monitor, an equipment from Met One Instruments. The Aerocet 531S is a full-featured, battery-operated, handheld mass monitor or particle counter. The unit simultaneously measures six mass concentration ranges (PM1, PM2.5, PM4, PM7, PM10 and TSP) or five particle count sizes (0.3, 0.5, 1.0, 5.0 and 10 μm). View sample history data in either mode. The particle counts from eight size ranges are converted to mass using a proprietary algorithm for typical density aerosols. Accommodation for special PM with different density is provided through user-programmable ‘K-factors’. The standard density value is adequate for most sampling environment. To measure, it was placed between 1 and 1.5 m above the ground level (breathing zone), switched in the environment of interest and the measured concentration read directly on the screen after particle capturing. Calibration tests were carried out using Hazdust™ sampler. Gravimetric method was employed in measuring filter weights before and after sampling. The obtained PM concentrations from Hazdust™ sampler were correlated against those obtained from the Aerocet 531S dust monitor. The details of the equipment calibration have been reported in our previous study (Adeniran et al., 2017b). Corrected values were calculated using equations obtained from regressing the measured concentration of Hazdust™ sampler against that of the Aerocet 531S. Zero count tests were carried out every 1 h using the zero filter kit supplied with the Aerocet 531S dust monitors. The measurements were carried out simultaneously in all the locations.

The concentration levels measured were averaged for 1 and 8 h. The extrapolated 24-h averaging concentration was obtained using an atmospheric stability formula (Al Smadi et al., 2009) given in Eq. 1:

$$ {C}_0=C\times F $$
(1)

where

C 0 :

the concentration at the averaging period t 0.

C :

the concentration at the averaging period t 1.

F :

factor to convert from the averaging period t 1 to the averaging period t 0 = (t 1 / t 0)n.

N :

0.28, the stability dependent exponent.

Spatial variation of PM

The spatial variation in the PM measured was plotted for the light and the heavy haze episodes. The PM concentration levels obtained were used as the input data for a grid-based contour map, to study the distribution of PM sizes of 1, 2.5 and 10 and the TSP in the residential, industrial and commercial areas of Ilorin metropolis for the light and heavy haze episodes. Surfer® 8.0, a mapping software from Golden Software, LLC, USA, was used for the PM spatial distribution mapping. Kriging gridding method was adopted for the interpolation of geographical data.

Human exposure assessment

Human exposure to airborne PM at the different sampling locations was assessed using the total respiratory deposition dose (TRDD) estimates and statutory limit breach (SLB) values of PM. The TRDD rates for PM10, PM2.5 and PM1.0 were estimated by using the mathematical relation given by Kumar and Goel (2016) as in Eq. 2:

$$ \mathrm{Deposited}\kern0.36em \mathrm{doses}\;\left(\mathrm{in}\kern0.17em \mathrm{thoracic},\mathrm{tracheobronchial},\mathrm{alveolar}\kern0.17em \mathrm{regions}\right)\kern0.24em \mathrm{of}\;\mathrm{PM}\;\mathrm{fractions}=\kern1.2em \left({V}_T\times f\right)\times {DF}_i\kern0.48em \times \kern0.36em {PM}_i\kern0.6em $$
(2)

where V T is the tidal volume, f is the frequency of breathing and DF i and PM i are the deposition fraction and mass concentration of particle size i in micrometre. Values of V T and f were taken as 800 cm3 per breath and 0.35, respectively, for men during light exercise conditions (Hinds 1999; Kumar and Goel 2016). Estimates of D fi for PM10, PM2.5 and PM1 were calculated according to the equation proposed by Hinds (1999) given in Eq. 3 and inhalable fraction (IF) as given in Eq. 4.

$$ DF= IF\times \left(0.058+\frac{0.911}{1+\exp\;\left(4.77+ In\;{d}_p\right)}+\frac{0.943}{1+\exp\;\left(0.508-2.58\; In\;{d}_p\right)}\right) $$
(3)
$$ \kern0.24em IF\kern0.36em =\kern0.36em 1-\kern0.36em 0.5\;\left(1-\frac{1}{1-0.00076\;{d}_p^{2.8}}\right)\kern0.36em $$
(4)

