The Impact of Coronavirus Disease of 2019 (COVID-19) Lockdown Restrictions on the Criteria Pollutants

Air pollution is accountable for various long-term and short-term respiratory diseases and even deaths. Air pollution is normally associated with a decreasing life expectancy. Governments have been implementing strategies to improve air quality. However, natural events have always played an important role in the concentration of air pollutants. In Australia, the lockdown period followed the Black Summer of 2019–2020 and coincided with the season of prescribed burns. This paper investigates the changes in the concentration of criteria pollutants such as particulate matter, nitrogen dioxide, ozone, and sulphur dioxide. The air quality data for the lockdown period in 2020 was compared with the pre-lockdown period in 2020 and with corresponding periods of previous years from 2016 to 2019. The results were also compared with the post-lockdown scenario of 2020 and 2021 to understand how the concentration levels changed due to behavioural changes and a lack of background events. The results revealed that the COVID-19 restrictions had some impact on the concentration of pollutants; however, the location of monitoring stations played an important role.


Introduction
According to the World Health Organization's (WHO's) report, seven million deaths occur every year because of exposure to air pollution [1]. Air pollution is a cause of one in every eight deaths globally [2]. Unfortunately, worldwide nine out of ten people are exposed to a high level of air pollutants [1]. The severity of the effects of air pollution on human health is immense and there is an immediate need to curb this problem, and effective management of ambient air pollution can lead to a substantial reduction of pollutant concentrations [3]. However, the unprecedented outbreak of coronavirus disease of 2019 (COVID-19) has made a remarkable breakthrough in unresolved air pollution management [4].
The WHO declared COVID-19 a global pandemic on 12 March 2020 [5][6][7]. All affected countries enforced lockdowns and preventive measures to stop the virus's spread [8,9]. However, the situation was a little different in Australia. From 23 March 2020, the Queensland Department of Health announced that some businesses would be required to close or limit their operation [10]. On 5 June 2020, the Premier announced that the Queensland Tourism and Accommodation Industry COVID- 19 Safe Plan had been approved, with the plan applying to thousands of tourism and accommodation businesses. This made Queensland's situation unique as there were some COVID-19-related restrictions, but no complete lockdown was enforced. We chose Rocklea monitoring station to collect most of the data on various air quality parameters such as PM10, PM2.5, NO2, and ozone. The SO2 monitoring data was collected from Springwood monitoring station. Meteorological parameters such as wind speed, wind direction, relative humidity, and ambient temperature were also collected from Rocklea monitoring station. Located approximately 500 metres southeast of the We chose Rocklea monitoring station to collect most of the data on various air quality parameters such as PM 10 , PM 2.5 , NO 2 , and ozone. The SO 2 monitoring data was collected from Springwood monitoring station. Meteorological parameters such as wind speed, wind direction, relative humidity, and ambient temperature were also collected from Rocklea monitoring station. Located approximately 500 metres southeast of the railway line in the grounds of the Oxley Common, the Rocklea monitoring station is surrounded by open parkland with light industry and residential uses surrounding the site. The site is wellplaced to represent the background air quality for Brisbane and often used to establish the background concentrations in the air quality dispersion modelling assessments for different industrial emission sources in Brisbane. To investigate the impact of restricted emissions and behavioural changes, we analysed the PM 2.5 , PM 10 , and NO 2 data from other monitoring stations such as Brisbane Central Business Districts (Brisbane CBD), Woolloongabba, and South Brisbane.

Analysis Methodology
The raw data with hourly averages was downloaded and cleaned for any error values due to instrument malfunction such as -9999 and -1111. The error values were replaced with empty data cells. The plots were generated using the OpenAir package in the R programming environment. To analyse the impact of COVID-19 restrictions, the pollutant concentration data were segregated into periods: pre-lockdown (1 January 2020 to 19 March 2020), lockdown period (20 March 2020 to 5 June 2020), and post-lockdown period (6 June 2020 to 31 December 2020). The different periods were again compared with the same periods for each of the last four years (2016 to 2019) and for the year 2021. The comparison of pollutant concentrations for one year does not give a very clear idea of the variation in the emissions of the pollutants as the variations in the pollutants are very much dependent on the meteorological parameters [20], hence the meteorology of the other years was also compared.

