Impact on Air Quality of the COVID-19 Lockdown in the Urban Area of Palermo (Italy)

At the end of 2019, the first cases of coronavirus disease (COVID-19) were reported in Wuhan, China. Thereafter, the number of infected people increased rapidly, and the outbreak turned into a national crisis, with infected individuals all over the country. The COVID-19 global pandemic produced extreme changes in human behavior that affected air quality. Human mobility and production activities decreased significantly, and many regions recorded significant reductions in air pollution. The goal of our investigation was to evaluate the impact of the COVID-19 lockdown on the concentrations of the main air pollutants in the urban area of Palermo (Italy). In this study, the trends in the average concentrations of CO, NO2, O3, and PM10 in the air from 1 January 2020 to 31 July 2020 were compared with the corresponding average values detected at the same monitoring stations in Palermo during the previous five years (2015–2019). During the lockdown period (10 March–30 April), we observed a decrease in the concentrations of CO, NO2, and particulate matter (PM)10, calculated to be about 51%, 50%, and 45%, respectively. This confirms that air pollution in an urban area is predominantly linked to vehicular traffic.


Introduction
The new severe acute respiratory syndrome-coronavirus (SARS-CoV2) has had an unprecedented impact around the world. The spread of the 2019 coronavirus disease (COVID- 19), initially identified in Wuhan in China, led to over one million cases worldwide in the first four months [1]. The virus has affected almost all countries on the planet (235 in total), causing so far over 34 million confirmed cases and about 1,030,000 deaths [2]. This has resulted in lockdown in many nations. The COVID-19 pandemic has substantially affected the human society, including healthcare, economic structures, and social relationships [3].
In Italy, between the months of February and March, a series of national and regional legislative measures were issued to stop the spread of COVID-19, which radically changed the daily habits and lifestyles of citizens. The lockdown and the related implemented measures led to a sudden drop in economic activities, including a fall in road transport in many cities and a drastic reduction in the movement of citizens through the adoption of smart-working practices wherever possible.
In Italy, the Decree of the President of the Council of Ministers (DPCM) of 9 March 2020 [4], also known as the "I stay at home" Decree, extended the measures on confinement and restriction on travel to the whole country and prohibited any form of gathering of people in public places or locations open to the public, also suspending sporting and cultural events. The DPCM of 21 March 2020 [5], known as "Italy's block", blocked non-essential activities and imposed a ban on moving to different municipalities. In Italy, lockdown and travel restrictions were maintained until 3 May 2020.
the Mediterranean basin, which considerably influences the concentration of PM 10 . Potential local pollutants are limited to emissions from traffic, domestic heating, and small manufacturing industries.
We decided to compare the data collected from monitoring stations during the period from 1 January 2020 to 31 July 2020 with the averages for the same period calculated for the five-year period of 2015-2019.
The CO, NO 2 , and O 3 values were recorded hourly, whereas the PM 10 values were recorded daily. Each hourly or daily value for the five-year period is the result of the average calculated from all the values recorded in the same hour or day in the considered five years. From the data of the five-year period, the values due to abnormal and/or exceptional weather conditions and not attributable to anthropogenic emissions were deleted.
The comparison was performed using data on pollutants collected from monitoring stations, shown in Table 1. In total, 213 days were analyzed for CO, NO 2 , O 3 , and PM 10 in urban stations. The selected urban stations are close to roads with different densities of vehicles, urban buses, and extra-urban buses. Gas analyses were performed using the municipal air quality monitoring network managed by Risorse Ambiente Palermo (RAP). Ambient CO concentration measurements were recorded according to European standard EN 14626; NO 2 concentration was analyzed according to European standard EN 14211 [21]. Ambient O 3 concentration measurements were recorded according to European standard EN 14625 via ultraviolet photometry. PM 10 was sampled with a beta gauge analyzer (Environnement S.A. MP101M, ENVEA, Italy), certified as equivalent to the reference method in accordance with EN12341 [22] (stations: Di Blasi and Giulio Cesare) and another beta gauge analyzer (OPSIS SM200, OPSIS, Sweden) for the Castelnuovo and Indipendenza stations.