SLB value, expressed as the ratio of measured ambient PM concentration to the statutory limit of ambient concentration, was calculated for PM2.5, PM10 and TSP using Eq. 5.

$$ SLB\kern0.36em =\kern0.6em \frac{M_p}{S_p} $$
(5)

where M p is the measured mass concentration of PM and S P is the statutory limit of the PM. The equation was used to compute SLB value, taking into consideration the statutory limits of the US Environmental Protection Agency (USEPA 2012), Federal Ministry of Environment (FMENV) (FEPA 1991), World Bank (World Bank 1998) and that of the World Health Organization (WHO) (WHO 2006, 2010) (Table 2).

Table 2 Statutory limits of particulate matter
Full size table

Results and discussion

Meteorological parameters

The wind data obtained from the University of Ilorin and Lower Niger Basin Development Authority are presented in Fig. 2a, b using 1870 data sets for raw data and binned data. Wind speed ranged from 0 to 6.811 m/s with a mean and standard deviation value of 2.08 ± 1.14 m/s. As shown in Fig. 2a, the wind velocities were generally low and majorly towards southern direction. The wind speed and directions obtained in the study area are a result of the annual alternating southward and northward passages of the Intertropical Convergence Zone (ITCZ) (Sossa et al. 2017). Anuforum (2007) posited that the intertropical discontinuity (ITD) is displaced southwards during the Harmattan season. Regions around the sub-Saharan West Africa that lies between the Sahel and Gulf of Guinea are usually affected.

Fig. 2
figure 2

Wind rose for Ilorin metropolis during the sampling period

Full size image

The appearance of dominating weak southern atmospheric wind flow inhibits the dispersion of particulate matter in the atmosphere, and this will in turn affect their removal from the atmosphere (Chaloulakou et al., 2005). The descriptive statistics of other meteorological parameters such as the duration of sunshine hours per day, the relative humidity and temperature are presented in Table 3. During the heavy haze days, the high turbidity caused by the dusty air and low wind velocity had strong attenuations on the solar radiation as the period of visible solar radiation ranged from 0.3 to 3.5 h. For the light haze days, visible solar radiation ranged from 3 to 4.1 h. The relative humidity of the metropolis ranged between 28 and 57%, while daily temperature ranged from 15 to 36 °C. The early morning and late night periods are the coldest, while the midday periods were warmer as a result of solar irradiation. No rainfall data were recorded during the study.

Table 3 Descriptive statistics of the meteorological parameters
Full size table

Concentration levels of PM

The reduction in visibility and soiling of surfaces are as a result of high concentrations of PM of various diameters in the atmosphere that were observed. The measured mass concentration of PM (PM1.0, PM2.5, PM10 and TSP) is presented in Fig. 3a–d. PM1.0 concentration level ranged from 34.5 to 73.4 μg/m3; PM2.5 ranged from 103.8 to 562.7 μg/m3; PM10 ranged from 775.6 to 7066.3 μg/m3, and the TSP concentration ranged from 1151.7 to 9907.8 μg/m3. The mean and standard deviation values obtained for TSP in the residential areas were 6058.9 (± 364.8), 3133.5 (± 675.7) and 6624.9 (± 2012.6) μg/m3, respectively, for R1, R2 and R3. Similar values obtained for PM10, PM2.5 and PM1.0 are, respectively, 4681 (± 295.4), 510.9 (± 30.0) and 68.7 (± 2.6) μg/m3 for R1; 2090.4 (± 316.4), 232.5 (± 19.3) and 44.3 (± 1.9) μg/ m3 for R2; and 4673 (± 1242.6), 425.7 (± 32.9) and 60.5 (± 3.1) μg/m3 for R3. Descriptive statistics of the concentration levels were presented in the box plot for the PM sizes.

