Monitoring of air quality before the Olympic Games Rio 2016

For the accomplishment of Olympic and Paralympic Games in Rio de Janeiro city (Rio 2016), the government of the Rio de Janeiro state has undertaken to monitor air quality before and during the events. In Beijing, China, and Athens, Greece, the air quality was monitored in Olympic venues in order to evaluate the athletes’ performance in relation to the environment in which they were exposed. This study has the same proposal to Rio de Janeiro, Brazil. The air quality scenario of the three previous years (2013, 2014, 2015) of Rio 2016 was considered. Coarse (PM10) and fine (PM2.5) particles and O3 were monitored continuously on the stations located near to competition venues, as required by International Olympic Committee (IOC). The levels registered ranged from 6 to 96 μg m-3 for PM10, 1 to 44 μg m-3 for PM2.5 and 121 to 269 μg m-3 for O3. These concentrations exceeded the national and international air quality standards. These high concentrations are associated with uncountable civil works to build Olympic arenas and the urban mobility ́s improvement. however, the concentrations for all the pollutants monitored in Rio de Janeiro city were in lower concentrations than in Beijing Olympic Games 2008.


INTRODUCTION
The Olympic and Paralympic Games were hosted in Rio de Janeiro city, between August 5 th and September 18 th of 2016. Rio de Janeiro is one of the 30 biggest megacities in the world, being the second biggest city in Brazil and third in South America. This city has about six million inhabitants and a population density of 5.265 hab km -2 (IBGe 2012). It has a fleet of 2.6 million vehicles (DeTRAN-RJ 2014) -almost one vehicle for every three inhabitants -and the second highest participation in the Gross Domestic Product (GDP) of the country (IBGe 2012).
With very similar characteristics to Athens, Greece, Rio de Janeiro city is also surrounded by the sea and mountains, which have an enormous impact in the pollutants atmospheric dispersion and, consequently, in the air quality. Although Rio de Janeiro is a coastal city, which helps the natural ventilation, its steep topography, parallel to the cost line, makes the air circulation more difficult (Soluri et al. 2007, Godoy et al. 2009). Another factor that influences the city's air quality is the irregular land-use. The city's climate is subtropical, with intense solar radiation and high temperatures, which favors the formation of secondary photochemical pollutants.
Rio de Janeiro city has undergone many changes, mainly in infrastructure and urban mobility, due to the World Cup and Olympic Games. The infrastructure works throughout the city began in 2008. Works included expansion of the international airport (2008) and hotel network (2010), maracanã stadium reform (2010), construction of subway (line 4) (2010) and exclusive bus routes (BRT) (2011) (Ventura et al. , 2017b. At each new choice of the Olympic Games' host cities, the demands imposed by the International Olympic Committee (IOC) about the monitoring air quality have increased. The chosen city must monitor air quality at least three years before the Olympics and inform the population the air quality indexes. The pollutant levels must be within the national standards during the Games to avoid influences on the athletes' performance. It is due to new studies about air quality and health (Brook et al. 2010, Rundell 2012. One of these studies was performed by the American heart Association (AhA), which advised against the practice of intense exercises in the presence of unsatisfactory air quality, even though the sportive medicine societies do not recommend cares with the practice of exercises in polluted places (Brook et al. 2010). Rundell (2012) evaluated the effect of pollution in athletes' performance and concluded that the practice of physical exercises in environments with high levels of pollutants reduces sharply the pulmonary and vascular function both in asthmatic individuals and healthy ones. Therefore, the practice of exercises in long-term in polluted places is associated to the reduction of pulmonary function, being able to induce vascular dysfunction, due to the systemic oxidative stress and in the nasal passages, and, consequently, reduce the athletes' performance (Rundell 2012).
Particulate matter (Pm) and O 3 are between the six pollutants (CO, Pb, SO 2 , NO 2 , and O 3 ) used to assess air quality and avoid damage to health and ecosystems (U.S.ePA 2012). Pm is considered the main pollutant involved in respiratory and cardiac problems (Dominici et al. 2006). Ozone can also have a negative impact on public health when present in the higher and moderate levels in the atmosphere (luna et al. 2014). In high concentrations, these pollutants can reduce the athletes' physical capacity and can cause discomfort for the population (Brook et al. 2010). Therefore, monitoring the air quality is essential to the athlete's performance and population health.
In 2011 Ventura et al. 2017b). The Olympic Stations and the competition venues can be visualized in Figure 1.
The goal of this study is to evaluate the results obtained by INeA and IOC regarding the air quality monitoring in Rio de Janeiro city 3 years before the Olympic and Paralympic Games (Rio 2016), in order to infer about health effects.

