Introduction

The coronavirus, namely COVID-19, was first identified in Wuhan City in the Hubei province in China in December 2019 after a local hospital warned about patients with pneumonia of an unknown etiology (Lu et al. 2020). This virus was first reported to the World Health Organization (WHO) on December 31, 2019, and has become a global public health problem. The COVID-19 pandemic has been deemed a global health crisis (Dantas et al. 2020; Sohrabi et al. 2020). By mid-June 2020, the pandemic encompassed more than 8 million infected people and nearly 450,000 deaths around the world. This pandemic has greatly affected human society, including health care, social relationships, and the structure of the economy as well as politics (Berman and Ebisu 2020).

The first confirmed COVID-19 case in South Korea on January 20, 2020, was a woman who returned from a trip to Wuhan, China, after which the speed of the spread of the secondary transmissions accelerated (Lim et al. 2020). The city of Daegu, South Korea, saw the biggest COVID-19 outbreak outside of China, with a total of nearly 6700 confirmed cases in March 2020, accounting for 90% of the total nationwide. Patient no. 31 of South Korea, who did not travel to any area affected by COVID-19, was the first confirmed case in Daegu. Following this patient, Daegu and the surrounding areas reported hundreds of new cases every day (Her 2020). However, South Korea was able to control the pandemic of new infections early on, maintaining low mortality rates. The effectiveness of the South Korean government actions was demonstrated by the development of a rapid test kit for the testing of a large number of people who were suspected to be positive with the COVID-19 virus, also indicating the high level of responsibility assumed by these people (Oh et al. 2020). To control the spread of COVID-19, the authorities of Daegu implemented various administrative regulations to prevent the outbreak of COVID-19. For example, on March 25, 2020, all citizens were requested to self-quarantine and maintain social distancing (Ma and Kang 2020).

Recent studies on the effects of a lockdown or social distancing on the air quality have attracted much interest. For instance, the association between the six criteria air pollutants (CAPs: PM2.5, PM10, SO2, CO, NO2, and O3) and COVID-19 patients in China has been investigated, and findings suggest that the short-term exposure to air pollutants and the risk of COVID-19 infection showed significant relationships (Zhu et al. 2020). The lockdown has led to a reduction of air pollutants mainly due to the low density of vehicles circulating in cities such as Milan, Italy, resulting in a sharp decrease in SO2 and NOx; however, O3 concentrations tended to increase because of low NO concentrations in the ambient air (Collivignarelli et al. 2020). This finding is consistent with another study in Barcelona, Spain (Tobías et al. 2020), in which O3 concentrations increased by up to 50%. In Sao Paulo, Brazil, the levels of pollutants at an urban site significantly declined during the lockdown period by 77% (NO), 54% (NO2), and 65% (CO) compared with a 5-year monthly mean (Nakada and Urban 2020). In addition, several previous studies have shown that the concentrations of PM2.5 and PM10 sharply decreased during the COVID-19 lockdown. These reductions of the CAP levels are one of the positive effects of the COVID-19 lockdown on the environment, not considering the negative ones on the socio-economy (Chauhan and Singh 2020; Mahato et al. 2020; Tobías et al. 2020).

In this study, the relationship between the new daily COVID-19 confirmed cases and hourly meteorological data (ambient temperature, relative humidity, and wind speed) in Daegu during the first 4 months of 2020 was investigated. The variation in hourly CAP (PM10, PM2.5, O3, NO2, CO, and SO2) levels before and during the pandemic outbreak was also evaluated.

Materials and methods

Daegu is one of the largest metropolitan cities in South Korea with a population of ~ 2.5 million and a total area of ~ 900 km2. In this study, all data were obtained for eight districts (Dong, Buk, Suseong, Seo, Jung, Nam, Dalseo, and Dalseong) in Daegu. Statistical data for the confirmed COVID-19 cases in Daegu classified by district were downloaded from the official websites of each administrative department (Table S1 in the Supplementary Information) for the period between February to April 2020.

To investigate the lockdown impacts on the air quality of Daegu before and during the outbreak, hourly CAP (14 stations: PM10, PM2.5, O3, NO2, CO, and SO2) and meteorological (one observatory: temperature, wind speed, and relative humidity) data were obtained from the Air Korea website (https://www.airkorea.or.kr) and the Korean Meteorological Administration (KMA) (https://data.kma.go.kr), respectively, for the first 4 months of 2020. The locations of the air monitoring stations and the meteorological observatory in Daegu are presented in Fig. 1 and Table S2. The information of measuring instruments for the CAPs and meteorological data is summarized in Table S3.

