Introduction

An infectious disease which is known as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or COVID-19 has been discovered in Wuhan, Hubei Province, China around the end of December 2019 (Du et al. 2020). In January 2020, it was quickly transmitted to several cities and outside China region (Li et al. 2021). In Indonesia, the first case of this virus was discovered in Jakarta city on March 2, 2020 (Ministry of Health of Indonesia 2020). Then, the cases increased rapidly with a total number of five thousand cases has been reported on May 17, 2020. Therefore, the Indonesian government implemented the large-scale social restriction (LSSR) policy to manage the COVID-19 spread. This policy aimed to decrease the large social activities to control the virus transmission. Due to the effect of this policy, all social, and commercial activities have been restricted including the closure of offices, schools, transportation systems, malls, cafes, and others. However, some places have been partially operated such as pharmacies, grocery stores, banks, and health clinics.

While as we know the cost of this control measure has declined the economic growth of countries. Several studies revealed that decreasing human activities during the COVID-19 period had an important role in improving air quality (Dentener et al. 2020; Singh et al. 2020). But, Wang et al. (2020) assumed that during heavy pollution, the reduction of human activities (i.e. transportation and industrial activities) could not stop the pollution when meteorological factors were not in favorable condition. NO2, CO, SO2, O3, and PM2.5 are mainly ambient air pollutants in urban areas and many studies have investigated severe health issues like respiratory and cardiovascular diseases, lung cancer, and pneumonia related to these pollutants (Kinney 2018; Mo et al. 2018). Even the World Health Organization reported that around 4.2 million deaths per year were due to acute and chronic respiratory diseases. For instance, many studies have revealed a strong relationship between the increased PM2.5 levels and acute health issues (Zhang et al. 2017; Neira et al. 2019). The PM2.5 levels and respiratory health issues, such as death rate, hospital admissions, and others showed a positive correlation (Cao et al. 2018; Lee et al. 2019; Rendana et al. 2021; Sipra et al. 2020).

Anthropogenic activities were a substantial source of air pollution all over the world and specifically in Jakarta. Several studies have reported the combustion of fossil fuels such as coal-fired power plants has majorly contributed to SO2 and NO2 pollution (Rendana 2021; Xu 2019; Zhao et al. 2018). In 2014, there were about 5.5 million motorized vehicles in Jakarta that they have emitted more than one million tons of nitrogen oxides. Moreover, Zhu et al. (2020) found positive correlations of PM10, PM2.5, NO2, O3, and CO with COVID-19 cases. High levels of those pollutants could exacerbate and increase the risk of death of COVID-19 patients because of cardiovascular and pneumonia diseases (Barnett-Itzhaki and Levi 2021; Rendana and Idris 2021).

Therefore, in this study, we analyzed the changes in air pollutants levels (PM2.5, SO2, NO2, CO, and O3) during the COVID-19 period (during and after the LSSR periods). Because under the LSSR situation the emissions from usual sources have been diminished and it was important to identify the remaining pollutants came from and also their pathways. The association between the air pollutants such as SO2, PM2.5, CO, O3, and NO2 and meteorological factors were also prominent to be studied since severe air pollution generally occurred in favorable meteorological conditions. It would give a new insight into the mechanisms of air pollutants in distinct meteorological situations. Furthermore, the study in changes air pollutants during the two different periods will assist to ensure the emissions sources that lead to the changes in air pollutant levels. There are many studies on air pollution during the lockdown period investigated around the world (Kumar 2020; Patel et al. 2020; Tobias et al. 2020; Kumari and Toshniwal 2020). In Jakarta, previous studies such as Pramana et al. (2020) and Pardamean et al. (2021) have analyzed the air pollutants concentration, but, they concentrated on the before and the lockdown phases. They also did not elaborate on the spatial variations of the diverse air pollutants. Thus, this current study is a preliminary investigation to analyze the association between the meteorological factors and air pollutants levels during and after the LSSR periods in Jakarta. These outputs could be used to arrange effective control strategies in solving the air pollution issue in Jakarta and other neighboring cities.

Materials and methods

Study area

Our study location is Jakarta city, it is the capital and populous city in Indonesia. It lies on the northwest coast of Java island and expands approximately 664.01 km2 (Fig. 1). Jakarta is the core of the culture, entertainment, politic, and economy of Indonesia. The current population of the city is around 10,562,088. Since the COVID-19 outbreak spread in Indonesia on March 2, 2020, Jakarta becomes an epicenter of COVID-19. This city has the greatest number of COVID-19 cases with more than a total of 7000 cases as reported by the Indonesian Ministry of Health on May 31, 2020. Therefore, as a control measure, the local government has implemented a large-scale social restriction policy to control the spread of COVID-19 in the city. The restriction policy has a good impact on suppressing the COVID-19 cases even though it will lead to socio-economic setbacks.