Analysis of Meteorological Parameters
The changes in the meteorological parameters are presented in Figures 2 and 3. The wind speed and wind direction are presented in the form on wind-rose plots for different monitoring periods of 2020 in Figure 2. In general, the prevalent winds were from the northeast and southwest quadrants. There was a lower frequency of winds from the northeast quarter, which could be due to seasonal variation during the lockdown period and early parts of the post-lockdown period.
ses 2022, 10, x FOR PEER REVIEW 4 o northeast quarter, which could be due to seasonal variation during the lockdown per and early parts of the post-lockdown period.     Figure 3 illustrates the daily average variation in ambient temperature and relati humidity. During the pre-lockdown period, temperatures were higher compared during lockdown and the post-lockdown period, such that the highest temperature in t post-lockdown period was 0.1 °C lower than the lowest temperature in the pre-lockdow period. The average temperature of the study periods was found to follow a monoto ically decreasing trend, having the highest average of 25.4 °C during the pre-lockdow period. The summer weather for Brisbane starts from December and winter starts fro June [21].
A similar trend was observed for wind speed. With an average wind speed of m/s, in the during-lockdown period, more frequent winds were from the south a southwest. According to the Bureau of Meteorology, Australia, Brisbane falls under t category of 'light winds' because the maximum daily average wind speed was 4 m even though Brisbane comes under a tropical cyclone risk area [22]. Nonetheless, hou wind speeds of up to 6.3 m/s were observed in Brisbane during the pre-lockdown a during-lockdown periods, which were mostly southerly winds. For the sudden inc ment of wind speed, we can consider Brisbane's topography, as well as the pr sure-gradient force, the vertical exchange of horizontal momentum in the boundary la er, and the Coriolis force as some of the factors affecting it [23]. In terms of wind spee the post-lockdown period was relatively calm, and slow and calm wind speeds are due northerly winds. The statistical comparison of the different meteorological paramete for the during-lockdown period for the years 2016 to 2021 are shown in Table S1. The were no substantial differences in the meteorological variables among years during lockdown period.

Analysis of Pollutants for 2020 Events
This section explains the differences in the 24 h average concentrations for the d ferent pollutants in 2020. Table S2 in the supplementary information summarises t changes in different statistical parameters for 24 h concentrations of all pollutants f 2020.  Figure 3 illustrates the daily average variation in ambient temperature and relative humidity. During the pre-lockdown period, temperatures were higher compared to during lockdown and the post-lockdown period, such that the highest temperature in the postlockdown period was 0.1 • C lower than the lowest temperature in the pre-lockdown period. The average temperature of the study periods was found to follow a monotonically decreasing trend, having the highest average of 25.4 • C during the pre-lockdown period. The summer weather for Brisbane starts from December and winter starts from June [21].
A similar trend was observed for wind speed. With an average wind speed of 1.5 m/s, in the during-lockdown period, more frequent winds were from the south and southwest. According to the Bureau of Meteorology, Australia, Brisbane falls under the category of 'light winds' because the maximum daily average wind speed was 4 m/s, even though Brisbane comes under a tropical cyclone risk area [22]. Nonetheless, hourly wind speeds of up to 6.3 m/s were observed in Brisbane during the pre-lockdown and during-lockdown periods, which were mostly southerly winds. For the sudden increment of wind speed, we can consider Brisbane's topography, as well as the pressure-gradient force, the vertical exchange of horizontal momentum in the boundary layer, and the Coriolis force as some of the factors affecting it [23]. In terms of wind speed, the post-lockdown period was relatively calm, and slow and calm wind speeds are due to northerly winds. The statistical comparison of the different meteorological parameters for the during-lockdown period for the years 2016 to 2021 are shown in Table S1. There were no substantial differences in the meteorological variables among years during the lockdown period.

Analysis of Pollutants for 2020 Events
This section explains the differences in the 24 h average concentrations for the different pollutants in 2020. Table S2 in the Supplementary Materials summarises the changes in different statistical parameters for 24 h concentrations of all pollutants for 2020.

