Ambient Particulate Matter Air Pollution in Mpererwe District, Kampala, Uganda: A Pilot Study

Air quality in Kampala, the capital of Uganda, has deteriorated significantly in the past two decades. We made spot measurements in Mpererwe district for airborne particulate matter PM2.5 (fine particles) and coarse particles. PM was collected on Teflon-membrane filters and analyzed for mass, 51 elements, 3 anions, and 5 cations. Both fine and coarse particle concentrations were above 100 µg/m3 in all the samples collected. Markers for crustal/soil (e.g., Si and Al) were the most abundant in the PM2.5 fraction, followed by primary combustion products from biomass burning and incinerator emissions (e.g., K and Cl). Over 90% of the measured PM2.5 mass can be explained by crustal species (41% and 59%) and carbonaceous aerosol (33%–55%). Crustal elements dominated the coarse particles collected from Kampala. The results of this pilot study are indicative of unhealthy air and suggest that exposure to ambient air in Kampala may increase the burden of environmentally induced cardiovascular, metabolic, and respiratory diseases including infections. Greater awareness and more extensive research are required to confirm our findings, to identify personal exposure and pollution sources, and to develop air quality management plans and policies to protect public health.


Introduction
Air pollution is an increasing problem in urban dwellings worldwide. The World Health Organization (WHO) ranks indoor air pollution (from biomass fuel combustion) and urban outdoor air pollution 10th and 14th, respectively, among 19 leading risk factors for global mortality [1]. There is indisputable epidemiologic evidence that air pollution exposure in humans increases not only the risk of respiratory conditions such as asthma and respiratory tract infections but also cardiovascular disease and metabolic diseases such as diabetes mellitus [2,3]. In the developing world, air pollution presents a particular threat to the health of populations in cities that are growing in a fast, poorly planned, and unregulated manner [4,5]. This is the case in Kampala, Uganda, where air quality has deteriorated significantly during the past two decades in parallel with the fast-paced economic development. Kampala is the capital of Uganda in East Africa, a country with a rapidly increasing population and expanding economic growth. With a growing middle class, Uganda, and specifically Kampala, has seen fast expanding vehicular traffic comprised of imported second hand and reconditioned cars, aging, exhaust-unregulated cars, trucks, buses, and motorcycles. Furthermore, the national economic strategy gives priority to the industrial and manufacturing sectors and has promoted and realized a rise in cottage and other larger industries within the city. In addition to the ever growing road traffic and unregulated industrial emissions, waste burning by individuals to manage uncollected waste, also contributes to the deterioration of air quality in Kampala.
However, reports on air pollution in Kampala are primarily anecdotal [6][7][8]]. An ongoing study shows an increase of ischemic heart disease from virtually absent up to the end of the 1990s to more than 10% of all admissions on the Mulago University Hospital cardiology ward in 2011. As part of an effort to preliminarily characterize airborne particulate matter (PM) within Kampala, we made spot measurements of ambient PM in a typical bustling, low-income, mixed commercial, and residential section of the city (Mpererwe District). This paper provides a summary of the findings.

