1 Introduction

Lockdowns applied worldwide to stop the spread of COVID-19 have aroused great interest in the scientific community due to the fact that the effects of air pollution resulting from these lockdowns are greater in some places than in others.

Ground-based and satellite observations over China—where the COVID-19 crisis began in late 2019—and northern Italy, since January 2020, have shown signs of declining concentrations of air pollutants harmful to human health, such as nitrogen dioxide (NO2) which is emitted mainly by vehicles and power plants burning fossil fuels and is closely linked to morbidity and mortality (NASA 2020; Huang et al. 2020; He et al. 2020). A similar downward trend in these countries has been demonstrated by the atmospheric particle concentrations.

It should be noted that according to the World Health Organization, 4.6 millions of people die each year from diseases directly related to air quality (Cohen et al. 2017). More specifically, 193,000 people died in Europe in 2012 from airborne particles (Ortiz et al. 2017). In addition, air pollution in China is responsible for 4000 preventable deaths every day or 1.6 million deaths in 2016 (Wang et al. 2012; Rohde and Muller 2015). Considering the above-mentioned figures for the first 2 months of lockdown in China a reduction in air pollution could indicate a 6 % reduction in mortality (i.e., about 100,000 people) (Dutheil et al. 2020; Varotsos and Krapivin 2020; Varotsos et al. 2021).

The changes observed in China and Italy were so large that they could not be explained by the variability caused by weather alone. In addition, the rapid spread of COVID-19 infection in selected areas of northern Italy is believed to be related to PM10 pollution due to airborne particles that may serve as carriers of pathogens (Setti et al. 2020).

The situation is more ambiguous when looking at the case of the USA, where air pollution has been reduced in the case of New York and other major cities due to COVID-19 lockdowns, but unlike most US cities a statistically significant change in air pollution. According to satellite data, the only city in the USA with a statistically significant improvement in air quality is Los Angeles, but there are skepticisms that the unusually heavy rainfall in recent weeks has contributed significantly (Schiermeier 2020).

Based on the above-mentioned experience, it should be of great interest to examine the impact of the lockdown applied in Greece in terms of air pollution, analyzing both ground-based and satellite data (Cracknell and Varotsos 2007; Cracknell and Varotsos 2011). This is the main purpose of the present paper.

2 Data and Analysis

For the purposes of this study, we used the available data gathered by the ground-based National Network for Monitoring Air Pollution operated by the Ministry of Environment & Energy (MEE) and local authorities (e.g., municipalities /counties). Figure 1a shows the locations of the network stations used in this study, which are installed in the greater Athens area and operate by the MEE. Figure 1b illustrates the locations of used stations installed in other major Greek cities operated by local authorities.

Fig. 1
figure 1

a The ground-based National Air Pollution Monitoring Network located in the greater Athens area. b Ground-based stations for air pollution monitoring located at Thessaloniki, Volos and Larisa (source: Google Earth)

Full size image

In this study, we focus on NO2, O3, PM10, and PM2.5 pollutants. Details of the measurements of each of them, such as the name of the station, the average measurement time, the measurement technique used for each pollutant, the equipment, and the unit of measurement, are given in Table 1.

Table 1 Information on each of the pollutants investigated in this study
Full size table

According to this information, these four pollutants are not available for all stations. It should be noted that all the equipment listed in Table 1 has been officially approved by the European Environment Agency. The data is taken from the database of the European Environment Agency’s (EEA) (https://discomap.eea.europa.eu/map/fme/AirQualityExport.htm, last access 18/6/2020).

Satellite data have been retrieved from the Ozone Monitoring Instrument (OMI) database (last access 20/6/2020). Specifically, the following time series were used in this study : the average range of NO2 tropospheric column (OMNO2d_003_ColumnAmountNO2TropCloudScreened) (last access 20/6/2020), the SO2 planetary boundary layer column amount (OMSO2e_003_ColumnAmountSO2_PBL) (last access 20/6/2020), the a Aerosol Optical Thickness (AOT) at 442 nm (OMAEROe_003_AerosolOpticalThicknessMW_442_0) (last access 20/6/2020) and the Absorbing Aerosol Optical Thickness at 442 nm (OMAEROe_003_AbsorbingAerosolOpticalThicknessM W_442_0) (last access 20/6/2020) over the greater Athens basin and Thessaloniki. The data were obtained through the Giovanni web-based application developed by the Goddard Earth Sciences Data and Information Services Center (GES DISC) (Acker and Leptoukh 2007).

The daily data presented, cover the period from March 1 to June 10 of the years mentioned in each specific case.

3 Results and Discussion

3.1 The Pattern of Air Quality in Greece

In the urban and industrial areas of Greece, air pollution has been a serious environmental problem related to the rapid and anarchic urbanization of cities and the increase of motor vehicles, central heating, and industrial facilities. In the 1960s and 1970s, there were serious problems with air pollution, especially in Athens (Varotsos and Cartalis 1991; Chattopadhyay et al. 2012; Varotsos et al. 2014). In most urban areas (e.g., Athens and Thessaloniki), air quality has improved significantly in recent years, mainly due to the replacement of polluting vehicles of old technology with those equipped with catalytic converters; hence, the environmental levels of smoke, CO, SO2, NOx (NO2, NO), and O3 decreased significantly. Policy reduction practices have effectively reduced primary (SO2, NOx, CO, smoke) and secondary (O3) pollutants. However, the economic crisis of recent years (2010–2018) has forced some residents to use low-cost wood and biomass as fuel for fireplaces and stoves, resulting in fog and high levels of airborne particles that are dangerous due to their toxic organic compounds, free radicals, and metals.