Fig. 1
figure 1

Locations of 14 air monitoring stations (black circle) and a meteorological observatory (blue triangle) in Daegu, South Korea

Full size image

The first COVID-19 confirmed case in Daegu, or patient no. 31 of South Korea, was reported on February 18, 2020. This patient had physical contact with other people for many days after showing COVID-19 symptoms, leading to the widespread pandemic in Daegu in the following weeks. Therefore, the study period for this paper is divided into (1) January 1–February 17 (before the outbreak) and (2) February 18–April 30 (during the outbreak). To understand the COVID-19 lockdown effects on the concentrations of CAPs, data during the outbreak in 2020 were also compared with those for the same months of 2019 and 2018.

The data distribution and correlations between the new daily COVID-19 confirmed cases and meteorological parameters were evaluated by applying the Shapiro-Wilk normality test and Pearson’s and Spearman’s correlation analyses, respectively, using SPSS Statistics 23.0 (IBM, USA). Tests to calculate statistically significant differences (t test and Mann-Whitney rank sum test) between the CAP data in different years were performed using SigmaPlot 12.0 (Systat Software Inc., USA).

Results and discussion

Overview of the COVID-19 pandemic in Daegu

Figure 2 shows the weekly spread of the COVID-19 pandemic during the outbreak period in Daegu. There were over 440 cases within one week (February 24) after the first confirmed case. The incubation of the virus lasts up to 14 days; therefore, the outbreak of COVID-19 in Daegu was at a peak in the second (March 2) and third (March 9) weeks. In particular, the number of confirmed cases significantly increased by 2600 and 2500 in the second and third weeks, respectively, leading to a total of over 5500 cases on March 9. However, efforts from the South Korean government, such as strict social distancing policies and advancements in medical technology, have controlled and prevented the rapid spread of the COVID-19 pandemic in Daegu. As a result, the new cases sharply declined from hundreds to tens during the following weeks in March and April. At the end of April 2020, a total of nearly 6900 confirmed cases were reported, accounting for ~ 60% of the cases throughout the entire country, when the evolution of the pandemic stabilized. A death rate of 2.5% (172 cases) from COVID-19 was reported for Daegu, indicating that the treatment methods and actions for the prevention of further spread taken by the government were highly effective.

Fig. 2
figure 2

Weekly data for COVID-19 confirmed cases and deaths from February 18 to April 30, 2020, in Daegu, South Korea

Full size image

Figure 3 illustrates the spatial distribution of the pandemic in Daegu during the outbreak period. The most affected districts were Dalseo, Nam, and Suseong, with the total number of cases for each district at approximately 1600, 1300, and 1100, respectively. The spread of COVID-19 in Daegu originated from patient no. 31, a 61-year-old Korean woman and a member of the Shincheonji religious organization. According to the Korea Centers for Disease Control and Prevention, the transmission pathways initiating from patient no. 31 were mainly within the Suseong and Nam districts. More specifically, patient no. 31 went to medical centers and hospitals in the Suseong district for a health check between February 7–17 and left the hospital to attend church in the Nam district on February 9 and February 16. At least 1000 other members of the Shincheonji Church were exposed to COVID-19 by patient no. 31. In addition, Dalseo (~ 560,000 people) and Suseong (~ 420,000 people) districts have the highest population in Daegu (Fig. 1). For this reason, Dalseo, Nam, and Suseong districts became the hotspots of the COVID-19 pandemic.

Fig. 3
figure 3

Spatial distribution map of the total COVID-19 confirmed cases from February 18 to April 30, 2020, in Daegu, South Korea

Full size image

Relationship between the new daily confirmed cases and meteorological conditions

The incubation period of COVID-19 symptoms normally lasts 14 days; hence, the meteorological conditions at the same time as (0-day lag), 7 days prior (7-day lag), and 14 days prior (14-day lag) to the outbreak were evaluated in this study. South Korea is located in the temperate zone of the northern hemisphere with winter-to-spring weather from February to April. In general, the weather conditions during these three periods are quite similar (Table 1) with low mean temperatures (8–10 °C), low wind speeds (2.6 m/s), and low relative humidity (58%).