Particulate Matter
The time-series 24 h average PM 10 and PM 2.5 concentrations are presented in Figure 4. The average 24 h PM 10 concentration at Rocklea dropped by 15.8% during the lockdown period compared to the pre-lockdown concentration. Similarly, the 24 h PM 10 concentration dropped by 8.2%, 3.1%, and 3.1% in Brisbane-CBD, South-Brisbane, and Woolloongabba, respectively. The average 24 h PM 10 concentration witnessed a reduction of 8.3% during the lockdown period. The maximum 24 h concentration in the pre-lockdown period was 48 µg/m 3 recorded on 20 February 2020. However, this concentration was the only peak recorded in contrast to the general trend. This could be due to some regional event that resulted in the elevated concentration. Similarly, the during-lockdown period also recorded two minor peaks of 32 and 23 µg/m 3 . The overall air quality levels did not change significantly on removing the peaks due to background events such as dust storms. The average 24 h PM 10 concentration dropped by 7.5% during the lockdown period compared to the pre-lockdown period on removing the peaks due to background events. Hence, the reduction in the PM 10  the world [24,25]. In addition to anthropogenic activities, the role of meteorology cannot be ignored. For example, there was a higher frequency of winds from the northeastern quadrant in 2020 during the lockdown period, as shown in Figure S1 (Supplementary Materials). Similarly, there were some notable differences in the frequency of the counts by wind direction for other years, as shown in Figures S2 and S3. to some regional event that resulted in the elevated concentration. Similarly, the dur-ing-lockdown period also recorded two minor peaks of 32 and 23 µ g/m 3 . The overall air quality levels did not change significantly on removing the peaks due to background events such as dust storms. The average 24 h PM10 concentration dropped by 7.5% during the lockdown period compared to the pre-lockdown period on removing the peaks due to background events. Hence, the reduction in the PM10 concentration could be due to the restricted transportation traffic and industrial emissions. Similar observations were made in other metropolitan cities around the world [24,25]. In addition to anthropogenic activities, the role of meteorology cannot be ignored. For example, there was a higher frequency of winds from the northeastern quadrant in 2020 during the lockdown period, as shown in Figure S1 (Supplementary Information). Similarly, there were some notable differences in the frequency of the counts by wind direction for other years, as shown in Figures S2 and S3. In comparison to the pre-lockdown period, the during-lockdown period witnessed an increase of 10% of the average 24 h PM2.5 concentration. The 24 h concentration during the post-lockdown period dropped marginally by 0.6% compared to during the lockdown period. The post-lockdown period also observed some peaks such as one of 25.6 µ g/m 3 on 7th June 2020, which could have resulted from pyrogenic emissions or other regional events. Corresponding peaks were also observed in the PM10 plot. Hence, lockdown had little to no effect on the PM2.5 concentration in the Rocklea area. This could In comparison to the pre-lockdown period, the during-lockdown period witnessed an increase of 10% of the average 24 h PM 2.5 concentration. The 24 h concentration during the post-lockdown period dropped marginally by 0.6% compared to during the lockdown period. The post-lockdown period also observed some peaks such as one of 25.6 µg/m 3 on 7 June 2020, which could have resulted from pyrogenic emissions or other regional events. Corresponding peaks were also observed in the PM 10 plot. Hence, lockdown had little to no effect on the PM 2.5 concentration in the Rocklea area. This could be due to several reasons such as regional events, and the fact that vehicle movement was not as prominent as in the rest of Brisbane due to its proximity to light industry. The lockdown restrictions imposed in Brisbane were not like the curfew-like conditions in the rest of the world. Essential industries were allowed to operate as usual. Further, some peaks were noticed in the PM 2.5 concentration levels during the lockdown period, which could be due to a regional event and pyrogenic emission episodes [26][27][28].
Therefore, comparison of different periods at other monitoring stations gives a good indication of the impact of restrictions on PM 2.5 and PM 10 concentration levels, as shown in Figure 5. The 24 h PM 2.5 concentration levels dropped by 13.5%, 12.3%, and 5.4% in Brisbane-CBD, South-Brisbane, and Woolloongabba, respectively. In general, Rocklea monitoring station recorded the lowest 24 h PM 10 concentration and the highest PM 2.5 concentration compared to the other monitoring stations. However, for the post-lockdown period, Rocklea monitoring station recorded the lowest 24 h PM 2.5 concentration compared to the other stations.
indication of the impact of restrictions on PM2.5 and PM10 concentration levels, as shown in Figure 5. The 24 h PM2.5 concentration levels dropped by 13.5%, 12.3%, and 5.4% in Brisbane-CBD, South-Brisbane, and Woolloongabba, respectively. In general, Rocklea monitoring station recorded the lowest 24 h PM10 concentration and the highest PM2.5 concentration compared to the other monitoring stations. However, for the post-lockdown period, Rocklea monitoring station recorded the lowest 24 h PM2.5 concentration compared to the other stations.