Fig. 3
figure 3figure 3

a Concentration levels of PM1.0. b Concentration levels of PM2.5. c Concentration levels of PM10. d Concentration levels of TSP

Full size image

The averaged daily and 8 hourly averaged concentrations of PM are presented in Tables 4 and 5, respectively. For the daily (24 h) averaged particulate concentration, the levels obtained were 14.17–30.14, 42.63–231.09, 318.53–2902.1 and 472.98–4069 μg/m3 for PM1.0, PM2.5, PM10 and TSP, respectively. For the 8 hourly averaged concentration, the values obtained are 19.27–41.00, 57.98–314.32, 433.25–3947.24 and 643.34–5534.50 μg/m3, respectively, for PM1.0, PM2.5, PM10 and TSP. The highest PM concentrations were obtained in R3 which is a student residential area with a lot of commercial activities. Aside from the Harmattan bound dust, other contributions to the high concentration of PM obtained include those from vehicular sources, biomass burning used in food preparation and dust resuspension from people walking. Previous study has identified heavy vehicular density in the city as a major source of airborne PM (Adeniran et al. 2017b). Total PM concentration obtained during the wet season in a previous study (Adeniran et al. 2017a) is about 11% of the TPMC obtained in the present study for the same location. The use of back-up generators in commercial buildings, roadside shops and those used in mobile telecommunications base transceiver stations also contribute to elevated airborne PM concentration levels (Adeniran et al., 2017).

Table 4 Twenty-four hourly averaged concentration of particulate matter
Full size table
Table 5 Eight hourly averaged concentration of particulate matter
Full size table

The sample diurnal variation of PM concentration levels are shown in Fig. 4. The early morning and the late night periods are generally characterized by higher concentrations. This is because the wind is usually calmer in the night and there is no solar irradiation. Lower temperature observed in the night periods also contributes to the concentration levels obtained. Cooler air is usually above making it difficult for pollutants to diffuse due to the difference in density (Janhäll et al., 2006; Sæbø et al., 2017). This phenomenon is termed temperature inversion.

Fig. 4
figure 4

Sample diurnal variation of particulate matter for hazy days

Full size image

The ratio of ultrafine particles (PM1.0) to the respirable fraction (PM2.5) to the inhalable fraction (PM10), the respirable fraction (PM2.5) to the inhalable fraction (PM10) and that of the inhalable fraction to the total suspended particles was considered because of the air pollution fractions and human exposure implications.

Figure 5 shows the calculated PM1.0/PM2.5, PM2.5/PM10 and PM10/TSP for PM levels measured. PM1.0/PM2.5 ratio ranged from 0.13 to 0.35; PM2.5/PM10 ratio ranged from 0.09 to 0.13, while PM10/TSP ratio ranged from 0.57 to 0.77. There are high percentages of coarse particles in TSP. This further corroborates the fact that concentration increases with increasing particle size due to agglomeration, accretion, coagulation and secondary formation processes. The high ratios of PM10/TSP suggest that crustal materials and soil resuspension have great influence on particulate levels during dry climatic conditions.

Fig. 5
figure 5

Ratios of particulate matter measured

Full size image

Effect of climatic parameters on haze pollution

Pearson’s correlation was done using the Predictive and Analytical Software (PASW version 18). The result of Pearson’s correlation to test relationships among the measured PM concentration levels and microclimatic parameters is presented in Table 6.

Table 6 Correlation coefficient matrix for particulates and microclimatic parameters
Full size table

Strong relationships exist among all the PM sizes as the correlation coefficient obtained among them varied from 0.8 to 1.0. This indicates that lower sizes are fractions of the higher diameter particles as earlier posited. Very strong negative correlations exist between the PM concentration levels and microclimatic parameters. The obtained correlation signifies that low levels (low wind speed, low temperature, low relative humidity, less hours of sunshine) have strong impact on increased levels of airborne PM. This correlation pattern is similar to those obtained by Chaloulakou et al., 2005 who observed that strong negative correlations exist among PM and climatic parameters such as no rainfall and calm wind speed. The prevalence of stagnant or weak wind flow regimes will favour the continued suspension and accumulation of particles brought by the trade winds and those emitted locally. This will further result in the elevation of ambient PM levels. As expected, strong correlation existed between the number of sunshine hours and temperature. Wind speed and relative humidity are highly correlated with a coefficient of 0.943. The atmosphere was dry, and wind velocity was relatively calm. This condition makes airborne PM to be stable in the atmosphere.

