1 Introduction

Air quality is deteriorating due to anthropogenic activities (e.g. vehicle exhaust, industrial emissions, fossil fuel combustion, resident, and household smoking and heating) (Feng et al. 2021; Ismail et al. 2021). Exhausts from motor vehicles cause about 50% of air pollution in Iran (Ghaffarpasand et al. 2020). Moreover, motor vehicles are responsible for about 36% of NOx concentrations in the atmosphere of china (Zhong et al. 2018). In Brunei Darussalam, 88% of the total annual CO comes from transport, while steel smelting contributed about 38% of the total annual CO emissions in China (Lee et al. 2018). Fossil fuel combustion contributed about 16% and 32% of PM10 and PM2.5 emissions in China, respectively (Jiang et al. 2020). Recently, Ji et al. (2020) stated that the transportation of NOx emissions is the major cause of O3 photochemical formation. However, emission data for Arab countries are scarce and fragmentary (Butenhoff et al. 2015; Harrison et al. 2016a, b; Basahi et al. 2017; Haiba and Hassan 2018). Air quality is determined by measuring the concentrations of PM10, PM2.5, Carbon monoxide (CO), Nitrogen dioxide (NO2), and ambient ozone (O3) (Chhikara and Kuma, 2020; Said et al. 2022). Variations in the concentration of different air pollutants could be detected by analyzing these air quality parameters. Moreover, meteorological conditions (e.g. temperature, wind speed, wind direction, and relative humidity) could affect the concentration levels of different air pollutants in a given period (Tello-Leal and Macias-Hernandez 2021). Economic and social activities were disrupted worldwide by the outbreak of the novel coronavirus (COVID-19), which was declared a pandemic on March 11, 2020 (WHO 2020a; b, c). This pandemic resulted in a tremendous lifestyle change. Many countries around the globe responded in an unprecedented way to protect the public health of their nations (e.g., quarantine, curfews, travel restrictions, etc.) (Rohrer et al. 2020; Farahat et al. 2021; Selcuk et al. 2021; Saddique et al. 2021; Hassan et al. 2021).

Jeddah is the second most polluted city in Saudi Arabia (Hassan et al. 2013; Harrison et al. 2016a, b). Air quality has mostly surpassed the air quality standards recommended by WHO and local authorities (National Centre of Meteorology "NCM") (Basahi et al. 2017; Harrison et al. 2016a, b; Ismail et al. 2021). Recently, Ali et al. (2021) stated that climatic indicators (wind speed, humidity, temperature, rainfall, and air quality) are strongly correlated with COVID-19 (Siciliano et al. 2020; Xu et al. 2020; Dogan et al. 2020; Fattorini and Regoli 2020; Ding et al. 2021).

In Saudi Arabia, due to the increase in positive cases, the government introduced a series of measures and strategies on March 22, 2020 to restrict human activities, as a proactive step, to prevent the spread of COVID-19. These include social distancing, the suspension of ritual activities and prayers in all mosques (including the two holly mosques), the closure of all schools, universities, and shopping malls, traffic and travel restrictions, and the suspension of construction sites. Moreover, the pandemic caused a sharp reduction in the number of visitors and pilgrimages (1000 visitors per day during the lockdown compared to about 2,000,000 during 2019 i.e., before the pandemic) (SPA 2021; Farahat et al. 2021).

Lockdown and reductions in human activities improved air quality in many countries worldwide (Aman et al. 2020; Bauwens et al. 2020; Gupta et al. 2020; Bashir et al. 2020; Chhikara and Kumar 2020; Marquès and Domingo 2022). On the other hand, it harmed the economy (e.g. travel restrictions, suspension of industries and construction, and a drastic collapse of almost all sectors). Although several studies were conducted in the USA, Asia, and Europe, to the best of our knowledge, no such study has been conducted in Saudi Arabia. Therefore, this study was carried out to fill this gap of knowledge.

The present study aimed to investigate the impact of the Lockdown and the reduction of economic activities during the pandemic COVID-19 on air quality in Saudi Arabia. Moreover, it was aimed to investigate the changing pattern of different pollutants before and during the lockdown as well as after the lockdown was eased.