Nitrogen Dioxide
Anthropogenic emissions from sources such as the combustion of fossil fuels (coal, gas, and oil), especially fuel used in vehicles and industrial activities, are mainly responsible for increasing level of nitric oxides (90 to 95%) and nitrogen dioxides (5 to 10 %) [29][30][31][32][33][34]. In the atmosphere, hydroperoxyl radicals (HO2) and organic peroxy radicals (RO2) react with nitric oxide (NO) to convert it to NO2; thus, increases in NO emissions lead to increases in NO2 concentrations [35]. In Brisbane, the lockdown conditions were not similar to the rest of world. They did not restrict local transportation to ensure social distancing, but for maintaining social distancing during travel people were using personal vehicles. Exhaust emissions from personal vehicles could be a major source of nitrogen oxide (NO) that is converted to nitrogen dioxide by atmospheric chemistry. Also, during the post-lockdown period there were 10%, 12%, and 24% increases in the use of private vehicles in January 2021, July 2021, and January 2022, respectively, which could be responsible for the increased levels of NO2 during the post-lockdown period compared to pre-lockdown conditions [36].
The 24 h average NO2 concentration level and diurnal trends at Rocklea monitoring station are presented in Figure 6. As observed, the 24 h average concentration shows an increasing trend during the lockdown period. The 24 h average NO2 concentration in-

Nitrogen Dioxide
Anthropogenic emissions from sources such as the combustion of fossil fuels (coal, gas, and oil), especially fuel used in vehicles and industrial activities, are mainly responsible for increasing level of nitric oxides (90 to 95%) and nitrogen dioxides (5 to 10%) [29][30][31][32][33][34]. In the atmosphere, hydroperoxyl radicals (HO 2 ) and organic peroxy radicals (RO 2 ) react with nitric oxide (NO) to convert it to NO 2 ; thus, increases in NO emissions lead to increases in NO 2 concentrations [35]. In Brisbane, the lockdown conditions were not similar to the rest of world. They did not restrict local transportation to ensure social distancing, but for maintaining social distancing during travel people were using personal vehicles. Exhaust emissions from personal vehicles could be a major source of nitrogen oxide (NO) that is converted to nitrogen dioxide by atmospheric chemistry. Also, during the post-lockdown period there were 10%, 12%, and 24% increases in the use of private vehicles in January 2021, July 2021, and January 2022, respectively, which could be responsible for the increased levels of NO 2 during the post-lockdown period compared to pre-lockdown conditions [36].
The 24 h average NO 2 concentration level and diurnal trends at Rocklea monitoring station are presented in Figure 6. As observed, the 24 h average concentration shows an increasing trend during the lockdown period. The 24 h average NO 2 concentration increased by 39% during the lockdown period compared to the pre-lockdown period. To further investigate the impact of restrictions, we analysed the changes in NO and NOx concentration levels at Rocklea monitoring station. The diurnal trends of NO, NO 2 , and NOx are presented in Supplementary Materials Figure S4. The NO levels increased by 32% during lockdown compared to pre-lockdown levels. The NO levels were relatively similar during the post-lockdown period, i.e., they increased marginally by 0.3% compared to during the lockdown period. This means vehicular movement was higher during lockdown and the post-lockdown period. This is well-supported by public surveys in Australia [36]. Figure 7 shows a comparison of diurnal patterns of NO 2 concentration levels at different monitoring stations. There is a clear similarity in the diurnal profile of NO 2 in the during-and post-lockdown periods. The NO 2 concentration levels at South Brisbane and Woolloongabba monitoring stations were substantially higher than at Rocklea monitoring station. This could be due to the placement of the monitoring stations, as they are close to a major highway. A similar trend was observed in the PM 2.5 and PM 10 emissions, as discussed in Section 3.2.1.