Gaseous Air Pollutants, CO, NO 2 , and O 3
The average concentrations of the gaseous air pollutants carbon monoxide (CO) and nitrogen dioxide (NO 2 ) were used to verify the impact on pollution of the closure of all activities in a city.
The main sources of CO emissions are linked to incomplete combustion processes of natural gas, diesel, or gasoline from engines and household heating [23]. In recent years, the levels of CO in Palermo city were below the limit of 10 mg m −3 (maximum daily eight-hour mean) fixed by Directive 2008/50/EC [24]. Figure 1 shows the temporal trend in the hourly average concentration of carbon monoxide in the atmosphere, detected for the period from 1 January 2020 to 31 July 2020 for the urban stations compared to the trend of the hourly average concentration resulting from the processing of the data of the five-year period of 2015-2019 for the same months.
From 10 March (DPCM 9 March, 2020 [4], "I stay at home") and even more evident from 22 March (DPCM 21 March, 2020 [5] "Italy's block"), which defined the total lockdown, CO concentrations were lower than the corresponding values of the five-year (2015-2019) period, confirming that the main source of CO was heavy vehicular traffic. After the end of the lockdown, we observed that CO concentrations tended to slowly increase until they reached very similar values to those of the month of July of the 2015−2019 period. In Figure 2a, we report the frequency distribution of two temporal trends, where the sharp reduction observed during the lockdown period is more evident. Figure 2b    The trend for the hourly concentrations in the city of Palermo showed a drop in the week preceding the Prime Minister's decree of 9 March 2020. Following the detection of COVID-19 cases since 20 February, the closure of all schools and grades had already been implemented.
From 10 March (DPCM 9 March 2020 [4], "I stay at home") and even more evident from 22 March (DPCM 21 March 2020 [5] "Italy's block"), which defined the total lockdown, CO concentrations were lower than the corresponding values of the five-year (2015-2019) period, confirming that the main source of CO was heavy vehicular traffic. After the end of the lockdown, we observed that CO concentrations tended to slowly increase until they reached very similar values to those of the month of July of the 2015−2019 period.
In Figure 2a, we report the frequency distribution of two temporal trends, where the sharp reduction observed during the lockdown period is more evident. Figure  From 10 March (DPCM 9 March, 2020 [4], "I stay at home") and even more evident from 22 March (DPCM 21 March, 2020 [5] "Italy's block"), which defined the total lockdown, CO concentrations were lower than the corresponding values of the five-year (2015-2019) period, confirming that the main source of CO was heavy vehicular traffic. After the end of the lockdown, we observed that CO concentrations tended to slowly increase until they reached very similar values to those of the month of July of the 2015−2019 period. In Figure 2a, we report the frequency distribution of two temporal trends, where the sharp reduction observed during the lockdown period is more evident. Figure 2b plots the deciles calculated for the two trends, showing that D5 was 0.2 mg m −3 for the 2020 distribution and 0.47 mg m −3 for the 2015−2019 distribution, representing a reduction of −56%.  From 1 January to 10 March, 2020, the trend in the hourly concentration of NO2 can be superimposed on that observed in the prior five-years. Since the DPCM of 9 March, 2020, a drastic decrease in the hourly concentration of NO2 was observed. Nitrogen dioxide in the atmosphere is a highly reactive pollutant emitted from burning fossil fuels (diesel, gasoline, and coal) [25,26]. This pollutant is considered a good indicator of air pollution in urban areas, as its concentration is heavily influenced by transport and mobility of citizens. At the end of the lockdown, some economic activities did not resume, schools and universities remained closed, and smart working was maintained. This reopening period coincided with the partial emptying of the city for the summer holidays. The density of motor vehicles decreased because of the use of bicycles and scooters. Figure 4 shows the frequency distribution and deciles of two temporal trends. The trend in NO2 concentration observed is similar for the two periods, but with a strong concentration reduction in 2020. The concentrations decreased significantly during the lockdown due to the stoppage of transport and low mobility [27,28].  