Fig. 1
figure 1

Location map of the Jakarta city

Full size image

Data sources

The satellite-based observation data for nitrogen dioxide (NO2), sulfur dioxide (SO2), carbon monoxide (CO), ozone (O3), and PM2.5 over Jakarta city were obtained from the website (https://giovanni.gsfc.nasa.gov/giovanni/). The NO2 data was measured by the Ozone Monitoring Instrument (OMI) aboard the NASA Aura satellite with spatial resolution 0.25° × 0.25°. The SO2 and PM2.5 data were measured by the MERRA-2 satellite with spatial resolution 0.5° × 0.625°. While the CO and O3 data were assessed by the Atmospheric Infrared Sounder (AIRS) on the NASA Aqua satellite with spatial resolution 1° × 1°. A total of four satellite data were obtained for each pollutant that was divided into four months (May–June during LSSR and July–August after LSSR period). All air pollutants data have been available in the form of raster data. The raster data were then processed using ArcGIS software to improve the clarity of the raster images. The air pollutants data from the satellite were compared with the air quality monitoring station around the study area. A good accuracy result has been acquired with R2 = 0.88, RMSE = 0.12 and MAE = 0.08. This indicated that the satellite data could be used to evaluate the changes in air pollutant levels. In addition, the data on meteorology were collected from the Meteorology, Climatology, and Geophysical Agency of Indonesia. The data included temperature, humidity, rainfall, sunshine hour, and wind speed that all provided in daily data. The meteorological data retrieval time was adjusted for air pollutant data during and after LSSR periods. This was important to link meteorological and air pollutant data.

Correlation between air pollutants and meteorological factors

The Pearson correlation test was carried out to analyze the association between the five air pollutants (NO2, PM2.5, O3, CO, and SO2) and the five meteorological factors (humidity, temperature, rainfall, sunshine hour, and wind speed) using the IBM SPSS Statistic 21 application.

Backward trajectory model

The backward trajectory model of air pollutants was analyzed using the HYSPLIT model ver.5.1.0. This study applied the Global Data Assimilation System (GDAS) one-degree archive tool for analysis and it set the backward trajectories for 48 h with different altitudes (100 m, 500 m, and 1000 m) which used the vertical velocity model method with 6 h interval.

Results and discussion

Changes in NO2, O3, CO, SO2, and PM2.5 concentrations

Globally, fossil fuel and biomass burning activities were the major sources of particulate matter while the gaseous pollutants were primarily produced by power plants (Zhang et al. 2017). The local government has implemented a lockdown (known as the large-scale social restriction) as a control measure to mitigate the COVID-19 spread in Jakarta. One of the restricted actions was socio-economic activity. During the socio-economic restriction, every business on large, medium, and small scales were deferred, local transport, offices, universities, schools, amusement places, restaurants, market, shopping centers were all shut. However, pharmacies, daily stores, and banks remained open with tight health protocols. These all situations directly contributed to the socio-economic decline. But besides that, it had good effects on air pollution, energy storage, and certainly less COVID-19 spread. Several studies have reported the COVID-19 lockdown had a beneficial impact in decreasing various air pollutants and greenhouse gaseous most of the time as they were co-emitted from the same activities (Ali et al. 2021; Lu et al. 2021).

In this study, a significant increase was recorded after the LSSR period in mean concentrations of O3, PM2.5, NO2, and CO were 59.4%, 21.2%, 16.2%, and 1.0%, respectively (Table 1). This result was in line with a previous study by Ikhlasse et al. (2021) that found an increase in NO2 (42%) and O3 (21%) concentrations after the lockdown period in France. In contrast, our study observed a reduction in the mean concentration of SO2 that was 19.1% after the LSSR period. We run the back trajectory analysis to know the cause of the reduction, we found the pollutants transported from east and central Java areas where these areas were implementing the lockdown, many industries were closed, only vital power plants that operated during this period thus it leads to the reduction of SO2 emissions from the areas. The SO2 value in our study was significantly higher than the value of SO2 in France during the same period with a value of 3.4% (Ikhlasse et al. 2021). In the Republic of Korea, Hu et al. (2021) found a similar result where they found a reduction of SO2 with a value of 12.1%. Contrarily, in Wuhan, China, the SO2 levels were found to increase after lockdown with a value of 3.8%. The increased SO2 was estimated because urban areas and industries in China were mainly burning without emission control tools such as uncontrolled metal processing plants which burned sulfide ores to produce metal oxides (Hu et al. 2021). By using spatial analysis, we found a notable increase in the mean concentration of NO2 (2.12 to 3.38 NO2 × 1015 molecule cm–2) and O3 (24.92 to 28.98 ppbv) over the south of Jakarta from during to after the LSSR period. While a significant decrease in the mean concentration of SO2 (19.57 to 15.82 µg/m3) was observed in the northeast and southeast of Jakarta (Fig. 2). A substantial decrease could result from the shutting down of the essential sectors such as thermal power plants and other industries in the area.