2 Materials and Methods

The lockdown in Saudi Arabia came into effect on March 22, 2020, and was eased on June 29, 2020. Therefore, this study was carried out to investigate the changing pattern of different pollutants during the months before the lockdown (January 1–March 21, 2020), the months during the lockdown (March 22–June 28, 2020), and the months after the lockdown was eased (June 29–December 31, 2020).

2.1 Sampling Site and Environmental Indicators

Jeddah is located on the Red Sea coast of Saudi Arabia (N 21° 67´, E 39° 15´). The meteorology is characterized by frequent dust storms, with limited rainfalls, and therefore, the climate is generally dry and warm (Hassan et al. 2013). Traffic is the primary mobile source of air pollution in Jeddah, while the main stationary sources include a desalinization plant, an oil refinery, a major harbor, and a power generation plant.

The environmental indicators (CO, NO2, O3, PM10, and PM2.5) have been considered to evaluate the deviation of air quality during the lockdown, pre-lockdown, and post-lockdown periods. The concentrations of these parameters have been compared between the lockdown session during 2020, and a similar time in the preceding year (i.e. 2019).

The sampling site was located in a residential area of Jeddah, at a height of 11 m above the ground level on the roof of a residential building. PM samples were collected onto Quartz microfiber filters (47 mm) using a sequential air sampler (Partisol Plus 2025-D, Thermo Fisher Scientific, Waltham, MA, USA) (Ismail et al. 2021). Filters were weighed before and after sampling, using a micro-balance (Model XPE206DR, Mettler Toledo, Muntinlupa, Philippines), to determine the mass concentrations of the PM2.5 and PM10 samples. Samples were collected daily for a 12-month period starting January through December 2020 covering the different four seasons as well as the periods before, during and post the lockdown. Each sample was collected every day for 24 h at flow rates of 15 and 10 L min−1 for PM10 and PM2.5, respectively. Different environmental parameters were recorded simultaneously (atmospheric pressure, atmospheric temperature, wind speed, and relative humidity).

Ambient ozone (O3), NOx, and CO concentrations were monitored simultaneously (8:00—20:00 h, Saudi local time) using a fully integrated air quality monitoring system POLLUDRONE SAMRT (Modbus, RS—485, Savoie Technol. France) that was located adjacent to the Partisol on the top of the residential building (Ismail et al. 2021).

2.2 Statistical Analysis

A one-way analysis of variance (ANOVA) was performed, using the SATATGRAPHICS statistical software package, to examine the temporal variations in PM10, PM2.5, CO, NO2, and O3 before, during, and after the lockdown.

The relationship between the concentrations of different pollutants and meteorological parameters (temperature, relative humidity, and absolute humidity) was assessed using correlation analysis.

To quantify the variations in air quality, in terms of either concentration, the relative change (RC) is defined using an Eq. that was suggested by Ji et al. 2020 (Eq. 1)

$${\text{RC}} = \frac{{C_{i} - C{}_{{{\text{ref}}}}}}{{C_{{{\text{ref}}}} }} \times 100,{\text{RC}} = \frac{{C_{i} - C{}_{{{\text{ref}}}}}}{{C_{{{\text{ref}}}} }} \times 100,$$
(1)

where Ci is the air pollutant concentration or index value in year i, and Cref is the value in the reference year 2019. Therefore, RCs indicate variations in a certain year relative to the value in the reference year. Moreover, the influence of the lockdown period and meteorological conditions on changes in the concentrations of pollutants was determined using correlation base Principal Component Analysis (PCA).

The dataset was log-transformed prior to analysis to be sure that they were normally distributed. This was verified by the Shapiro–Wilk normality test (Menebo 2020).

3 Results and discussion

Temporal variations of PM2.5 (Fig. 1), PM10 (Fig. 2), and NO2 (Fig. 3) were observed in air quality samples collected daily for 1 year from January to December 2020, with the elevated concentrations observed in the pre- and post-lockdown periods. During the lockdown (March 22–June 28, 2020), concentrations were significantly lower than before and post lockdown periods. The annual concentrations of PM2.5, PM10 and NO2 were 21.9 μg m−3, 123.08 μg m−3 and 33.57 nl l−1, respectively, during the sampling period (Figs. 1, 2, and 3).