Table 2 reports the average values for CO and NO2 pollutants at the individual monitoring stations. The measured values showed that the link between improved air quality and traffic density was more visible in the DB and GC monitoring stations, with variations during the lockdown between −58% and −43% for CO and −37% and −54 % for NO2. From 1 January to 10 March 2020, the trend in the hourly concentration of NO 2 can be superimposed on that observed in the prior five-years. Since the DPCM of 9 March 2020, a drastic decrease in the hourly concentration of NO 2 was observed. Nitrogen dioxide in the atmosphere is a highly reactive pollutant emitted from burning fossil fuels (diesel, gasoline, and coal) [25,26]. This pollutant is considered a good indicator of air pollution in urban areas, as its concentration is heavily influenced by transport and mobility of citizens. At the end of the lockdown, some economic activities did not resume, schools and universities remained closed, and smart working was maintained. This reopening period coincided with the partial emptying of the city for the summer holidays. The density of motor vehicles decreased because of the use of bicycles and scooters. Figure 4 shows the frequency distribution and deciles of two temporal trends. The trend in NO 2 concentration observed is similar for the two periods, but with a strong concentration reduction in 2020. The concentrations decreased significantly during the lockdown due to the stoppage of transport and low mobility [27,28]. From 1 January to 10 March, 2020, the trend in the hourly concentration of NO2 can be superimposed on that observed in the prior five-years. Since the DPCM of 9 March, 2020, a drastic decrease in the hourly concentration of NO2 was observed. Nitrogen dioxide in the atmosphere is a highly reactive pollutant emitted from burning fossil fuels (diesel, gasoline, and coal) [25,26]. This pollutant is considered a good indicator of air pollution in urban areas, as its concentration is heavily influenced by transport and mobility of citizens. At the end of the lockdown, some economic activities did not resume, schools and universities remained closed, and smart working was maintained. This reopening period coincided with the partial emptying of the city for the summer holidays. The density of motor vehicles decreased because of the use of bicycles and scooters. Figure 4 shows the frequency distribution and deciles of two temporal trends. The trend in NO2 concentration observed is similar for the two periods, but with a strong concentration reduction in 2020. The concentrations decreased significantly during the lockdown due to the stoppage of transport and low mobility [27,28].  Table 2 reports the average values for CO and NO2 pollutants at the individual monitoring stations. The measured values showed that the link between improved air quality and traffic density was more visible in the DB and GC monitoring stations, with variations during the lockdown between −58% and −43% for CO and −37% and −54 % for NO2.  The measured values showed that the link between improved air quality and traffic density was more visible in the DB and GC monitoring stations, with variations during the lockdown between −58% and −43% for CO and −37% and −54 % for NO 2 .  Table 3. The calculated percentage changes show a strong decrease in carbon monoxide and nitrogen dioxide between the periods considered. The two-tailed paired t-test showed that the observed differences among the three groups were statistically significant at p < 0.05. Ozone levels have increased as a result of human activities. Photochemical processes influenced by anthropogenic emissions of ozone precursors (NO 2 and volatile organic compounds (VOCs)) have caused the current tropospheric ozone levels to be substantially higher compared to the natural background levels [29]. The critical aspect of ozone is related to its complex photochemistry, as the rates of ozone formation and accumulation are non-linear functions of the mixture of VOCs and NO x in the atmosphere [30]. Under certain conditions, NO x reduction can lead to higher ambient ozone levels [31]. Motor vehicles are the main source of NO x emissions in an urban area; the increase in ozone production has been linked to the reduction in vehicular traffic [32,33]. The increase in ozone concentration that occurred during the lockdown is highlighted in Figures 5 and 6.