Methods
To characterize the content of air pollution PM, we made spot measurements in Mpererwe District, Kawempe South Division, at Kalerwe Market, 80 meters east from Gayaza Kampala Road in Kampala City (Figure 1-Map). Employing a Sioutas Personal Cascade Sampler (SKC Inc., Eighty Four, PA, USA) with a flow rate of 9 liters per minute, we collected fine particles (PM 2.5 , or particles with aerodynamic diameters less than 2.5 m) and coarse particles (PM 7-2.5 ) (this fraction contains total suspended particles with aerodynamic diameters greater than 2.5 m) during two 24-hour sampling periods (6:30 p.m. to 6:30 p.m., Local Standard Time [LST]) between December 31st, 2012, and January 1st, 2013, and between January 2nd, 2013, and January 3rd, 2013. PM were collected on Teflon-membrane filters (37 mm, 2 mm pore, Pallflex Gelman Scientific, Ann Arbor, MI, USA). The personal PM monitor as well as a Leland Legacy Pump (SKC Inc., Eighty Four, PA, USA) were installed 3 meters above ground level, on the outside wall of a single story structure under a 50 cm protruding metal sheet roof. PM 2.5 and PM coarse samples were removed from the monitors at the end of each sampling period, stored in plastic petri dishes, sealed and transferred to the USA for further analysis. Filters were equilibrated under a temperature (21.5 ± 1.5 ∘ C) and relative humidity (35 ± 5%) controlled environment before gravimetric analysis following the US Environmental Protection Agency (EPA) protocol at the Environmental and Occupational Health Sciences Institute (EOHSI) in New Jersey. PM 2.5 samples were submitted for elemental analysis by X-ray fluorescence (XRF) including 51 elements from Na to U (see Table 1 and [10]). Three anions (Cl − , NO 3 − , and SO 4 2− ) and five cations (NH + , Na + , Mg 2+ , Ca 2+ , and K + ) were measured for PM 2.5 by ion chromatography (IC) [11]. Particle loadings from the personal PM monitor are centered on a slit, so only 18 elements were analyzed by inductively coupled plasma-mass spectrometry (ICP-MS).

Results
Based on the PM mass, fine and coarse particle concentrations were 104.9 g/m 3 and 132.7 g/m 3 for the sample from December 31, 2012 and 103.7 g/m 3 and 208.1 g/m 3 for the sample from January 2, 2013, respectively. PM 2.5 mass concentrations and chemical compositions are listed in Table 2, which includes typical marker species representing major PM 2.5 sources, such as soil (Al, Si, Ca, Fe, and Ti), oil combustion (V and Ni), biomass combustion (K + ), traffic (Pb and Zn), incinerator (Cl), and secondary PM (SO 4 2− ). A full list of all measured PM 2.5 species is in Table 1. The relative abundance of major PM 2.5 species is similar across the two sampling days. Markers for crustal/soil are the most abundant, followed by primary combustion sources, such as biomass burning and incinerator emissions. Elevated water-soluble to total potassium ratio (K + /K = 0.91) for both PM 2.5 samples confirms an abundance of biomass burning during the sampling period [12]. An exploratory PM 2.5 mass reconstruction was conducted based on major species measured, that is, SO 4 2− , NO 3 − , Al, Si, Ca, Fe, and Ti, and closing the mass balance by assuming the remaining mass was contributed by carbon species [13]. Figure 2 illustrates the percent contributions of SO 4 2− , NO 3 − , crustal sources, and carbon. Over 90% of measured PM 2.5 mass can be explained by crustal species and carbonaceous aerosol. Table 3 shows the elemental concentrations measured for PM coarse samples. Abundant (>200 ng/m 3 ) crustal species (e.g., Al, Ca, and Fe) are found with large sample-to-sample variations.
The meteorological conditions during the sampling period are presented in Table 4. Mean temperatures across the four sampling dates ranged from 21.7 ∘ C to 24.4 ∘ C, with dew point temperatures between 18.9 and 20.0 ∘ C and mean wind speeds between 6.4 and 8.1 Km/hr.

