Temporal Trend of PM 10 and Associated Human Health Risk over the Past Decade in Cluj-Napoca City, Romania

: The human health risk associated with PM 10 exposure was assessed for the residents of Cluj-Napoca city, Romania, for a best case-scenario based on the monthly average PM 10 and for a worst-case scenario based on the monthly 90th percentile of PM 10 concentration. As no toxicity value for PM 10 was available, for the calculation of the hazard quotient, the toxicity value was considered to be equal to the annual limit value (40 µ m / m 3 ) set in the European Union (EU), and to air quality guidelines (20 µ m / m 3 ) set by the World Health Organization ( WHO ). The daily PM 10 concentrations for the period 2009–2019, at the four monitoring stations existing in Cluj-Napoca, were obtained from the National Air Quality Monitoring Network. The annual PM 10 values ranged between 20.3 and 29.5 µ g / m 3 , and were below the annual limit value (40 µ g / m 3 ) set by European and national legislation, but above the annual air quality guideline (20 µ g / m 3 ) set by WHO . Generally, the monthly PM 10 concentrations were higher from October to March than in the rest of the year. The monthly air quality index ( AQI ) showed the good to moderate quality of the air during the whole decade; however, there were days when the air quality was unhealthy for sensitive population groups. The air quality was more or less constant during the warm months, and improved signiﬁcantly for the cold months from 2009 to 2019. In the best-case scenario, calculated using the EU annual limit value for PM 10 , the potential non-carcinogenic chronic health risk was present only in 2009 and 2010, but in the worst-case scenario, in each year there were periods, especially in the cold months, in which health risk was present. When considering the WHO air quality guidelines in the calculation of the health risk, the potential non-carcinogenic chronic health risk was present between October and March in each year in the best-case scenario, and in most of the months in the worst-case scenario.


Introduction
Despite the important measures taken to reduce pollutant emissions in the last few decades, air quality remains a major issue of concern all over the world [1,2]. Exposure to air pollution may limit value (50 µm/m 3 ) were measured by 22% of the monitoring stations in 17 out the 28 EU member states, while levels above WHO AQG (20 µm/m 3 ) were measured by 51% of the monitoring stations in the majority of the countries [4]. Furthermore, around one fifth of the European urban population was exposed to PM 10 levels above the EU daily limit value, and half of it to PM 10 above AQG [4]. Air quality in Romania represents a European and national concern. In this regard, the EU opened the infringement procedure against Romania in 2018 for not taking measures to reduce atmospheric pollution in the major cities. Bucharest, the capital city of Romania, received penalties for PM 10 levels exceedance, while Cluj-Napoca city received the same for NO 2 . In 2015, the European Environmental Agency estimated 27,280 premature deaths in Romania associated with PM, ozone and nitrogen dioxide, out of which 25,400 were attributed to PM concentrations [27]. The urban population of Romania exposed to PM 10 values above the EU daily limit values decreased from 53.1% in 2013 to 21.4% in 2017 [28]. This decrease could be the consequence of the initiation of air monitoring and pollution abatement measures requested by Law 104/2011, which transpose into the national legislation the Directive 2008/50/EC ambient air quality and cleaner air for Europe [29]. Numerous studies reported the level of PM in cities all over the world [1,[30][31][32][33]. Human health risk assessment studies regarding PM 10 and toxic elements in PM 10 were also carried out in several cities [34][35][36][37][38][39]. These studies revealed that although PM concentrations do not exceed air quality standards, the health impacts associated to air pollution exposure are still important [40]. Therefore, the assessment of the human health risks caused by exposure to airborne PM is an important step in order to control and mitigate urban air pollution. As the chemical composition of PM depends on a high number of factors, such as geography, season, climate and combustion sources, and differs from region to region, the health risk assessment in different cities could give a better insight into the population exposure and health risks [18]. In Romania, Leitte et al. presented the adverse effects of air pollution on the respiratory tract for the inhabitants of Drobeta Turnu Severin [39], while Dunea et al. showed the impact of PM 2.5 and associated metals on the health of children in the industrial area of Targoviste [40]. However, to the best of our knowledge, no health risk assessment associated with PM 10 has been conducted in Romania.