Table 1 Hourly meteorological conditions (mean ± standard deviation) on the same day as (0-day lag), seven days prior (7-day lag), and 14 days prior (14-day lag) to the outbreak of the COVID-19 pandemic in Daegu, South Korea
Full size table

The human-to-human transmission routes of COVID-19 are the inhalation of aerosols or droplets (Jayaweera et al. 2020) and contact with fomites (WHO 2020). Virus-laden droplets and aerosols can travel under the effects of wind speed by distances from less than 2 m with normal exhalation (Chen et al. 2020; Ong et al. 2020) and more than 6 m with cough and sneeze (Jayaweera et al. 2020). However, the lifetime of COVID-19 is 1–5 and 1–18 h in aerosols and on the furniture surfaces (made of copper, cardboard, stainless steel, or plastic), respectively (van Doremalen et al. 2020). Therefore, an increase in wind speed at 0-day lag could have dispersed the virus and enhanced its transmission through human inhalation. Our result also indicates that wind speed was positively correlated to the new daily cases in Daegu at 0-day lag (r = 0.595, p < 0.01) (Table 2), which is also reported in previous studies (Bashir et al. 2020; Coşkun et al. 2021). Under low wind speeds, virus-laden droplets may deposit quickly on the surface of clothes and/or furniture, favoring the transmission through contact with fomites. Thus, there was a significantly negative correlation between the new daily cases and wind speed at 7-day lag (r = − 0.246, p < 0.05) (Table 2).

Table 2 Spearman’s correlation analysis between the new daily COVID-19 confirmed cases and meteorological parameters at the same time as (0-day lag), 7 days prior (7-day lag), and 14 days prior (14-day lag) to the outbreak of the COVID-19 pandemic in Daegu, South Korea
Full size table

According to a previous study on the regional climate impacts, the transmission of COVID-19 was highly favored in temperate and dry zones; ambient temperature and relative humidity showed significantly negative correlations with new cases (Méndez-Arriaga 2020; Qi et al. 2020). These findings are in line with those in our study. The new daily cases in Daegu were negatively correlated to mean temperature and relative humidity at 0-day lag (r = − 0.400 and r = − 0.316, p < 0.01) and 7-day lag (r = − 0.244, p < 0.05 and r = − 0.316, p < 0.01), respectively (Table 2). Although temperature and relative humidity played a vital role in the spread of COVID-19, the results in our study are different from those in a previous study (Auler et al. 2020); in tropical Brazil, high temperatures (28 °C) and intermediate relative humidity (80%) increased the spread of COVID-19. As the incubation period for COVID-19 is up to 14 days, the relationship between the new daily cases and meteorological conditions was somewhat unclear at 14-day lag. In particular, the correlations were not significant for wind speed and relative humidity (p > 0.05) and shifted from negative to positive for mean temperature (Table 2). In short, meteorological conditions (e.g., temperature, wind speed, and relative humidity) may significantly affect the transmission of COVID-19 (Bashir et al. 2020); however, the degree and direction of these effects may be different between tropical and temperate climates. These results indicate that the existence and evolution of the coronavirus may be complex. Further investigation is necessary to fully understand the behavior of this virus.

Lockdown effects on CAPs

The CAP concentrations during the outbreak of COVID-19 (February 18–April 30) in 2020 were compared with those for the same months in 2019 and 2018 in Daegu (Table 3). Except for the relatively constant level of O3 for all 3 years, all other CAPs showed clear reductions. These trends are consistent with those in previous studies (Abdullah et al. 2020; Collivignarelli et al. 2020; Dantas et al. 2020; Nakada and Urban 2020). In particular, there was a decrease of 24.3% (PM10), 29.6% (PM2.5), 20.0% (NO2), 9.51% (CO), and 11.5% (SO2) between CAP levels in 2020 and the average CAP levels for 2019 and 2018. Wind speed and precipitation between February 18 and April 30 in these three consecutive years were also considered to evaluate the impacts of meteorological conditions on the CAP levels in Daegu. Wind patterns (Fig. S1 in the Supplementary Information) were similar for all three years with an average wind speed of 2.4 m/s, whereas precipitation reduced from 237 mm in 2018 to 129 mm in 2019 and then to 87 mm in 2020. Therefore, the decreasing trend in CAPs may not be completely influenced by meteorological conditions. To investigate the impacts of the COVID-19 lockdown on the current changes in air quality in Daegu, the following results and discussion focus on the CAP levels between January 1–February 17, 2020 (before the outbreak) and February 18–April 30, 2020 (during the outbreak).