Fig. 1
figure 1

Temporal variations in the concentration of PM2.5 (µg m−3). Red lines state the beginning and the end of lockdown in Jeddah city

Full size image
Fig. 2
figure 2

Temporal variations in the concentration of PM10 (µg m−3). Legend as Fig. 1

Full size image
Fig. 3
figure 3

Temporal variations in the concentration of NO2 and O3 (nl L−1). Legends as Fig. 1

Full size image

The average PM2.5 concentration was 22.2 µg m−3 during the pre-lockdown period (January, February and the first 3 weeks in March 2020). This concentration was declined during the lockdown period (April –July 2020) and reached 15.5 µg m−3 (Table 1 and Fig. 1). The lockdown resulted in a 31.7% reduction in PM2.5 concentration (Table 2). PM2.5 concentration increased to 23.42 µg m−3 in the post-lockdown period (Table 1). Similarly, PM10 concentrations were 128.8 and 82.40 µg m−3 in pre-lockdown and during lockdown periods, respectively (Table 1 and Fig. 2). The PM10 concentration increased in post-lockdown period (117.06 µg m−3) (Table 1). Table 2 shows that the lockdown caused a 36% reduction in PM10 concentration. The PM2.5/PM10 ratio did not vary significantly in the pre-, during and post-lockdown periods of COVID-19, indicating insignificant changes in coarse particles due to reduced emissions from anthropogenic activities (Feng et al. 2021). Weather conditions were almost the same before and during the lockdown, confirming the stability of the PM2.5 fraction of PM10. The restriction on travel, suspension of industrial and economic activities caused a significant reduction in primary emissions during the lockdown, i.e. limited oxidation in the atmosphere (Feng et al. 2021).

Table 1 Concentrations of different pollutants during the different periods of COVID-19 lockdown (means + SE)
Full size table
Table 2 The impact of COVID-19 lockdown on the during lockdown/pre-lockdown ratio of different air quality parameters in different regions over the world
Full size table

CO and NO2 concentrations dropped from 5.39 mg m−3 and 43.8 nl L−1, respectively, in the pre-lockdown period to 3.26 mg m−3, and 16.30 nl L−1, respectively, during the lockdown period (Table 1). Subsequently, they increased to 5.98 mg m−3 and 29.84 nl L−1 in the post-lockdown period, respectively (Table 1). The lockdown resulted in a 39.58% and 62.78% reduction in the concentrations of CO and NO2, respectively, when compared with concentrations in the pre-lockdown (Table 2). The reduction in pollutants due to the containment measures implemented by the COVID -19 pandemic was very significant. The concentration of CO decreased by about 40% during the closure, with combustion processes being the main emitters, in our case road traffic in urban areas, especially diesel and gasoline vehicles. In addition, Table 2 shows a simplified comparison of the effect of the COVID-19 Lockdown implementation on air quality in selected cities around the world. It clearly shows that the lockdown measures had a positive effect on air quality around the world. All pollutants studied exhibited significant reductions during the lockdown, with the exception of O3. Recently, El Sheekh and Hassan (2021) found that the reduction in concentrations of pollutants was significantly (p ˂ 0.01) correlated with the decrease in transportation, industrial processes and other economic activities in Egypt. The main contributors to air pollution (vehicular emissions, various industrial processes, and commercial activities) were adversely affected by the lockdown measures worldwide (Bauwens et al. 2020; Feng et al. 2021; Viteri et al. 2021). The results of the present study are consistent and in agreement with previous studies that travel restrictions during the lockdown caused the most significant reduction in NO2.

The average mass concentrations of PM2.5 and PM10 during the sampling period exceeded the annual average levels recommended by the World Health Organization (WHO) for PM2.5 and PM10 (15 μg m−3 and 25 μg m−3, respectively) (Table 3). Moreover, the recorded levels exceeded the recommended levels by Environmental Protection Agency (EPA) for PM2.5 and PM10 (15 and 50 μg m−3, respectively). However, PM2.5 levels were lower than the levels recommended by National Center for Meteorology (NCM) (30 μg m−3), while PM10 concentrations were higher than the levels recommended by NCM (80 μg m−3) (Table 3).