Spatial variation of PM

The PM concentration level spatial variation contours obtained using kriging interpolation method in Surfer® v8 are presented in Figs. 6, 7, 8 and 9 for the light and heavy haze episodes. The visibility range considered for the light haze episodic days is 5 less than or equal to 10 km, while that of heavy haze episodic days is 0.5 less than or equal to 5 km. In general, several hotspots of high particulate concentrations were identified in the eight maps for PM1.0, PM2.5, PM10 and TSP.

Fig. 6
figure 6

a Spatial variation of TSP for light haze episode. b Spatial variation of TSP for heavy haze episode

Full size image
Fig. 7
figure 7

a Spatial variation of PM10 for light haze episode. b Spatial variation of PM10 for heavy haze episode

Full size image
Fig. 8
figure 8

a Spatial variation of PM2.5 for light haze episode. b Spatial variation of PM2.5 for heavy haze episode

Full size image
Fig. 9
figure 9

a Spatial variation of PM1.0 for light haze episode. b Spatial variation of PM1.0 for heavy haze episode

Full size image

The observed spatial variation trend observed in the metropolis further conforms with the position of Anuforom (2007) who posited that dust haze depends on latitude. The concentration increases from south to north. However, antropogenic emissions such as those from smoke stacks in the industrial region, vehicular movement, fuel combustion and other activities have influence on the particulate level on spatial scale. Some areas also fall in between mountains where dispersion of PM is further impeded by the weak flow of wind. PM especially the fine fractions (PM1.0 and PM2.5) tend to stay airborne for a longer period.

Exposure assessment of PM

Exposure assessment of the measured PM was done using the SLB value (comparison with the regulatory limit values) and total respiratory deposition dose.

Statutory limit breach values

In light of health concerns, several authorities have set exposure standards or guidelines for PM. Environmental Protection Agency (EPA) set a 24-h threshold limit value (24-h TLV) of 35 and 150 μg/m3 for PM2.5 and PM10, respectively, while World Health Organization (WHO) set a 24-h TLV of 25 and 50 μg/m3 for PM2.5 and PM10, respectively. For TSP, a 24-h TLV of 250 and 80 μg/m3 was set by the Federal Ministry of Environment (now the National Environmental Standards and Regulations Enforcement Agency (NESREA)) and the World Bank, respectively. A SLB value, calculated using the threshold limits, that exceeds 1 is worrisome. SLB values of the heavy and light haze episodes are presented in Tables 7 and 8.

Table 7 Average SLB values for heavy haze episodes
Full size table
Table 8 Average SLB values for light haze episodes
Full size table

For the heavy haze episodes, the SLB values with reference to WHO TLV of 25 μg/m3 for PM2.5 ranged between 2.4 and 9.2 with an average and standard deviation of 5.1 (± 2.4) while similar value when compared with the USEPA TLV of 35 μg/m3 was 1.7–6.6 with an average and standard deviation of 3.7 (± 1.7). USEPA SLB values for PM10 were 3.9–19.3 with an average and standard deviation of 9.4 (± 4.9). WHO SLB values for PM10 were 11.7–58.0 with an average and standard deviation of 28.2 (± 14.7). The PM concentration levels obtained also exceeded the TSP TLV of 80 μg/m3 for World Bank in the multiples of 12.0–50.9 with an average and standard deviation of 27.1 (± 12.9). The Nigerian FEPA TLV of 250 μg/m3 for TSP was breached in the folds of 3.8–16.3 with an average and standard deviation of 8.7 (± 4.1).

For the light haze episodes, the SLB values with reference to WHO TLV of 25 μg/m3 for PM2.5 ranged between 1.7 and 8.0 with an average and standard deviation of 4.0 (± 2.1) while similar value when compared with the USEPA TLV of 35 μg/m3 was 1.2–5.7 with an average and standard deviation of 2.8 (± 1.5). USEPA SLB values for PM10 were 2.1–11.9 with an average and standard deviation of 5.6 (± 3.3). WHO SLB values for PM10 were 6.4–35.8 with an average and standard deviation of 16.7 (± 10.0). The PM concentration levels obtained also exceeded the TSP TLV of 80 μg/m3 for World Bank in the multiples of 5.90–28.8 with an average and standard deviation of 14.2 (± 7.8). The Nigerian FEPA TLV of 250 μg/m3 for TSP was breached in the folds of 1.9–9.2 with an average and standard deviation of 4.5 (± 2.5).