Table 3 Comparison of the CAP concentrations during selected periods in Daegu, South Korea
Full size table

Boxplots of the CAP concentrations before and during the outbreak of COVID-19 in Daegu are illustrated in Fig. 4. There were statistically significant differences between the CAP concentrations of these two periods (Mann-Whitney rank sum test, p < 0.01). PM10 slightly declined by 3.75% from 38.9 to 37.5 μg/m3. This reduction percentage is approximately one-third of that in Rio de Janeiro, Brazil (Dantas et al. 2020), and one-tenth of that in Milan, Italy (Collivignarelli et al. 2020). Asian dust, mostly originating from natural sources (e.g., deserts and loess areas), has been reported to abundantly appear during these months in South Korea (Choi et al. 2012; Kim 2008; Kim et al. 2009; Matsumoto et al. 2006). According to the KMA, in Daegu, there were several days with typically high levels of PM10, which are defined as Asian dust. For example, the maximum hourly concentrations of PM10 on February 22 (09:00–20:00, UTC+9) and March 19 (12:00–17:00, UTC+9) were 124 and 122 μg/m3, respectively. Therefore, the reduction of PM10 in Daegu during the spread of COVID-19 was less than that in Rio de Janeiro and Milan. Regarding PM2.5, its decreasing trend is more evident than that of PM10 with a decrease of 30.9% from 27.9 to 19.3 μg/m3 in Daegu and comparable with that of PM2.5 in many cities of Malaysia (with a decrease of 28.3%) (Abdullah et al. 2020) but less than that of PM2.5 in Milan (with a decrease > 40%) (Collivignarelli et al. 2020). The strict social distancing policy was implemented in Daegu to prevent human-to-human transmission of the virus, leading to a large limitation in the circulation of vehicles. Thus, NO2 and CO, which have been reported to be mainly produced by vehicular emissions (Clarke et al. 2014), reduced significantly by 36.7% from 25.8 to 16.3 ppb and 43.7% from 726 to 409 ppb, respectively. Previous studies have indicated that NO2 decreased by 47% in Milan (Collivignarelli et al. 2020) and 50% in Barcelona, Spain (Tobías et al. 2020), which are greater decreases than that in Daegu. However, the reduction of CO in Daegu is similar to that in Milan (46%) (Collivignarelli et al. 2020) and Rio de Janeiro (40%) (Dantas et al. 2020) and less than that in Sao Paulo, Brazil (65%) (Nakada and Urban 2020). Regarding SO2, a reduction of 21.3% from 3.4 to 2.7 ppb in Daegu is comparable with that in Milan (25%) (Collivignarelli et al. 2020). Although Daegu is not a major industrial city, its neighboring cities (e.g., Ulsan and Pohang) include the largest industrial zones in South Korea. During the outbreak of COVID-19, the closure of factories may have caused a decline in SO2 emissions in those cities, which in turn influenced the air quality of Daegu by short-range transport.

Fig. 4
figure 4

Concentrations of CAPs before (January 1–February 17, 2020) and during the COVID-19 pandemic outbreak (February 18–April 30, 2020) in Daegu, South Korea

Full size image

In South Korea, NO2 and CO have been reported to usually be higher from December to March in Daegu (Jo and Park 2005), Daejon (Jung et al. 2017), Seoul (Pandey et al. 2008), and Ulsan (Clarke et al. 2014), whereas the level of SO2 increases from November to February in Daegu (Jo and Park 2005) due to seasonal factors (e.g., low temperatures and increasing demand for heating). Therefore, the South Korean government introduced several policies at the end of 2019 to improve the air quality of the whole country. According to the Ministry of Environment, the Special Measures to Respond to High Levels of Fine Dust and the Fine Dust Seasonal Management System were introduced in November 2019 to reduce the intensity and recurrence of high concentrations of PM2.5 during the typical period of its occurrence (MOE 2020). In particular, the operation of coal-fired power plants has been suspended, and the upper limit restrictions on air pollutant emissions have been applied to many industrial facilities. Furthermore, the 5th Grade Vehicle Operation Restrictions (in terms of air pollutant emissions) and the Alternative No-Driving System of Public Vehicles have been issued for citizens and implemented in many metropolitan cities. These actions from the government may also be partly responsible for the improvements in air quality in Daegu during the outbreak of the pandemic in 2020.