This research aimed to assess the health risks associated with PM 10 for the residents living in one of the largest urban agglomerations in Romania. The air quality index (AQI) was applied in the urban agglomeration of Cluj-Napoca, in order to analyze the temporal variation of the pollution level of PM 10 from 2009 to 2019, and to identify the seasons when air quality can become an issue for the local population. The health risk assessment, according to EU and WHO recommendations, was conducted for the first time in a Romanian city. The health risk assessment will offer information to local governments regarding the efficiency of air quality management. The results of this study can support decision-makers in implementing and developing better strategies and regulations to improve air quality and to mitigate the effects on human health.
Located in the Somes Mic river valley, surrounded by forests and grasslands, Cluj-Napoca has a continental climate characterized by warm, dry summers and cold winters. The city is an important trade and tourist center and a hub of the European network roads, connecting the country with Western Europe. It has a large industrial park with modern facilities, and the second main airport in Romania, after Bucharest, is located 9 km east of the center of the city. The local topography, mainly the Somes Mic River that crosses the city, influences the dominant NW (~15%), NE (~12%), W (~10%) and SW (~10%) wind direction [42]. Further, the dominant wind direction can favor the accumulation processes or the upwind emissions.

Air Quality Data
Cluj-Napoca has 4 stations belonging to the National Air Quality Monitoring Network that monitors the traffic (CJ-1), urban (CJ-2), suburban (CJ-3) and industrial (CJ-4) influences on the air pollution ( Figure 1). At the CJ-1, CJ-2 and CJ-3 stations, the PM10 concentration is measured gravimetrically using Skypost Tecora (Cogliate, Italy) low volume samplers, while at the CJ-4 station it is done by optic measurements using an LSPM10 (Unitec, Veggiano, Italy) analyzer. In this study, the assessment of PM10 air pollution in Cluj-Napoca was carried out using the daily PM10 concentrations provided by the operational monitoring stations. The daily PM10 was further aggregated to monthly and annual concentrations for the period 2009-2019. Data are publicly available on the National Air Quality Monitoring Network website [44].

Air Quality Index
The air quality can be comprehensively presented to citizens using the AQI, calculated according to Equation (1). While the AQI is normally used to report the daily air quality at each monitoring station, in this study it was used to assess the monthly variability and to underline the multi-annual trends using all four operational stations in Cluj-Napoca city, considering that a single monitoring station is not representative for a city and did not reflect spatial variations of PM10. Therefore, the daily AQI was calculated based on the average PM10 concentration obtained for the operational stations and the EU daily limit value (50 µg/m 3 ) [45,46]. The daily AQI was further aggregated to obtain the monthly AQI.
High AQI values indicate a high level of air pollution with PM10. The AQI values below 100 are considered satisfactory, while values above 100 are considered to pose health risks [45]. At the CJ-1, CJ-2 and CJ-3 stations, the PM 10 concentration is measured gravimetrically using Skypost Tecora (Cogliate, Italy) low volume samplers, while at the CJ-4 station it is done by optic measurements using an LSPM10 (Unitec, Veggiano, Italy) analyzer. In this study, the assessment of PM 10 air pollution in Cluj-Napoca was carried out using the daily PM 10 concentrations provided by the operational monitoring stations. The daily PM 10 was further aggregated to monthly and annual concentrations for the period 2009-2019. Data are publicly available on the National Air Quality Monitoring Network website [44].