Australia [36]. Figure 7 shows a comparison of diurnal patterns of NO2 concentration levels at different monitoring stations. There is a clear similarity in the diurnal profile of NO2 in the during-and post-lockdown periods. The NO2 concentration levels at South Brisbane and Woolloongabba monitoring stations were substantially higher than at Rocklea monitoring station. This could be due to the placement of the monitoring stations, as they are close to a major highway. A similar trend was observed in the PM2.5 and PM10 emissions, as discussed in Section 3.2.1  However, the post-lockdown period observed a decreasing trend as the average 24 h concentration dropped by 10% compared to the during-lockdown period. The NO2 to NOx ratio increased by 2% during the lockdown period compared to the pre-lockdown period. On the other hand, the NO2 to NOx ratio decreased by 6% during the post-lockdown period. The lower share of NO2 levels for different periods in NOx could be related to NO2 changes due to seasonal variation and the increments in NO concentration levels could be related to more vehicles on the roads [36]. A similar pattern was also observed in Sydney by Duc et al. [37]. The authors mentioned that the variation in NO2 concentration could be due to the photochemical reaction change involving nitrogen Australia [36]. Figure 7 shows a comparison of diurnal patterns of NO2 concentration levels at different monitoring stations. There is a clear similarity in the diurnal profile of NO2 in the during-and post-lockdown periods. The NO2 concentration levels at South Brisbane and Woolloongabba monitoring stations were substantially higher than at Rocklea monitoring station. This could be due to the placement of the monitoring stations, as they are close to a major highway. A similar trend was observed in the PM2.5 and PM10 emissions, as discussed in Section 3.2.1  However, the post-lockdown period observed a decreasing trend as the average 24 h concentration dropped by 10% compared to the during-lockdown period. The NO2 to NOx ratio increased by 2% during the lockdown period compared to the pre-lockdown period. On the other hand, the NO2 to NOx ratio decreased by 6% during the post-lockdown period. The lower share of NO2 levels for different periods in NOx could be related to NO2 changes due to seasonal variation and the increments in NO concentration levels could be related to more vehicles on the roads [36]. A similar pattern was also observed in Sydney by Duc et al. [37]. The authors mentioned that the variation in NO2 concentration could be due to the photochemical reaction change involving nitrogen However, the post-lockdown period observed a decreasing trend as the average 24 h concentration dropped by 10% compared to the during-lockdown period. The NO 2 to NO x ratio increased by 2% during the lockdown period compared to the pre-lockdown period. On the other hand, the NO 2 to NO x ratio decreased by 6% during the post-lockdown period. The lower share of NO 2 levels for different periods in NO x could be related to NO 2 changes due to seasonal variation and the increments in NO concentration levels could be related to more vehicles on the roads [36]. A similar pattern was also observed in Sydney by Duc et al. [37]. The authors mentioned that the variation in NO 2 concentration could be due to the photochemical reaction change involving nitrogen oxides, carbon monoxide, volatile organic compounds, and ozone from Australian summer months (December to February) to Australian autumn months (March to May) and winter months (June to August), which results in the reduction of ambient temperature. As the solar insolation decreased during Australian winters, there was a reduction in photochemical activities, and ozone was rapidly consumed by nitrogen monoxide (NO) and nitrogen dioxide produced [38].
The increasing trend of NO 2 levels is also evident in the diurnal pattern which gives the finer details of the average change in the NO 2 concentration levels at each hour of the day. This type of analysis has been used by several authors to assess the impact of different ambient conditions on the pollutant concentration. The NO 2 concentration levels were typically higher in the morning and then dropped during the mid-day before peaking in the evening at around 18:00 h. This increment in the NO 2 concentration levels was more dominant during the lockdown period.