Table 1 The mean concentrations of air pollutants in Jakarta during LSSR (May–June 2020) and after LSSR (July–August 2020) periods
Full size table
Fig. 2
figure 2

Satellite-retrieved average levels of a SO2, b NO2, c CO, d O3, e PM2.5 for the months May–June 2020, f SO2, g NO2, h CO, i O3, and j PM2.5 for the months July–August 2020 in Jakarta

Full size image

The spatial distribution of PM2.5 concentration during the LSSR period was higher in the northeastern part of Jakarta with the mean concentration was 33 μg/m3, then it increased to 48 μg/m3 after the LSSR period (Fig. 3). In contrast, another study in France reported that PM2.5 levels decreased after the lockdown period with a value of 37.9% (Ikhlasse et al. 2021). A similar result was also shown in several cities in India (Roy and Singha 2021). According to Chauhan and Singh (2020) that the alteration of the level of PM2.5 relied on some factors such as dust events, seasonal distribution, traffic, and other human activities.

Fig. 3
figure 3

The 48 h backward trajectory of air transport pathway a during the LSSR and b after the LSSR periods in Jakarta. AGL: Above Ground Level

Full size image

A substantial increase in PM2.5 could be caused by the increase in levels of NO2 and CO emitted from vehicles and power plants (Carslaw et al. 2019). Additionally, because of the high population density in eastern Jakarta city, the emission from the residential sectors could also be a major source of high PM2.5 concentrations. Overall, the mean concentration of PM2.5 has increased by 21% (Table 1). Several studies revealed the increase in PM2.5 had a direct association with the application of lockdown policy (Mo et al. 2021; Das et al. 2021), we also believed, the increased emission scenario could be resulted from loosening LSSR policy such as movement control and other industrial activities. In Jakarta, the location of industries concentrated in the northern part of Jakarta has shown the areas with the highest PM2.5 levels. The recorded PM2.5 values from the districts in the Jakarta city after the LSSR period showed significantly higher than the lockdown period for all of the districts (Fig. 2). The reasons for such high PM2.5 values in this area were emissions from construction sites, fields, unpaved roads, smokestacks, or fires.

Association between air pollutants and meteorological factors

Table 2 showed the relationship between the level of air pollutants and meteorological factors analyzed by the Pearson correlation test. The level of SO2, PM2.5, and CO were positively correlated with temperature during the COVID-19 period. The SO2 showed the greatest correlation coefficient followed by CO and PM2.5 (Table 2). However, NO2 and O3 were negatively correlated with temperature. These results were consistent with another study by Sulaymon et al. (2021) in Wuhan, China during the lockdown period.

Table 2 The correlation between the mean values of each meteorological variable and each air pollutant concentration during LSSR (May–June 2020) and after LSSR (July–August 2020) periods
Full size table

The correlation between NO2, PM2.5, O3, and CO was strong while the SO2 had moderate correlations with humidity. During the COVID-19 period, only PM2.5 and NO2 had a strong correlation with rainfall, while other pollutants showed a weak correlation. Furthermore, SO2 and NO2 showed a high negative correlation between sunshine and wind speed. PM2.5, CO, and O3 showed a weak correlation with wind speed. The wind condition could govern the pollutant distribution. The air pollutants would be more congested when the area had lower wind speed (Coccia 2021).

Furthermore, all pollutants had a strong relationship with temperature during and after the LSSR periods. The wind speed was strongly related to NO2 and SO2 but weakly related to PM2.5, CO, and O3 (Table 2). The association between humidity and all air pollutants after the LSSR period varied from weak to moderate correlation. Only SO2 and NO2 showed a high negative correlation with rainfall and sunshine, while other pollutants showed a weak correlation after the LSSR period.