Air Quality Index
The air quality can be comprehensively presented to citizens using the AQI, calculated according to Equation (1). While the AQI is normally used to report the daily air quality at each monitoring station, in this study it was used to assess the monthly variability and to underline the multi-annual trends using all four operational stations in Cluj-Napoca city, considering that a single monitoring station is not representative for a city and did not reflect spatial variations of PM 10 . Therefore, the daily AQI was calculated based on the average PM 10 concentration obtained for the operational stations and the EU daily limit value (50 µg/m 3 ) [45,46]. The daily AQI was further aggregated to obtain the monthly AQI.
High AQI values indicate a high level of air pollution with PM 10 . The AQI values below 100 are considered satisfactory, while values above 100 are considered to pose health risks [45].

Human Health Risk Assessment
To estimate the human health risk related to PM 10 exposure by inhalation, the hazard quotient (HQ) was calculated according to Equation (2) [26]: where EC is the exposure concentration of PM 10 (µg/m 3 ) calculated according to Equation (3), and TV is the toxicity value [26]. As no toxicity value for PM 10 was found in the literature, the calculations were made assuming that the TV was equal to the EU annual limit value of 40 µm/m 3 (HQ 1 ), and to the WHO AQG of 20 µm/m 3 (HQ 2 ).
where CA is the monthly average PM 10 concentration (µg/m 3 ) for the best-case scenario and the monthly 90th quartile for the worst-case scenario, ET is the exposure time (hours/day), EF is the exposure frequency (days/year), ED is the exposure duration (year) and AT is the averaging time calculated as ED × 365 days/year × 24 h/day. For the exposure of residents, ET was considered 24 h/day, EF was 350 days/year and ED was 30 years for adults [26]. Non-carcinogenic risk includes all the adverse health effects in the organism, excluding cancer caused by exposure factors. The used safety benchmark level for HQ is 1. Thus, the exposure to PM 10 could induce non-carcinogenic chronic effects if HQ > 1, while no non-carcinogenic health risk is expected if HQ < 1 [47].

Air Quality
The basic statistics for PM 10 Table 1. The annual average, calculated based on daily values, ranged between 20.3 and 29.5 µg/m 3 , this being below the annual limit value set by the EU legislation (40 µg/m 3 ), but above the AQG set by the WHO (20 µg/m 3 ) [4,29].   The highest annual maximum was measured in 2010 (111 µg/m 3 ), while the lowest annual maximum in 2012 (54.8 µg/m 3 ). During the monitoring period, the annual average concentration was comparable, ranging between 20.3 and 29.5 µg/m 3 , but the annual maximum almost halved from 111 µg/m 3 in 2010 to 62.6 µg/m 3 in 2019. The difference between the average and median annual PM 10 value decreased from 2009 to 2019, indicating a reduction of the days with high PM 10 values in favor of days with low PM 10 values. The positive skewness of the data also confirms this decrease. Highly skewed data were obtained for 2009, 2010 and 2019, the years that also have the highest numbers of days with PM 10 exceeding the daily limit value. The daily maximum PM 10 threshold value (50 µg/m 3 ) was not exceeded more than 35 times in the studied years. The highest number (29) of breaches occurred in 2009, of which 20 breaches occurred during October-December. With few exceptions, in each year, the majority of breaches occurred between October and March.