Ozone
The presence of nitrogen oxides (NO x ), volatile organic compounds (VOCs), and solar irradiation are the cause of the formation of the secondary pollutant ozone [33]. The composition of photochemically aged air advected into a region (urban or rural) and mixed with local emissions has long been recognised as a key control on the production rates of ozone and other photochemical species. For O 3 production, nitrogen dioxide serves as a precursor [24]. The oxidation of NO by O 3 , and combustion processes such as industrial boilers, vehicles, and ships are responsible for the direct emission of NO 2 , even if there was a restriction related to the pandemic on public transit [25,38,39]. The 24 h average ozone concentration levels and the hourly diurnal profile are presented in Figure 8.
peaking in the evening at around 18:00 h. This increment in the NO2 concentration levels was more dominant during the lockdown period.

Ozone
The presence of nitrogen oxides (NOx), volatile organic compounds (VOCs), and solar irradiation are the cause of the formation of the secondary pollutant ozone [33]. The composition of photochemically aged air advected into a region (urban or rural) and mixed with local emissions has long been recognised as a key control on the production rates of ozone and other photochemical species. For O3 production, nitrogen dioxide serves as a precursor [24]. The oxidation of NO by O3, and combustion processes such as industrial boilers, vehicles, and ships are responsible for the direct emission of NO2, even if there was a restriction related to the pandemic on public transit [25,38,39]. The 24 h average ozone concentration levels and the hourly diurnal profile are presented in Figure  8. As observed in Figure 8, the maximum daily concentration in the pre-lockdown period was 60.8 µ g/m 3 in addition to some of the other peaks. In contrast, the maximum daily concentration during lockdown and post-lockdown was observed to be 53.1 and 68.1 µ g/m 3 . The average 24 h concentration dropped by 7.2% during the lockdown period compared to the pre-lockdown period and then increased by 12.4% during the post-lockdown period. On the other hand, the reduction in the median 24 h ozone levels was relatively lower (4.4%).
Similar trends were observed in the diurnal profile (right side of Figure 8). Peak ozone was observed during 10:00 to 16:00 h with the minimum concentration in the morning during 06:00 to 08:00 h. This agrees with the local diurnal temperature profiles and the typical NOx titration from local motor-vehicle-related emissions in the morning. The decreasing trend of ozone levels during the lockdown period could also be attributed to the seasonal variation as higher ozone levels occur in the warmer months (October to March) and peak ozone is usually recorded in January. In addition, the ambient emission As observed in Figure 8, the maximum daily concentration in the pre-lockdown period was 60.8 µg/m 3 in addition to some of the other peaks. In contrast, the maximum daily concentration during lockdown and post-lockdown was observed to be 53.1 and 68.1 µg/m 3 . The average 24 h concentration dropped by 7.2% during the lockdown period compared to the pre-lockdown period and then increased by 12.4% during the post-lockdown period. On the other hand, the reduction in the median 24 h ozone levels was relatively lower (4.4%).
Similar trends were observed in the diurnal profile (right side of Figure 8). Peak ozone was observed during 10:00 to 16:00 h with the minimum concentration in the morning during 06:00 to 08:00 h. This agrees with the local diurnal temperature profiles and the typical NOx titration from local motor-vehicle-related emissions in the morning. The decreasing trend of ozone levels during the lockdown period could also be attributed to the seasonal variation as higher ozone levels occur in the warmer months (October to March) and peak ozone is usually recorded in January. In addition, the ambient emission concentrations also change seasonally as the solid-fuel heating and temperature-dependent emissions change due to weather changes [40]. Some evidence comes from a similar study conducted on ten major world cities from February to April. Sydney and Perth were included in this study and a comparison was made of February to April 2020 and 2019. The authors observed a reduction in NO 2 , CO, and PM in February, March, and April but an increase in O 3 in February and April [37,41]. It is not an unexpected observation that median O 3 levels increased during the lockdown period. Mostly in volatile organic compounds (VOC), a limited regime of ozone formation occurs and, because of reduced levels of NO x , the photochemistry reaction rate tends to increase and ozone levels also increase [42].

Sulphur Dioxide
The variation in the 24 h average concentration and the hourly diurnal profile are presented in the Figure 9. The pre-lockdown period witnessed a series of high 24 h SO 2 concentration levels during 19 January to 24 January 2020 when the concentration ranged between 10 and 37 µg/m 3 . Similarly, several peaks were observed during lockdown and post-lockdown, such as a 24 h average concentration of 18 µg/m 3 on 30 April and 22 µg/m 3 on 31 October 2020 . The average 24 h concentration during the lockdown period decreased by 18% whereas a negligible increment of 0.5% was observed during the post-lockdown period. As observed in the diurnal profile of the SO 2 concentration levels ( Figure 6, right Processes 2023, 11, 296 9 of 14 hand side plot), the SO 2 concentration levels peaked between 16:00 and 19:00 h, which could be due to vehicular movement. It could be inferred that SO 2 levels dropped during the lockdown period due to less traffic activity. The registered peak could be due to the regional wildfire events [43].
The variation in the 24 h average concentration and the hourly diurnal profile are presented in the Figure 9. The pre-lockdown period witnessed a series of high 24 h SO2 concentration levels during 19 th January to 24 January 2020 when the concentration ranged between 10 and 37 µ g/m 3 . Similarly, several peaks were observed during lockdown and post-lockdown, such as a 24 h average concentration of 18 µ g/m 3 on 30 th April and 22 µ g/m 3 on 31 st October 2020. The average 24 h concentration during the lockdown period decreased by 18% whereas a negligible increment of 0.5% was observed during the post-lockdown period. As observed in the diurnal profile of the SO2 concentration levels ( Figure 6, right hand side plot), the SO2 concentration levels peaked between 16:00 and 19:00 h, which could be due to vehicular movement. It could be inferred that SO2 levels dropped during the lockdown period due to less traffic activity. The registered peak could be due to the regional wildfire events [43].

Annual Trend Analysis of the Pollutants for 2016 to 2019 and 2021
The comparison of pollutant concentrations for one year does not give a very clear idea of the variation in the emissions of the pollutants as they are very much dependent on meteorological parameters [20]. As discussed above in Section 3.2, the changes in the criteria pollutants' concentration, especially NO2 and O3, were dependent on the seasonal variation between pre-lockdown and lockdown periods. Thus, comparison of the same period of different years gives more confidence to investigate if lockdown restrictions really affected the changes in the pollutants' concentrations. The comparison of the 24 h average concentration for different pollutants is presented in Figures S5 to S9 in the Supplementary Information Figure 10 presents the comparison of PM10, PM2.5, and PM Ratio (PM2.5/PM10) for different periods from 2016 to 2021. The extreme high concentrations (24 h level of greater than 150 µ g/m 3 for PM10 and greater than 50 µ g/m 3 for PM2.5) were excluded for better representation of these plots. As observed, the median values for 24 h PM10 concentrations during the pre-lockdown period remained similar for 2018 and 2019 but var-