Table 3 Annual air quality standards
Full size table

On contrary to the other pollutants, O3 levels were higher (51.35 nl L−1) during lockdown than the pre- and post-lockdown concentrations (32.39 and 33.36 nl L−1, respectively) (Table 2). The average concentering of O3 during the sampling period was 32.69 nl l−1 (Fig. 3). The increased O3 concentration during the lockdown, compared to the pre-lockdown levels, is an evidence of the occurrence of the secondary chemical reactions in the atmosphere (Feng et al. 2021). Sicard et al. (2020) found that ambient O3 levels were amplified in many cities around the world during the COVID-19 pandemic, and they owed this increase in O3 levels to the reduction in NO2 levels. Hassan et al. (2013) found that reductions in NO2 concentrations resulted in an increase in ambient O3 levels in Jeddah, indicating the NO scavenging effect on O3 (Jain et al. 2021). Chang et al. (2018), found similar results in Malaysia, they found that the ozone concentration was somehow higher during the weekend and afternoon hours which was associated with lower NO2 concentrations at that times. Recently, Shi and Brasseur (2020) found that the decrease in ambient NO2 was associated with an increase in tropospheric O3 in China. Moreover, Xu et al. (2020) stated that the formation of O3 in the urban environment involves complex processes with no direct source of emission but is produced through reactions involving NOx, CO, and volatile organic compounds which serve as precursors. The lower O3 concentration after the lockdown was due to the more favorable weather conditions and the weakening of secondary reactions, which can also explain the variations in PM2.5 and PM10 in this study. Chu et al. (2021) found that the concurrent reductions in NOx and PM2.5 concentrations led to an increase in O3 concentrations across China during the COVID-19 pandemic. They suggested coordinated control of other pollutants. The elevated O3 levels could be directly attributed to the relative high air temperature (Hassan et al. 2013) and low wind speed during the COVID-19 lockdown in 2020 as well as to the emission sources (Feng et al. 2021). Ozone pollution in Jeddah must be further investigated in a future study.

It was observed that the levels of PM2.5, PM10 and NO2 were tremendously increased on September 23 (Figs. 1 and 2). These elevated levels of these pollutants coincided with the National Day of Saudi Arabia when Saudis celebrate on the streets and drive their cars. The elevated concentration of these pollutants are ascribed to an increased traffic on September 23. In contrast, the concentration of O3 decreased on September 23 (Fig. 3).

Table 4 shows that PM2.5, PM10 had a very strong positive correlation with each other (r = 0.99, P < 0.001). In addition, statistically moderate positive correlations were found between PM2.5 and CO (r = 0.66, P ˂ 0.05) and between PM10 and CO (r = 0.74, P ≤ 0.01). However, a correlation between particulate matter and NO2 and O3 was insignificant (P ˃ 0.05) (Table 4). Furthermore, the statistical analysis of wind speed (WS) revealed a very strong correlation (0.89 ≥ r ≥ 0.91) with PM factions and other pollutants. There was a strong negative correlation coefficient between WS and O3 (r =  − 0.79, P < 0.01) (Table 4). Temperature exhibited a consistently moderate correlation with all the air pollution variables (0.71 ≥ r ≥ 0.79), except for NO2, the relationship was insignificant (P ˃ 0.05). The high correlation of PM2.5 with PM10, CO, and NO2 during the lockdown, reflecting the common origin of these species from fossil fuel combustion (Wang et al. 2014). Moreover, the positive correlations between O3, and PM2.5 as well as PM10, indicating the simultaneous formation of secondary O3 and PM by photochemical reactions under favorable weather conditions. Similar correlations among pollutants were also observed interannually, but the correlations were slightly weaker than those in the lockdown. Our results are consistent and in agreement with those of several studies that demonstrate significant relationships between meteorological conditions (temperature, humidity, and wind speed) and air pollutants (PM2.5, PM10, CO NO2, and O3) (Bolâno-Ortiz et al. 2020; Dogan et al. 2020; Tello-Leal and Macias-Hernandez 2021, Said et al. 2022).