O3 is known as a secondary pollutant produced from solar radiation and its precursors (e.g., NO2, CO, and volatile organic compounds—VOCs). Previous studies in Daegu (Jo and Park 2005) and Ulsan, South Korea (Clarke et al. 2014), have shown that O3 reaches its annual maximum level in the spring season (March, April, and May). Additionally, under the condition of a low VOC/NOx ratio in the urban city (VOC-limited), a decrease in NOx may lead to the suppressed titration of O3, increasing the formation of tropospheric O3 (Jo and Park 2005; NRC 1991; Park et al. 2018). In our study, the concentration of O3 significantly increased by 71.1% during the outbreak of COVID-19 compared with the period right before the outbreak with an observed reduction of 36.7% in NO2 (Table 3). However, O3 showed very similar levels on January 1–February 17 (~ 19 ppb) and February 18–April 30 (~ 33 ppb) in 2020, 2019, and 2018. An increase in ambient temperature, mostly because of the solar radiation, from 3.9 ± 4.2 °C (before the outbreak) to 10 ± 5.4 °C (during the outbreak) may enhance the photochemical formation of O3 (Mukherjee et al. 2020). In addition, generation pathways from natural processes have also been proposed to explain the increase in O3 during these three months (Monks 2000). For example, O3 has been reported to increase in the spring due to a phenomenon at the northern-hemisphere scale (Simmonds et al. 1997) and the intrusion of stratospheric O3 into the troposphere (Levy II et al. 1985; Tu et al. 2007). Therefore, the influence of the COVID-19 lockdown on the sharp increase in O3 is still unclear in our study. This pattern observed for O3 may be due to annual natural processes rather than the reduction of NO that consumes tropospheric O3 as mentioned in other studies (Collivignarelli et al. 2020; Tobías et al. 2020).

To investigate the spatial impacts of the COVID-19 lockdown on the air quality of Daegu, spatial distribution maps of the average CAP concentrations before and during the outbreak are shown in Fig. 5. The most affected districts by the pandemic (Dalseo, Nam, and Suseong districts) showed the most apparent reductions in the CAP levels. As explained above, an increase in the O3 concentration may be influenced by several factors other than the lockdown; therefore, this pattern is not discussed further. The average decrease in concentrations of PM10 (9.17%), PM2.5 (35.1%), NO2 (36.8%), and CO (50.3%) was the highest in the Suseong district (sites 2, 3, 9, and 12), whereas the highest decrease in SO2 (55.9%) was observed in the southeastern part of the Dalseo district (site 13) in Daegu. These tendencies can be explained by the restrictions on mobility and/or the partial and total lockdowns for all workplaces and educational facilities due to the rapid spread of COVID-19 in these three districts during the first half of March 2020.

Fig. 5
figure 5

Spatial distribution maps of the average CAP concentrations before (January 1–February 17, 2020) and during the COVID-19 pandemic outbreak (February 18–April 30, 2020) in Daegu, South Korea. Contour maps were drawn using ArcGIS 10.8 (ESRI Inc., USA), and inverse distance weighting (IDW) was applied as an interpolation method

Full size image

Conclusion

The COVID-19 pandemic struck Daegu, South Korea, resulting in a total of ~ 5500 cases within the first 3 weeks of its outbreak (February 18–March 9, 2020). The South Korean government responded quickly to the pandemic by imposing partial and total lockdowns of the city to prevent human-to-human transmission of the virus; therefore, the situation was stabilized in April 2020 with a total of ~ 6900 cases and a death rate of 2.5% (172 cases). Three central districts were the most affected areas by COVID-19, with 1600 cases in Dalseo, 1300 cases in Nam, and 1100 cases in Suseong. Meteorological conditions were vital factors for the spread of COVID-19. In particular, the new daily cases were significantly negatively correlated to the mean ambient temperature and relative humidity at both 0-day and 7-day lag, while the relationship between the new daily cases and wind speed was significantly positive at 0-day lag and negative at 7-day lag. Actions taken by the government to protect human health from the pandemic have also directly influenced the air quality of Daegu.

The concentration of PM10 during the outbreak only slightly decreased compared with that before the outbreak, whereas the levels of PM2.5, NO2, CO, and SO2 significantly decreased over the entire city. This decreasing trend was the most apparent in the three central districts (Dalseo, Nam, and Suseong). However, an increase in O3 was not clearly affected by the lockdown, but other factors were involved. In addition to the major effects of the lockdown, policies by the government to improve the air quality for South Korea since the end of 2019 have partly influenced the decreasing trend in CAPs.

In conclusion, the city lockdown has had a negative impact on the socio-economy but a positive impact on the air quality of Daegu. The findings in our exploratory study can serve as fundamental knowledge to both scientists and the government to understand the behavior of the virus and to prevent its rapid spread. As a further study, the influence of the lockdown on the levels of hazardous air pollutants (e.g., polycyclic aromatic hydrocarbons, metals, and VOCs) and their health risks needs to be investigated.