Annual Trend Analysis of the Pollutants for 2016 to 2019 and 2021
The comparison of pollutant concentrations for one year does not give a very clear idea of the variation in the emissions of the pollutants as they are very much dependent on meteorological parameters [20]. As discussed above in Section 3.2, the changes in the criteria pollutants' concentration, especially NO 2 and O 3 , were dependent on the seasonal variation between pre-lockdown and lockdown periods. Thus, comparison of the same period of different years gives more confidence to investigate if lockdown restrictions really affected the changes in the pollutants' concentrations. The comparison of the 24 h average concentration for different pollutants is presented in Figures S5 to S9 in the Supplementary Materials.  In contrast to PM10 observations, the PM2.5 trends were higher. The literature research has shown that PM2.5 particles represent a main pollutant emitted from wildfire smoke constituting approximately 90% of the total particle mass [44][45][46][47]. A higher 24 h average on certain days in 2020 was observed (as shown in Figure S6). The PM ratio (ratio of PM2.5 to PM10) was higher during the lockdown period in 2020 compared to other years. This shows that during the lockdown period in 2020 PM2.5 was impacted by background events of which the impact of the COVID-19 lockdown restrictions dominated. However, this was not the case for rest of Australia. The greater metropolitan region in In contrast to PM 10 observations, the PM 2.5 trends were higher. The literature research has shown that PM 2.5 particles represent a main pollutant emitted from wildfire smoke constituting approximately 90% of the total particle mass [44][45][46][47]. A higher 24 h average on certain days in 2020 was observed (as shown in Figure S6). The PM ratio (ratio of PM 2.5 to PM 10 ) was higher during the lockdown period in 2020 compared to other years. This shows that during the lockdown period in 2020 PM 2.5 was impacted by background events of which the impact of the COVID-19 lockdown restrictions dominated. However, this was not the case for rest of Australia. The greater metropolitan region in New South Wales state observed a 12% reduction in PM 2.5 during the lockdown period in 2020 compared to 2019 [37].
As observed in Figure 11, the median of the 24 h NO 2 concentration levels during the lockdown period of 2020 was 14% higher than the average of the corresponding median values for 2016 to 2019. This was mainly due to the high-than-average NO 2 concentration during 2017 and 2018. Duc et al. [37] attributed the higher NO 2 concentrations during similar periods of 2017 and 2018 to higher temperatures. However, this was not the case in Brisbane as the median and mean temperature were not significantly different for all years except for the pre-and during-lockdown periods of 2021, as shown in Figure 12. In contrast to PM10 observations, the PM2.5 trends were higher. The literature research has shown that PM2.5 particles represent a main pollutant emitted from wildfire smoke constituting approximately 90% of the total particle mass [44][45][46][47]. A higher 24 h average on certain days in 2020 was observed (as shown in Figure S6). The PM ratio (ratio of PM2.5 to PM10) was higher during the lockdown period in 2020 compared to other years. This shows that during the lockdown period in 2020 PM2.5 was impacted by background events of which the impact of the COVID-19 lockdown restrictions dominated. However, this was not the case for rest of Australia. The greater metropolitan region in New South Wales state observed a 12% reduction in PM2.5 during the lockdown period in 2020 compared to 2019 [37].
As observed in Figure 11, the median of the 24 h NO2 concentration levels during the lockdown period of 2020 was 14% higher than the average of the corresponding median values for 2016 to 2019. This was mainly due to the high-than-average NO2 concentration during 2017 and 2018. Duc et al. [37] attributed the higher NO2 concentrations during similar periods of 2017 and 2018 to higher temperatures. However, this was not the case in Brisbane as the median and mean temperature were not significantly different for all years except for the pre-and during-lockdown periods of 2021, as shown in Figure 12.  On the other hand, the ozone concentrations were higher in the pre-lockdown and lockdown periods of 2020 compared to the similar periods in the preceding years. However, the post-lockdown period in 2020 and 2021 recorded lower median ozone concentration levels compared to 2016 to 2019 levels. This inconsistency in the trend of ozone levels between 2019 and 2020 at different sites could be due to different meteorological conditions, as well as the photochemistry mechanism. The ozone levels were particularly higher in the pre-lockdown period of 2020 and post-lockdown period of 2019 which overlapped with the pyrogenic events and heatwave episodes of 2019-2020 in southeast Australia [48,49]. In contrast to the other pollutants, SO2 concentration levels were highest in 2021.