The observed decreasing trend could be explained by the existence of some pollution reduction measures related to traffic restriction in the city center, and to the decrease of emissions from domestic heating following the increase in ambient temperatures and reduction of the number of frost days in Cluj-Napoca during the winters. Temperature fluctuation is the main factor that favors the increase in PM 10 concentrations, while the wind speed influences the dispersion of pollutants ( Figure 2). Both temperature (r = −0.68) and wind speed (r = −0.44) are negatively correlated with the PM 10 concentration. The highest monthly average wind speed was recorded in spring (4.51 m/s), while the lowest monthly average wind speed was recorded in December (3.13 m/s). Generally, the highest PM 10 concentrations were recorded when the average monthly temperature was below 0 • C and the intensity of wind was low.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 6 of 13 The observed decreasing trend could be explained by the existence of some pollution reduction measures related to traffic restriction in the city center, and to the decrease of emissions from domestic heating following the increase in ambient temperatures and reduction of the number of frost days in Cluj-Napoca during the winters. Temperature fluctuation is the main factor that favors the increase in PM10 concentrations, while the wind speed influences the dispersion of pollutants ( Figure 2). Both temperature (r = −0.68) and wind speed (r = −0.44) are negatively correlated with the PM10 concentration. The highest monthly average wind speed was recorded in spring (4.51 m/s), while the lowest monthly average wind speed was recorded in December (3.13 m/s). Generally, the highest PM10 concentrations were recorded when the average monthly temperature was below 0 °C and the intensity of wind was low. . Generally, the monthly PM10 concentrations were higher in the cold seasons than in the warm seasons. A possible explanation for high PM10 levels in winter could be the accumulation of PM in the low atmosphere due to the reduction of mixing layer thickness, determined by meteorological conditions. Household heating and the use of solid road deicing products during winter could also have increased the PM10 emissions [1]. Both the maximum and the 90 th percentile of monthly PM10 (Figure 3b) showed decreasing trends, confirming the lowering of the value and frequency of high concentrations. In the case of the 90 th percentile, the higher values were measured between October and March, and a decreasing trend in monthly PM10 from 2009 to 2019 was observed.
To assess the spatial variability of PM10 concentration in Cluj-Napoca, the average PM10 for each monitoring station was calculated for the periods in which all four stations functioned. The average PM10 concentrations were comparable for the CJ-2 (urban) and CJ-3 (suburban) stations, with values 3 3 . Generally, the monthly PM 10 concentrations were higher in the cold seasons than in the warm seasons. A possible explanation for high PM 10 levels in winter could be the accumulation of PM in the low atmosphere due to the reduction of mixing layer thickness, determined by meteorological conditions. Household heating and the use of solid road deicing products during winter could also have increased the PM 10 emissions [1]. Both the maximum and the 90th percentile of monthly PM 10 ( Figure 3b) showed decreasing trends, confirming the lowering of the value and frequency of high concentrations. In the case of the 90th percentile, the higher values were measured between October and March, and a decreasing trend in monthly PM 10 from 2009 to 2019 was observed.
To assess the spatial variability of PM 10 concentration in Cluj-Napoca, the average PM 10 for each monitoring station was calculated for the periods in which all four stations functioned. The average PM 10 concentrations were comparable for the CJ-2 (urban) and CJ-3 (suburban) stations, with values of 26.8 µg/m 3 and 26.7 µg/m 3 , respectively. These values were almost two times higher than the average PM 10 concentrations found for the CJ-1 (traffic, 13.6 µg/m 3 ) and CJ-4 (industrial, 15.1 µg/m 3 ) stations. The low PM 10 values were measured by the industrial and traffic background stations, which confirms that industry and traffic are responsible for a lower share of PM 10 pollution than the other urban pollution sources, such as heating and constructions/demolitions. Appl. Sci. 2020, 10, x FOR PEER REVIEW 7 of 13 confirms that industry and traffic are responsible for a lower share of PM10 pollution than the other urban pollution sources, such as heating and constructions/demolitions.  [35]. The PM10 values measured in Cluj-Napoca were lower than those from Bucharest [48,49].
The average monthly AQI (Figure 3c) ranged between 15 (May 2019) and 87 (December 2009), indicating good to moderate air quality. However, the 90 th percentile (Figure 3d) ranged between 22 (May 2019) and 147 (January 2010), indicating that there were months when the air quality was unhealthy for sensitive population groups. Similar to the PM10 values, good air quality was found in the warm months, the air quality decreasing in the cold months. Furthermore, an air quality improvement from 2009 to 2019 was observed, especially for the cold months. The number of months with average AQI above 50 decreased from eight months in 2010 to four months in 2019, while instances of 90 th percentile of AQI above 100 were identified in four months in 2010, but in only one month between 2011 and the present.