Backward trajectory analysis

This study used the central coordinate of Jakarta (6.2088° S, 106.8456° E) to estimate the backward trajectory of air pollutants during and after LSSR periods from May to August 2020. Our study used the Global Data Assimilation System (GDAS) one-degree horizontal resolution dataset for analysis and it set the backward trajectories for 48 h with different altitudes (100 m, 500 m, and 1000 m) which used the vertical velocity model method with 6 h interval. Another study by Su et al. (2015) found a significant difference between trajectories resulted from GDAS1 (one-degree horizontal and temporal resolution of 3 h) and GDAS0P5 (0.5° horizontal resolution and temporal resolution of 3 h) datasets. It can be caused by the difference in vertical motion calculations and the lack of vertical velocity in GDAS0P5.

Figure 3 showed the 48 h of average backward trajectories of the air pollutant transport in Jakarta. The air mass of trajectory (Trj) 1, 2, and 3 were all from short-distance transport from the southeast, with the transport trajectory ratio of 35%, 33%, and 32%, respectively. Overall, the transport trajectory ratio was not much different during and after LSSR periods. For instance, trajectory 3 that was from the east of Java, transported from southeast to Jakarta, had the lowest portion of all trajectories. The height altered during the pollution transport of trajectory 3 was low and hold below 1000 m. However, trajectories 1 and 2 slowly raised from a vertical height of 100 m to 3000 m. The transport trajectory of pollutants at the height of 100 m was the same as 500 m. If we compared the air trajectory transport during LSSR with after LSSR periods, the air transport pathway indicated almost the same pattern. Only after LSSR, air transport pathway differed in the height of all trajectories (1, 2, and 3) tended to be lower and maintained below 1000 m.

The source of air mass during and after the LSSR periods was not so scattered, with more than 30% of air mass coming from east Java, bringing the lower number of pollutants to Jakarta along the course. Trajectory 1 and 2 transported a higher number of pollutants originating from areas in central Java and other surrounding areas (68%). The ratio of this air mass was high enough, it largely traversed through dense areas with many factories, industries, and other emission origins. These activities contributed to more pollutants being carried to Jakarta along the pathway that aggravated the air quality in Jakarta. The air mass after LSSR mostly came from central Java had the highest proportion (69%) and had the longest distance. Meanwhile, the air mass of trajectory 3 passed over the Java Sea and most of the particles reacted with water vapor through dilution process.

According to Fig. 2, NO2, O3, and PM2.5 tended to exhibit a same pattern. In contrast, SO2 and CO showed a slight distinct pattern. There were several possible factors that could affect this occassion. Firstly, the characteristic of air pollutant, high level of SO2 was mainly observed in industrialized areas, such as Jakarta city, as well as in other neighbor cities. As for CO, the main sources were originated from the combustion process, high CO levels were usually same as SO2, which was found in densely urbanized areas.

Renuka et al. (2020) explained the concentrations of SO2 and CO in the lower atmosphere were not in abundance, accounting for the season and region, both could not be maintain for long period. This interval was too concise for these gaseous to disseminate globally. Thus, in neighboring areas, where both high and medium emissions of SO2 and CO were observed, a significant difference in the levels of SO2 and CO could occur in the atmosphere. Furthermore, SO2 and CO usually emitted in developed industrial areas, and also had a short distance transport from those areas with heavier air pollutants to neighbor areas, as presented by the backward trajectory analysis. Unlike SO2, high particulate matter (PM) concentration was mainly found in similar areas where NO2 and O3 had high concentrations, but due to the much longer residence time of aerosol particles in the atmosphere, they tended to be carried over long distances thus their concentrations were higher than other gaseous like SO2 and CO.

Conclusions

The impact of LSSR on air pollution due to the COVID-19 pandemic in Jakarta was assessed by comparing the levels of air pollutants from May to August 2020 using satellite-based data. During the implementation of the LSSR policy, the spatial variation of air pollutant concentrations in Jakarta was low. The restriction policy has helped to reduce emissions of greenhouse gases and aerosols caused by human activities. After the LSSR period, the spatial variation of NO2, PM2.5, O3, and CO levels raised by 59.4%, 21.2%, 16.2%, and 1.0%, respectively over Jakarta. The air transport pathway were all from short-distance transport from the southeast to Jakarta, it was mainly from east and central Java areas. The increase in NO2, PM2.5, O3, and CO levels quickly after the LSSR was an obvious hint that control measures must be carried out to improve air quality. Otherwise, we would revert to the same situation before the pandemic.