MATERIALS AND METHODS
In order to execute the monitoring of air quality in the Olympic and Paralympic Games arenas, INeA standardized measurement to determine the pollutant levels (Pm 10 , Pm 2.5 and O 3 ). The air quality monitoring stations were distributed in four sportive regions: Recreio dos Bandeirantes and Jacarepaguá (Barra da Tijuca), Gericinó and Campo dos Afonsos (Deodoro), lagoa, leblon and Urca (Copacabana), as well as maracanã,  analysis. The filter mass was weighed before and after collection on an analytical balance (mettler e., Zürich. Switzerland, ± 0.0001 g), as described in the Brazilian Technical Standard (NBR 13412) method.
To better understand some air quality levels, it was shown in Table SII the meteorological data of each Olympic stations from 2013-2015, such as temperature (T), relative humidity (Rh), wind speed (WS) and direction (WD). The meteorological data used in the statistical analyzes were averages of 24 h. Besides, it was showed the preferential wind direction by Wind Roses ( Figure S1).

STATISTICAl ANAlYSeS
To verify the correlation between the pollutants measured by the Olympic station and meteorological data, the correlation matrix was used. The Pearson correlation coefficient quantifies the association between two variables. When the correlation coefficient is greater than 0.3 there is correlation between the variables. This technique was applied to the data of Pm 10 , Pm 2.5 , O 3 , T, Rh and WS, for the average daily data for criteria pollutants concentrations and meteorological parameters, except for ozone which adopted maximum daily concentration. Besides, boxplot was used to represent the distribution of the Pm 10 and Pm 2.5 concentrations for each sport regions from 2013 to 2015.

RESULTS AND DISCUSSION
In Brazil, the air quality standards are set by the National environment Council through Resolution n. 03, published in 1990 (CONAmA 1990). The annual mean Pm 10 concentration standard is 50 μg m -3 and the daily concentration standard is 150 μg m -3 and for O 3 the hourly concentration standard is 160 μg m -3 . Brazilian legislation has not been updated since its creation in the 1990s. Presently, this resolution is being reviewed, and the limits for Pm 2.5 will be included. Although the Brazilian legislation has not adopted daily and annual air quality standards to Pm 2.5 yet, this pollutant has been monitored for semiautomatic stations in Rio de Janeiro state since 2011 (Ventura et al. 2017b). This study used the air ambient standards (AAS) set by U.S.ePA to evaluate the results because these standards include health criteria and also socioeconomic aspects. For the short-period exposure, it was adopted the daily standard of 35 µg m -3 and for long period, the annual standard of 12 µg m -3 .

PARTICUlATe mATTeR
The annual average Pm 2.5 concentrations measured from 2013 to 2015 in Rio de Janeiro city ranged from 6 to 20 µg m -3 . These levels were lower than the ones registered in Beijing, China, during the Olympic Games (46.7 µg m -3 ) and Pre-Olympic period ( The sampling sites located close to urban roads are influenced by heavy traffic (Table  II). Urca's station was the sampling point with less vehicular influence while lagoa was more impacted. The annual average Pm 2.5 concentrations observed at maracanã ranged from 6 to 13 µg m -3 , which are similar to the previous study from 2003 to 2005 (9.8 to 11 µg m -3 ) (Soluri et al. 2007, Godoy et al. 2009). On the other hand, previous studies performed in Recreio dos Bandeirantes registered lower concentrations of Pm 2.5 (7.7 µg m -3 ) (Soluri et al. 2007, Godoy et al. 2009) than in this one (11 to 16 µg m -3 ). Since 2008, this site has been received more influence of heavy traffic due to the construction of an expressway for bus circulation (BRT-TransOeste) and constructions lUCIANA m.B. VeNTURA et al.