The high PM10 levels reported all over Europe indicate that despite the progress made toward meeting the air quality standards, air pollution is still a critical issue. Even though Cluj-Napoca is not included in the top most polluted cities across Europe, there are still concerns about the air quality, especially during the cold season. In order to reduce the health risk associated with air pollution, effective air quality policies and tailored pollution abatement measures are needed, especially in cases of high urban agglomerations. In urban areas, most of the measures address the transport sector  [35]. The PM 10 values measured in Cluj-Napoca were lower than those from Bucharest [48,49].
The average monthly AQI (Figure 3c The high PM 10 levels reported all over Europe indicate that despite the progress made toward meeting the air quality standards, air pollution is still a critical issue. Even though Cluj-Napoca is not included in the top most polluted cities across Europe, there are still concerns about the air quality, especially during the cold season. In order to reduce the health risk associated with air pollution, effective air quality policies and tailored pollution abatement measures are needed, especially in cases of high urban agglomerations. In urban areas, most of the measures address the transport sector by encouraging the use of public transport or low pollution vehicles, such as electric cars or bicycles, and by limiting heavy vehicles' access to the city center and enhancing road infrastructures. The reduction of waste burning together with the shift towards low emission fuel-use for residential heating, along with emission-control and -law enforcing, could further reduce PM 10 emissions.

Human Health Risk Assessment
The exposure concentrations (EC) for the study periods, calculated based on monthly average PM 10  The non-carcinogenic chronic HQ 1 calculated based on the EU limit value obtained for the residential best-case scenario ranged between 0.2 and 1.1 in all studied years, indicating potential non-carcinogenic chronic health risk in the first years of the study (2009 and 2010), and no health risks for the other study years (2011-2019). However, for the worst-case scenario, the HQ 1 values were higher (0.2-1.6), with monthly peaks with values above the safety benchmark being found in each year. The monthly variation of HQ 1 in the best-case ( Figure 4a) and worst-case scenarios (Figure 4c) was similar to that in the PM 10 concentration, being higher in the cold months than in the warm months. In the best-case scenario, non-carcinogenic health risk was identified only in December 2009, while in the worst-case scenario, non-carcinogenic health risk was identified in each year, mainly in the winter seasons. In both exposure scenarios, over the years, a clear decreasing trend from 2009 to 2019 was observed.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 8 of 13 reduction of waste burning together with the shift towards low emission fuel-use for residential heating, along with emission-control and -law enforcing, could further reduce PM10 emissions.

Human Health Risk Assessment
The exposure concentrations (EC) for the study periods, calculated based on monthly average PM10 concentration, ranged from 7. The non-carcinogenic chronic HQ1 calculated based on the EU limit value obtained for the residential best-case scenario ranged between 0.2 and 1.1 in all studied years, indicating potential non-carcinogenic chronic health risk in the first years of the study (2009 and 2010), and no health risks for the other study years (2011-2019). However, for the worst-case scenario, the HQ1 values were higher (0.2-1.6), with monthly peaks with values above the safety benchmark being found in each year. The monthly variation of HQ1 in the best-case ( Figure 4a) and worst-case scenarios (Figure 4c) was similar to that in the PM10 concentration, being higher in the cold months than in the warm months. In the best-case scenario, non-carcinogenic health risk was identified only in December 2009, while in the worst-case scenario, non-carcinogenic health risk was identified in each year, mainly in the winter seasons. In both exposure scenarios, over the years, a clear decreasing trend from 2009 to 2019 was observed. The non-carcinogenic chronic HQ2 calculated based on the WHO AQG obtained for the best-case scenario ranged between 0.4 and 2.1 in all studied years, indicating a potential non-carcinogenic chronic health risk between October and March in each year, while for the worst-case scenario, the The non-carcinogenic chronic HQ 2 calculated based on the WHO AQG obtained for the best-case scenario ranged between 0.4 and 2.1 in all studied years, indicating a potential non-carcinogenic chronic health risk between October and March in each year, while for the worst-case scenario, the HQ 2 values were higher (0.5-3.2), exceeding the security benchmark in most of the months. The monthly HQ 2 values, both in the best-case ( Figure 4b) and worst-case scenarios (Figure 4d), were higher in the cold months than in the warm months, but slightly decreased from 2009 up to the present, without frequently falling below the security benchmark, indicating a non-carcinogenic risk for almost the whole period. The high HQ values suggest that environmental PM 10 levels, although below the legislative thresholds, are likely to induce various chronic pathologies.