Table 4 Pearson correlation coefficient matrix for main variables using a dataset during the lockdown
Full size table

Table 5 shows non-significant variation between meteorological conditions (WS, T and RH) before and during the lockdown, except for relative humidity. Non-significant variation indicates that the decrease in air pollution during the lockdown period was not solely dependent on meteorological conditions (Navinya et al. 2020). However, significant variation in relative humidity could be attributed to the coastal weather of Jeddah with its warm sea (Qari and Hassan 2017). Moreover, the lockdown was implemented between the spring and summer season (April–July) and the slight rise in temperature (not significant, P ˃ 0.05) is linked to the timing of the season. Patlakas et al. (2019) reported that the temperature is strongly correlated with the increase in the percentage of relative humidity in the atmosphere (Cichowicz et al. 2020; Hassan et al. 2021).

Table 5 Variation in meteorological parameters (mean ± SE)
Full size table

Although pollutant concentrations decreased during the lockdown, the concentrations of PM2.5, PM10, NO2, and O3 were still above WHO annual mean limit levels. This confirms that not only emission from traffic that contribute to increases in air pollutants, but also stationary sources from the industries, with fossil fuel combustion playing a key role in the complex mix of sources (Kerimray et al. 2020). Feng et al. (2021) stated that emissions from stationary sources (e.g. coal-fired power factories) were not reduced compared to emissions from mobile sources (e.g. traffic). Desalination plants in Jeddah were working with full capacity to provide Saudis with the necessary amounts of drinking water, and this could be a major and an important source of pollution emission in Jeddah. Moreover, electricity and heat were still provided as usual by the thermal power plants to ensure normal supply during the lockdown. One could argue that the systematic analysis of the data has limitations because of the general seasonal characteristics of the region and the corresponding months of the previous year. However, we analyzed air quality index (AQI) in the corresponding months of the previous year (2019) using the equation of Ji et al. (2020) to test this hypothesis. The relative changes (RC) exhibited decreases AQI (negative RC), which implied improved air quality during the lockdown in 2020 (Fig. 4). Moreover, Principal Component Analysis (PCA) for different pollutants and meteorological conditions in 2019 and the same period in 2020 during the lockdown period is presented in Fig. 5. Component 1 and 2 accounted for 57.11%, and 54.05%, of the total variation in 2019 and in 2020, respectively. There was an influence of COVID-19 lockdown on PM10, CO, and O3 with also a positive correlation between these pollutants and temperature (Fig. 5A). On the other hand, PCA for 2019 showed a different situation, with a positive correlation with humidity and wind speed (Fig. 5B).

Fig. 4 
figure 4

Relative Changes (RCs) in seasonal mean AQI against the reference year 2019

Full size image
Fig. 5
figure 5

Principal component analysis for the influence of meteorological data and pollutants concentration before and during COVID-19 lockdown (A) and during 2019 (B)

Full size image

To the best of our knowledge, this study provides the first comprehensive analysis of the variations in PM2.5, PM10, CO, NO2, and O3 concentrations before, during, and after the lockdown in Jeddah, Saudi Arabia. The results of the present study clearly show that exhaust emissions from vehicles and industrial plants were tremendously decreased when the Saudis were enforced to be home quarantined to curb the virus spread. As a result, their rate of entry into the atmosphere was reduced. Generally, curbing anthropogenic activities could improve air quality.

4 Conclusions

In this work, the improvement of urban air quality due to the COVID-19 mitigation measures was investigated for the first time in Saudi Arabia. The time series for PM10, PM2.5, CO, and NO2 during quarantine showed a clear reduction of pollutant concentrations which is directly associated with anthropogenic emissions. The targeted emission controls significantly improved air quality in the city. Although meteorological conditions did not change significantly during the lockdown, they still affected changes related to pollutant concentrations.

Emission controls should be more stringent to ensure that air quality improvements are permanent and not temporary, e.g., by controlling traffic, and industry activities, as well as introducing green transportation programs such as green commuting programs.