AIR QUAlITY RIO 2016
An Acad Bras Cienc (2019) 91(1) e20170984 5 | 14 of Arenas (Sport Parks) and Villages' Athletics for the Olympic Games (Ventura et al. , 2017b, resulting in high levels of pollutants. In 2014, the annual average Pm 2.5 concentrations were the highest registered in all sites. For this year, lagoa, engenho de Dentro, maracanã and Recreio dos Bandeirantes levels exceeded the annual air quality standard adopted by U.S.EPA (12 μg m -3 ). On the other hand, the levels for all sites decreased in 2015. These data show the influence of the civil works in the air quality during 2014. By 2015, the works were almost completed, therefore the levels decreased. however, the annual average Pm 2.5 concentrations in the last three years did not exceed the U.S.ePA AAS.
Pm 10 annual averages for all sampling sites are present in Table III. In Campo dos Afonsos, Gericinó, leblon sites were registered annual average concentration below the limit (50 μg m -3 )    (Table IV). The negative value means that these variables are inversely proportional, i.e., when the temperature increases an expansion in air volume occurs, resulting in a decrease in relative humidity.
During the three years of monitoring before the Olympic Games in Copacabana, maracanã and Deodoro regions, the daily Pm 10 concentrations varied from 6 to 96 µg m -3 , which are lower than Brazilian and U.S.ePA guidelines (150 µg m -3 ). In the same period, daily Pm 2.5 concentrations varied from 1 to 44 µg m -3 in Barra da Tijuca and Copacabana regions. In the maracanã region, on the other hand, the concentrations varied from 2.5 to 87 µg m -3 . These results enhance the influence of the sea breeze in this pollutant dispersion, as well as the characteristic of the emission source in the concentration levels (vehicles, industries, marine aerosols, and others). In these sites, the levels exceeded the U.S.ePA AAS (35 µg m -3 ). Copacabana region is close to the sea and Guanabara Bay, which favors air entrance in the continent and consequently, atmospheric pollutants dispersion (Ventura et al. 2017a). For this reason, lowest monthly mean concentrations for both Pm 10 and Pm 2.5 were observed.
From 2013 to 2015, an increase in the annual average Pm 10 concentrations was verified in all sampling sites (Table III) monthly data of particulate matter are a good indicator of atmospheric pollution, not only because it shows a direct standard measure for the concentration of dust in the air, but also because of direct relation to the human health (leys et al. 2011). Pm 10 seasonality was also studied by Gioda et al. (2016), who had already verified through the time series analysis of this pollutant in Rio de Janeiro city, that there was a trend to increase during the austral wintertime (June 21 to September 21). This trend was related to low rainfall and air mass stagnation. Therefore, the Olympics occurred during the period with higher pollutant levels. This is worrying because, during the exercise practice, the quantity of fine particles deposited in the respiratory tract is inversely proportional to its size. In other words, the smaller the particles, the bigger the quantity deposited, which during a physical practice can come to be five times superior to the ones registered when the body is resting (Daigle et al. 2003). Figure S2 presents box plots for the distribution of Pm 10 and Pm 2.5 concentrations for each sport regions from 2013 to 2015. It was observed that the Pm 10 data set was very similar. however, in maracanã Pm 2.5 concentrations showed a large variation in its data set, due some days in this region lUCIANA m.B. VeNTURA et al.
In maracanã and Copacabana region were monitored Pm 2.5 and Pm 10 . When it was applied Pearson Correlation matrix between these pollutants (Table IV), it was observed high correlations (> 0.7). According to Silva et al. (2016), this result indicates that these data are highly related and most likely have the same emission sources and/or similarity in formation and consumption processes. Figures 2 and 3 presented monthly average Pm 10 and Pm 2.5 distributed in the sportive regions to verify the particulate matter behavior on the different seasons, especially in winter, the period of the Olympics. According to the t-student test with 95% confidence level, the highest concentrations were registered from may to September, as expected, due to the unfavorable conditions to the atmospheric pollutants' dispersion. On the other hand, lower concentrations were measured in November and December, during the rainy season. In Beijing, as well as in Rio de Janeiro city, the winter was also the season less prone to the pollutant's dispersion, resulting in average Pm 2.  In all sport regions, ozone showed a positive correlation with Pm 2.5 and Pm 10 (Table IV). According to martins et al. (2015), although Pm does not directly participate in the process of ozone formation, environments with higher Pm concentrations block solar radiation due to light scattering and decrease the photochemical dissociation of NO 2 . This photochemical dissociation of NO 2 initiates the process of ozone formation by forming atomic oxygen.
Deodoro region presented the highest hourly ozone concentrations compared to other study regions (Table V). This can be explained for the high concentrations of the precursor gases emitted by vehicles in this area, within more than 10 ton NOx and 2 ton VOCs per year (INeA 2016b). moreover, in this region, a big mountain prevents the pollutant dispersion, as it was also observed by martins et al. (2015), Tsuruta et al. (2017) and Geraldino et al. (2017) in recent studies in west region in Rio de Janeiro city. martins et al. (2015) evaluated three stations (Bangu, Irajá and Campo Grande) in west region in Rio de Janeiro city, located along Avenida Brasil, which is the main traffic route, with more than 50,000 vehicles per day (INeA 2016b). This highway is the same that passes near the air quality station in Deodoro. The researchers concluded that these sites were influenced by vehicle emissions, with no significant influences from stationary sources.
In Bangu station, neighborhood Deodoro region was observed by Silva et al. (2016) the highest ozone concentrations and the largest number of air quality standard exceedances in Rio de Janeiro city in 2012 and 2013. According to Tsuruta et al. (2017), it is justified because this region is characterized by higher temperatures and poor air circulation in comparison to the other regions of the city. Geraldino et al. (2017), who also studied Bangu, concluded that the mountains in the southern and northern part of the west region contribute to the increase of surface temperature and the accumulation of local and transported pollutants. Then, high ozone concentrations are mainly associated with adverse meteorological and topographical conditions.
Ozone episodes, which are when the national standards are exceeded, was verified by Geraldino et al. (2017) in Rio de Janeiro city, in special in Bangu, with more frequency with weak surface winds, wind speeds lower than 2 m s -1 and from east (e), similar to observed in Gericinó station in 2013, where occurred the second biggest violation number during 3 monitored years. however, most lUCIANA m.B. VeNTURA et al.