The differences between the human health risk measurements from the air quality monitoring stations for the period 2009-2019, in Cluj-Napoca city, were observed both for the best-case and worst-case scenarios ( Figure 5). Thus, HQ 1 values, calculated based on the EU limit values, did not exceeded the unity threshold at any of the monitoring sites, suggesting no potential health risk, but they were higher at the urban (CJ-1) and suburban (CJ-3) background monitoring stations than at the traffic (CJ-1) and industrial (CJ-4) stations. Similarly, in the case of HQ 2 , the health risk was higher at the urban and suburban stations than at those monitoring the traffic and industrial background, but in this case potential non-carcinogenic risk might possibly appear. A similar pattern was observed for the worst-case scenario, both for HQ 1 and HQ 2 when the unity threshold was above 1, indicating the presence of potential non-carcinogenic risk.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 9 of 13 for almost the whole period. The high HQ values suggest that environmental PM10 levels, although below the legislative thresholds, are likely to induce various chronic pathologies. The differences between the human health risk measurements from the air quality monitoring stations for the period 2009-2019, in Cluj-Napoca city, were observed both for the best-case and worst-case scenarios ( Figure 5). Thus, HQ1 values, calculated based on the EU limit values, did not exceeded the unity threshold at any of the monitoring sites, suggesting no potential health risk, but they were higher at the urban (CJ-1) and suburban (CJ-3) background monitoring stations than at the traffic (CJ-1) and industrial (CJ-4) stations. Similarly, in the case of HQ2, the health risk was higher at the urban and suburban stations than at those monitoring the traffic and industrial background, but in this case potential non-carcinogenic risk might possibly appear. A similar pattern was observed for the worst-case scenario, both for HQ1 and HQ2 when the unity threshold was above 1, indicating the presence of potential non-carcinogenic risk.

CJ-1 CJ-2 CJ-3 CJ-4 CJ-1 CJ-2 CJ-3 CJ-4
Best Similar to Cluj-Napoca, a non-carcinogenic risk associated with PM10 exposure was reported in Krakow, Poland and Shanghai, China [10,50]. Opposite to Cluj-Napoca, no human health risk associated with exposure to PM10 for the general population, and a low risk for the sensitive population groups, were found for Pretoria West, South Africa [51].
Generally, during the warm months, the average HQ1 and HQ2 values were more or less constant, while during the cold months they varied widely. The obtained data suggest that, under the best-case scenario, presently, the inhabitants of Cluj-Napoca are unlikely to develop adverse health effects after PM10 exposure. However, in the worst-case scenario, the occurrence of adverse chronic health effects is still possible. The cumulative effects of PM10 with other air pollutants could increase the probability of manifesting adverse chronic non-carcinogenic health effects, although the additivity or synergism of the air pollutants is not clearly elucidated, and is influenced by the meteorological conditions [52][53][54]. A possible mitigation measure could be the reduction in the hours spent in urban environments with high levels of PM10 or other pollutants, and an increase in time spent in clean outdoor environments such as parks and green areas.