Total respiratory deposition dose

The average TRDD values are presented in Fig. 10 for PM10, PM2.5 and PM1.0. The TRDD for PM1.0 were 9.68 ± 0.37, 6.29 ± 0.28, 8.53 ± 0.43, 8.78 ± 0.59, 8.26 ± 0.38, 5.97 ± 0.46, 6.95 ± 1.18, 6.29 ± 0.45, 6.48 ± 1.04 and 6.08 ± 1.52 μg/h for locations R1, R2, R3, I1, I2, I3, C1, C2, C3 and C4, respectively. The highest TRDD rate was obtained at R1, and it was about 1.63 times of the value obtained for I3. The TRDD value for PM2.5 at R1 was 4.20 times higher than those for C4. TRDD values obtained for PM2.5 were 149.8 ± 8.8, 69.1 ± 5.9, 124.8 ± 9.6, 116.2 ± 15.5, 109.0 ± 4.8, 57.2 ± 3.5, 53.8 ± 6.4, 37.7 ± 2.90, 52.4 ± 4.4 and 35.7 ± 4.6 μg/h for locations R1, R2, R3, I1, I2, I3, C1, C2, C3 and C4, respectively. TRDD values estimated for PM10 were 1329.1 ± 83.9, 594.5 ± 98.8, 1326.6 ± 352.7, 1080.9 ± 194.9, 935.6 ± 67.8, 620.8 ± 194.7, 392.9 ± 32.9, 279.7 ± 62.8, 474.3 ± 169.4 and 302.5 ± 51.6 μg/h for locations R1, R2, R3, I1, I2, I3, C1, C2, C3 and C4, respectively. As obtained for PM1.0 and PM2.5, the highest TRDD value for PM10 was also obtained at R1; the value obtained is 4.75 times higher than the value obtained at C2. The TRDD values obtained were on hourly basis. Multiples of the dose may be inhaled when exposed for longer hours. Road side users, market women and people who work for long hours are at greater risk. Respiratory-related ailments are often common during the hazy days. Epidemiological studies have shown that PM sizes with aerodynamic diameter less than 10 μm (PM10) have adverse effect on the respiratory system as they can reach the conductive airways.

Fig. 10
figure 10

a TRDD levels of PM10. b TRDD levels of PM2.5. c TRDD levels of PM1.0

Full size image

Conclusion

Mass concentrations of PM (PM1.0, PM2.5, PM10 and TSP) measured during the December 2016 haze episodes in 10 locations spatially distributed within Ilorin metropolis, Nigeria, reached levels considered hazardous to public health. PM1.0 concentration level ranged from 34.5 to 73.4 μg/m3; PM2.5 concentration levels were between 103.8 and 562.7 μg/m3; PM10 ranged from 775.6 to 7066.3 μg/m3, and the TSP concentration ranged from 1151.7 to 9907.8 μg/m3. The appearance of dominating weak southern atmospheric wind flow observed inhibited the dispersion of PM. Pearson’s correlation signifies that low levels of climatic parameters (low wind speed, low temperature, low relative humidity, less hours of sunshine) have strong impact on increased levels of airborne PM. The observed spatial variation trend observed in the metropolis indicated that PM concentration in the atmosphere increases with latitude. The concentration increases from south to north. SLB values obtained show that the averaged concentration of all the PM sizes exceeded the statutory limits set by regulatory agencies for PM2.5, PM10 and TSP. The highest hourly TRDD rates for PM1.0, PM2.5 and PM10 were obtained at R1, while the lowest were found in I3 for PM1.0, C4 for PM2.5 and C2 for PM10. In general, particulate air pollution measured reached levels considered hazardous to public health. The present study has served as an indicator to the likely adverse consequences of elevated dose of PM during annual haze events. Public and government attention should be drawn to the inherent dangers in repeated and localized exposures. Dissemination of real-time air quality information should be made available to the public as this will contribute to increase in transparency and environmental awareness.