Particulate Matter PM10
Particulate matter in urban areas is mainly composed of mineral dust, metals, metalloids, sea salts, ammonium nitrate and sulfate, organic compounds, and elemental carbon [37]. Anthropogenic airborne particulate matter comes from a variety of sources, including traffic, industry, commerce, and domestic heating [23].
The daily variation in PM10 from 1 January to 31 July for the average values of 2015-2019 and for 2020 is shown in Figure 7

Particulate Matter PM10
Particulate matter in urban areas is mainly composed of mineral dust, metals, metalloids, sea salts, ammonium nitrate and sulfate, organic compounds, and elemental carbon [37]. Anthropogenic airborne particulate matter comes from a variety of sources, including traffic, industry, commerce, and domestic heating [23].
The daily variation in PM10 from 1 January to 31   The temporal analysis of the O 3 level showed an opposite behavior during the block (63 µg·m −3 ) compared to the average values of NO 2 (21 µg·m −3 ) at the same monitoring station. An increase in O 3 is usual during spring and summer due to the higher solar radiation (in terms of intensity and daily duration), which promotes the photolysis of NO 2 [34,35]. The increase in ozone concentrations registered in Palermo during the lockdown is linked to two factors: the decrease in NO 2 and the increased solar radiation [36].

Particulate Matter PM 10
Particulate matter in urban areas is mainly composed of mineral dust, metals, metalloids, sea salts, ammonium nitrate and sulfate, organic compounds, and elemental carbon [37]. Anthropogenic airborne particulate matter comes from a variety of sources, including traffic, industry, commerce, and domestic heating [23].
The daily variation in PM 10 Table 4. The GC, IND, and DB sites showed higher percentage changes of −43%, −44%, and −53%, respectively, between 1 March and 30 April. These results can be attributed to the reduction in vehicular traffic as a consequence of the closure of all schools and universities and the decrease in workers' movements. This hypothesis is confirmed by several previous studies conducted in the same area, in which the background value for PM 10 was estimated to be around 23 µg·m −3 on the basis of a linear correlation between PM 10 and CO and NO 2 [38,39]. Thus, the constant presence of a fraction of PM 10 , not strictly related to vehicle emissions, could result from soil and road dust particle resuspension. Dongarra et al. [39] confirmed that about 50% of particulate matter is produced by road traffic.
The reported data highlight the changes in air quality during the lockdown period. Strict alternative measures could be envisaged, such as short-term blocks, for pollution reduction in urban air. Table 5 shows that the decreases in CO, NO 2 , and PM 10 concentrations can be evaluated starting from the first six days of block. A statistical evaluation was conducted, which highlighted percentage changes of concentrations between 31% and 66% for CO, between 30% and 52% for NO 2 , and between 24% and 59% for PM 10 . The t-test confirmed that the variations observed CO, NO 2 , and PM 10 between the lockdown period and the previous five years were statistically significant (2015-2019; p < 0.05). Results of the parametric t-test at p < 0.05.

Pre-Lockdown
To verify the findings of the analysis of this study and the reliability and representativeness of the 2015-2019 database, a statistical analysis was also conducted to check the percentage changes for each pollutant from 1 January to 29 February 2020. During this period, the city of Palermo was not affected by any restrictions related to the COVID-19 pandemic. The results obtained showed that in the period considered, the differences between the average and the median between the two data series were practically negligible, and the variations were all less than 6% in absolute value for the arithmetic means and 10% for the medians, as shown in Table 6. The data obtained confirmed, despite the many variables involved, the reliability and representativeness of the 2015-2019 database used for the comparison. CO data expressed in mg·m −3 , NO 2 data expressed in µg·m −3 , PM 10 data expressed in µg·m −3 .

Conclusions
Air pollution studies should be considered as part of an integrated approach for sustainable development and the protection of human health. The national lockdown during the COVID-19 pandemic provided an opportunity to work on improving air quality. The results of this study demonstrate that anthropogenic activities strongly influence air quality. The lockdown period produced a sharp reduction in all pollutants that are closely linked to vehicular traffic. In the urban area of Palermo (Italy) from 10 March to 30 April, the concentrations of CO, NO 2 , and PM 10 decreased to about 51%, 50%, and 45%, respectively. The period of forced block produced decreases in the concentrations of CO, NO 2 , and PM 10 from the first six days of lockdown. The lack of a decrease of tropospheric ozone during the block due to non-linear chemical effects showed that these reductions will remain challenging, even with effective policies to reduce primary pollutants.
The findings reported here are a useful indication for competent authorities to rethink existing regulatory plans and provide assurance for the implementation of rigorous alternative measures such as short-term blocks to produce a real improvement in air quality. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.