Discussion
To the best of our knowledge, there is no known data/sample collection and public information available on type and concentrations of particulate matter in Kampala's air and studies of the associated human health impacts are extremely limited. There is clear evidence that air pollution in Kampala has increased in the last decade ( Figure 3) and that awareness about this problem is mounting [6][7][8]. The limited findings of the current study clearly support these observations and indicate that exposure to ambient air in Kampala may raise public health concerns such as the burden of cardiovascular and respiratory diseases. PM 2.5 mass concentrations observed     [16]. In a measurement campaign, Chow et al. (2002) reported that the mean and the maximum PM 2.5 mass concentrations in Mexico City were 35.2 and 183.7 g/m 3 , respectively [17]. It should be pointed out that the average concentrations from world cities mentioned above and in  Figure 2: Twenty-four-hour average PM 2.5 compositions of major components based on mass reconstruction, where fine soil = 2.2Al + 2.49Si + 1.63Ca + 2.42Fe + 1.94Ti [9]. Ammonium (NH 4 + ) concentrations are low, so only sulfate and nitrate are included. The remaining PM 2.5 mass is assumed to be carbon. the literature were the monthly, seasonal, or annual average of measured daily PM 2.5 concentrations. Therefore, caution needs to be exercised when comparing our results and the data presented in the literature.
Coarse particles collected in Kampala were primarily composed of crustal elements (i.e., Al, Ca, and Fe), indicating soil dust as the major source. Unlike PM 2.5 , coarse particles are formed in the atmosphere through mechanical forces, for example, the resuspension of road dust. Indeed, many unpaved road sections around the sampling area were observed by the authors. Other activities, such as agriculture, can also contribute to airborne coarse particles. These findings warrant further studies to examine the air pollutant source contributions.
Various anthropogenic sources appear to contribute to the elevated PM 2.5 levels in Kampala. In 2011, the Uganda Revenue Authority (URA) estimated there were 635,656 vehicles in Uganda and about 50% of those were operating in greater Kampala [18]. Between 4,000 and 5,000 vehicles enter the country monthly from various countries [8,19]. These vehicles are mostly second-hand and purchased reconditioned from Japan. To date, there are no age restrictions and requirements for emission restrictions in place that exist in industrialized countries [20,21]. According to the Uganda National Environment Management Authority (NEMA), all the resulting daily traffic jams result not only in loss of valuable health and time but also lead to a loss of about 500 6 Journal of Environmental and Public Health million Uganda Shillings (about US $200,000) in burnt fuel; that is, about 140,000 liters of fuel are burnt by idling cars every day [18].
Furthermore, the Bank of Uganda estimates that by 2003 there were 800,000 small scale enterprises in the entire country of Uganda, providing employment and income generation opportunities to low income sectors of the economy, but at the same time, potentially contributing to air pollution. In addition, there is no mandatory garbage collecting system in Kampala, Uganda. The majority of inhabitants has been and continues to simply burn their waste which includes a lot of plastics and results in toxic smoke.
The increased source emissions of PM 2.5 may result from the rapid urbanization. The National Development Plan 2010-2015 estimates that Uganda's population could double to nearly 61 million in about 22 years [22]. The city has spread from the original 7 hills it sat on at independence in 1962, to approximately 50 now. City planners at the time planned urban infrastructure for approximately 500,000 residents. Today, Kampala has approximately 1,819,200 inhabitants (extrapolated from a population of 1,420,200 in 2008 and a population of 1,659,600 on June 16th 2011) and an additional estimated 2.5 million commuters from outside populating the roads of the city every day. Ever since independence no major infrastructural changes were initiated to accommodate this increase. The City's governing bodies are unable to cope with this massive population influx and the resulting consequences such as overcrowding, uncollected garbage, increased individual motor-vehicle traffic in absence of an organized public transport system, mostly on unpaved roads and uncontrolled informal markets, and small industries.
Given the evidence of abundant vehicle engine emissions and biomass burning in the PM samples from this study, future studies will have to include analysis of organic and elemental carbon as well as carbon speciation to further understand the source attribution and its potential health impacts. The current study is limited by a small sample size, limited PM speciation, assessment of PM at one location only, and by lack of seasonal assessments. The results of this pilot study are indicative of unhealthy air in Kampala and suggest that a large number of people may be threatened by cerebro-and cardiovascular as well as respiratory tract diseases and even diabetes mellitus resulting from air pollution exposure. Greater awareness and more extensive research on air pollution in Kampala are required to confirm our findings, to identify pollution sources that contribute to personal exposure, and to develop air quality management plans and policies to improve public health.

Disclosure
Melody Corry, who has M.S. and cand. Ph.D. degrees deceased five weeks before manuscript submission.