The obtained results highlighted the contribution of local meteorology and winter sources to the PM10 concentration. Although the PM10 concentrations for Cluj-Napoca city generally did not exceed air quality standards, this study reveals that the human health risks resulting from exposure to air pollution are still possible. The health risk assessment method used in this study has a number of Similar to Cluj-Napoca, a non-carcinogenic risk associated with PM 10 exposure was reported in Krakow, Poland and Shanghai, China [10,50]. Opposite to Cluj-Napoca, no human health risk associated with exposure to PM 10 for the general population, and a low risk for the sensitive population groups, were found for Pretoria West, South Africa [51].
Generally, during the warm months, the average HQ 1 and HQ 2 values were more or less constant, while during the cold months they varied widely. The obtained data suggest that, under the best-case scenario, presently, the inhabitants of Cluj-Napoca are unlikely to develop adverse health effects after PM 10 exposure. However, in the worst-case scenario, the occurrence of adverse chronic health effects is still possible. The cumulative effects of PM 10 with other air pollutants could increase the probability of manifesting adverse chronic non-carcinogenic health effects, although the additivity or synergism of the air pollutants is not clearly elucidated, and is influenced by the meteorological conditions [52][53][54]. A possible mitigation measure could be the reduction in the hours spent in urban environments with high levels of PM 10 or other pollutants, and an increase in time spent in clean outdoor environments such as parks and green areas.
The obtained results highlighted the contribution of local meteorology and winter sources to the PM 10 concentration. Although the PM 10 concentrations for Cluj-Napoca city generally did not exceed air quality standards, this study reveals that the human health risks resulting from exposure to air pollution are still possible. The health risk assessment method used in this study has a number of limitations, some of which are due to the fact that this approach focuses only on PM 10 without considering simultaneous exposure to other air pollutants, and thus underestimates the risk. The interactions between different pollutants were also neglected. The non-carcinogenic health risk was calculated based on several assumptions regarding the exposure time, exposure frequency and exposure duration that are not specific for the studied population, and neglects the different behaviors of individuals inside the studied population. Furthermore, we assumed that the average daily PM 10 concentrations measured by the for air quality monitoring stations are representative for the whole city. Despite these limitations, the findings may contribute to a better understanding of the potential health risks associated with human exposure to PM 10 in urban areas, providing valuable information for air quality managers and specialists. The next steps should include sources apportionment and the assessment of their contribution to PM 10 concentration, as well as assessing the carcinogenic human health risk of toxic compounds associated with PM 10 .

Conclusions
The average annual PM 10 level in Cluj-Napoca city, ranging between 20.3 and 29.5 µg/m 3 , was below the legislative threshold, and had a decreasing trend from 2009 to 2019, while the annual maximum almost halved from 111 µg/m 3 (2010) to 62.6 µg/m 3 (2019). Generally, the monthly PM 10 concentrations were higher in the cold seasons compared with the warm seasons. The average monthly AQI ranged between 15 (May 2019) and 87 (December 2009), indicating good to moderate air quality. However, the 90th percentile indicated several months when the air quality was unhealthy for sensitive population groups. The non-carcinogenic chronic risk, calculated based on the EU limit value, obtained for the best-case scenario ranged between 0.2 and 1.1, indicating potential non-carcinogenic chronic health risk for 2009 and 2010, and no health risks for 2011-2019, while for the worst-case scenario, monthly values above the safety benchmark were found in each year. The non-carcinogenic chronic risk calculated based on the WHO AQG for the best-case scenario was higher, indicating a potential non-carcinogenic chronic health risk between October and March in each year, while for the worst-case scenario this exceeded the security benchmark in most of the months. The decreasing trend of the HQ values in the last decade could indicate the existence of air quality improvement measures; however, the implementation of more mitigation measures could further reduce the health risk, especially in the case of sensitive population groups. The measures must mainly target heating systems, taking into account the high concentrations of PM 10 during the winter, and in particular the type of fuel and its quality.