AIR QUAlITY RIO 2016
An Acad Bras Cienc (2019) 91(1) e20170984 10 | 14 Olympic stations showed wind direction to susoutheast (SSe) with a wind speed of 1 m s -1 . Despite the reduction of sulfur contents in diesel from 500 ppm to 10 ppm from 2011 to 2013, the introduction of catalysts in the new trucks and the reduction of NOx levels, ozone levels were not affected. A similar observation was done by a group of researchers (Adame et al. 2008, Wilson et al. 2012) who analyzed trends in ozone levels in the european troposphere. They indicated that ozone average levels have been increasing, despite reductions in pollutants, which impact ozone formation. However, they also identified year-byyear variations, caused by climate and weather events, and suggested they could be masking the impact of emission reductions on long-term ozone trends. This study confirmed that the relationship between ozone and its precursors is complicated. It is caused by the fact that meteorological and chemical processes can interact over a remarkably wide range of temporal and spatial scales.  Figure 4 showed the concentration of O 3 in hourly average through the 3 years. So as verified by martins et al. (2015), the O 3 concentrations begin to increase after sunrise and peaked between 13 -15 h, which occurs the most intense solar radiation along with high NOx levels (INeA 2016a). The highest hourly O 3 concentrations were measured from November to February, when levels as high as 200 µg m³ were registered ( Figure 5). high temperatures, low formation of clouds and high incidence of solar radiation are the main reasons for the results obtained. These are common phenomena in this period of the year that naturally helps the formation of O 3 in the atmosphere especially in subtropical climate regions, like the Rio de Janeiro city (Silva et al. 2018). The lowest O 3 concentrations were observed during the early morning due to weak sunlight as well as the consumption processes prevailing over the forming processes ( Table IV were significant statistically. According to Silva et al. (2016), although the process of ozone formation is initiated photochemically, the correlation was also high for the temperature. This fact was justified because solar radiation and temperature are two meteorological parameters that are interconnected and highly correlated.
Ozone episodes usually occur during hot and sunny days (spring and summer), where strong solar radiation and high temperatures favor the reactions that lead to the ozone formation and various other oxidizing species (Finlayson-Pitts and Pitts 2000). Therefore, these days were not expected during the August of 2016, when the Olympic Games happened, because, in the winter of the last three years (2013, 2014 and 2015), the hourly maximum O 3 concentrations did not reach the Brazilian standard (160 µg m³).

CONCLUSION
In order to reduce the impact of air quality on athletes' performance, the State of Rio de Janeiro installed monitoring networks to determine the levels of the main pollutants (Pm 10 , Pm 2.5 , and O 3 ) before and during Olympic Games Rio 2016. This study presents data from 2013 to 2015, a period before Rio 2016, which occurred the main changes in the city's infrastructure (e.g., construction of roads, hotels, arenas etc). According to the results, concentrations of Pm 10 and Pm 2.5 had increased in this period (2013-2015) when compared to previous years. This input is mainly caused by the uncountable civil works initiated in Rio de Janeiro city to Olympic arenas construction and urban mobility improvements. This scenario was also verified in other host-cities, like Athens, in Greece, in the preparation the 2004 Olympic Games. however, the concentrations observed for all pollutants monitored in Rio de Janeiro city were much lower than that measured in Beijing, in 2008. But, it is worth to mention, that the O 3 and Pm 2.5 were the pollutants that exceeded the air quality standards more often. Therefore, they were considered the critical pollutants to be controlled during Rio 2016. lastly, it is expected that with the end of all civil works initiated due to the 2016 Olympic Games, the air quality in Rio de Janeiro city will improve, especially during the sportive event.  Figure S1 -Wind rose from 2013-2015 data of the stations: 1 -Jacarepaguá; 2 -Recreio dos Bandeirantes; 3 -Campos dos Afonsos; 4 -Gericinó; 5 -Urca; 6 -leblon; 7 -lagoa; 8maracanã; 9 -engenho de Dentro; 10 -Downtown. Figure S2 -Boxplot for the Pm 10 and Pm 2.5 concentrations for each sport regions from 2013 to 2015.