Conclusions
Pyrogenic emissions such as bushfires, prescribed burns, and agricultural burning along with the COVID-19-related restrictions and post-lockdown behavioural changes have resulted in episodes of substantial changes (positive or negative) in air quality pa- On the other hand, the ozone concentrations were higher in the pre-lockdown and lockdown periods of 2020 compared to the similar periods in the preceding years. However, the post-lockdown period in 2020 and 2021 recorded lower median ozone concentration levels compared to 2016 to 2019 levels. This inconsistency in the trend of ozone levels between 2019 and 2020 at different sites could be due to different meteorological conditions, as well as the photochemistry mechanism. The ozone levels were particularly higher in the pre-lockdown period of 2020 and post-lockdown period of 2019 which overlapped with the pyrogenic events and heatwave episodes of 2019-2020 in southeast Australia [48,49]. In contrast to the other pollutants, SO 2 concentration levels were highest in 2021.

Conclusions
Pyrogenic emissions such as bushfires, prescribed burns, and agricultural burning along with the COVID-19-related restrictions and post-lockdown behavioural changes have resulted in episodes of substantial changes (positive or negative) in air quality parameters. The various criteria pollutants for different periods in 2020 were analysed and their comparison with the corresponding periods in 2016 to 2019 and 2021 was conducted. The following conclusions can be drawn from this study: 1.
The air quality levels during the lockdown period improved; however, some pollutant concentrations were largely impacted by background events.

2.
Meteorology also played an important role in the changes in pollutant concentration across different periods. However, across the different years, meteorological conditions were only marginal different. For example, there was a higher frequency of winds from the northeastern quadrant in 2020 during the lockdown period. 3.
The average 24 h PM 10 concentration dropped by 3.1% to 15.8% during the lockdown period compared to the pre-lockdown concentration levels at different monitoring stations.

4.
The lockdown period had no impact on the PM 2.5 concentration levels at Rocklea monitoring station. In comparison to the pre-lockdown period, the during-lockdown period witnessed an increase of 10% in the average 24 h PM 2.5 concentration. The 24 h concentration during the post-lockdown period dropped marginally by 0.6% compared to during the lockdown period. The PM 2.5 could be largely impacted by background events such as dust storms and bushfires. The 24 h PM 2.5 concentration levels dropped by 13.5%, 12.3%, and 5.4% in Brisbane-CBD, South-Brisbane, and Woolloongabba, respectively.

5.
The 24 h average NO 2 concentration increased by 39% during the lockdown period compared to the pre-lockdown period. The median of the 24 h NO 2 concentration levels during the lockdown period of 2020 was 14% higher than the average of the corresponding median values for 2016 to 2019. This was mainly due to the higherthan-average NO 2 concentration during 2017 and 2018. 6.
The average 24 h SO 2 concentration during the lockdown period decreased by 18%, whereas a negligible increment of 0.5% was observed during the post-lockdown period. The diurnal profile of the SO 2 concentration levels showed an incremental trend in the evening which could be due to vehicular movement.
The implication from COVID-19-related restrictions showed that the improvements in the concentration of criteria pollutants was small but measurable. Natural activities such as bushfires impacted the concentration substantially and contributed to elevated background levels. Hence, due to the location of monitoring stations and contextual natural events, the transition from combustion vehicles to electric or hybrid vehicles may not be enough for reducing pollutant concentrations in the future and biogenic sources such as dust and bushfires also play a critical role in air quality management in urban areas.
Supplementary Materials: https://www.mdpi.com/article/10.3390/pr11010296/s1, Figure S1: Windrose plots for pre-lockdown period for different years, Figure S2: Windrose plots for duringlockdown period for different years, Figure S3: Windrose plots for post-lockdown period for different years, Figure S4: Diurnal trends of NO, NO2 and NOx concentration levels at Rocklea monitoring station for year 2020, Figure S5: Calendar plot for Average 24-h PM10 concentration, Figure S6: Calendar plot for Average 24-h PM2.5 concentration, Figure S7: Calendar plot for Average 24-h NO2 concentration, Figure S8: Calendar plot for Average 24-h Ozone concentration, Figure S9: Calendar plot for Average 24-h SO2 concentration; Table S1. Data Attributes for Meteorological Parameters, Table S2. Statistical Data Attributes for pollutant parameters.  Data Availability Statement: The data for this study can be accessed from the Queensland Government website (https://www.data.qld.gov.au, last accessed on 15 December 2022).