3.2 The Pattern of the Air Quality in Athens

In particular, the high population density, the intense sunshine, and the special topography—a basin surrounded by mountains—create in the city of Athens and in the wider area air pollution problems characterized by the following:

  • Since 1984, there has been a decrease in the concentrations of certain air pollutants and mainly primary pollutants, such as CO and SO2.

  • The issue of air pollution in Athens focuses mainly on tropospheric Ο3 produced by the combination of intense sunshine with ozone precursors.

  • The PM10 concentration exceeds limits for more than 24 h in some stations, mainly due to the dust transported from Sahara.

  • Higher concentrations of SO2, NO2 and CO are observed in the city center. Exceeding the annual NO2 limit was observed in limited locations. SO2 and CO concentrations during the winter are higher than in the summer. Maximum CO values coincide with traffic peaks.

  • Τhe combination of topography, low mixing height, and weak local southwest winds (sea breeze effect) often leads to O3 accumulation in the northern and northeastern suburbs, especially during the summer afternoon. Thus, O3 concentrations are usually higher in suburban locations.

As mentioned in the Introduction, epidemiological studies in European cities have documented correlations between exposure to biomass combustion products, air pollutants, and some adverse health effects.

3.3 The Impact to the Air Pollution of the COVID-19 Lockdown: the Case of Athens

As of March 10, the government suspended all educational institutions nationwide and then, on March 13, proceeded with closing all cafes, bars, museums, shopping malls, sports facilities, and restaurants in the country. On March 16, all retail stores and places of worship were also closed. Additionally, from 6 p.m. on March 23, all unnecessary travels across the country were restricted. On April 4, these restrictions were extended until April 27, and on April 23, they were extended once again until May 4. A final extension by May 10 was decided by the government authorities. May 11 is considered the first day of the partial withdrawal of the restrictive measures.

Under these conditions, levels of air pollution are expected to have dropped significantly. In the following, we use the available air pollution measurements to verify this expectation.

3.3.1 Nitrogen Dioxide

Only about 1% of the total amount of NO2 found in cities’ air is formed naturally, by lightning, plants, soil, and water. It is a key air pollutant contributing to the formation of photochemical smog. The main source of this is the burning of fossil fuels, and when it comes to cities, it comes mainly from the motor vehicle emissions. Its other sources are petrol and metal refining, electricity generation from coal-fired power stations, other manufacturing industries, and food processing.

Ground-Based Data

Figure 2 shows the NO2 levels in the greater area of the Athens basin as derived from the measurements of the ground-based air pollution monitoring network stations located in the region. The time period of the NO2 concentrations presented is 1 March, a few days before lockdown until 10 June 2020. For comparison reasons, these NO2 concentrations levels are plotted together with those reflecting the normal NO2 concentrations of the entire available record in EEA database, i.e., 2013–2019. The inspection of Fig. 2 reveals that the reduction of NO2 levels during the lockdown period does not indicate a clear step-by-step or rapid reduction that could be expected after the abrupt application of shutdown. It is noteworthy that the NO2 concentrations during lockdown are in many cases into the standard deviation zone of its normal long-term levels or close to their lowest limits. Except for a short period, about 10 days after the start of the withdrawal measures, i.e., the “opening”, when high NO2 concentration values were observed in almost all stations, the mean daily concentration values followed the same pattern as the “lockdown” period. In most of the days studied after opening, the observed NO2 concentrations fall within the range 1σ of the 2013–2019 mean value.

Fig. 2
figure 2

The average daily concentration of NO2 in different stations (Aristotelous (a), Patision (b), Peristeri (c), Pireaus-1 (d), Nea Smirni (e), Marousi (f), Lykovrisi (g), and Ag. Paraskevi (h)) of the ground-based network located in the greater Athens area before, during and after the national lockdown due to the Coronavirus disease (COVID-19) outbreak (red dots), along with the long-term mean observed in 2013–2019 (blue dots) and error bars 1σ. The lockdown was introduced on March 12 (red vertical line) and the first part of the opening was on May 11 (green vertical line)

Full size image

Satellite Data

The above-mentioned result obtained from the ground-based NO2 concentrations is confirmed by the NO2 satellite observations illustrated in Fig. 3. More specifically, no clear lockdown signal is detected at the NO2 concentrations in the wider Athens area by satellite observations. Most of the NO2 observations lie within the range (μ−1σ, μ+1σ) leading to the conclusion that the claim that the lockdown period favored the environment due to the reduction of air pollution is not sufficiently confirmed in the case of Athens.

Fig. 3
figure 3

Time series of daily mean area averaged NO2 tropospheric column density over greater Athens basin before, during and after national lockdown due to coronavirus disease (COVID-19) outbreak (red dots), along with the long term mean observed during 2005–2019 (blue dots) and the 1σ error bars. The lockdown was introduced on March 12 (red vertical line) and the first part of the opening was on May 11 (green vertical line)

Full size image

3.3.2 Ozone from Ground-Based Data

Surface O3 is a secondary air pollutant formed by a series of chemical reactions in the presence of sunlight. The photochemical reactions of nitrogen oxides and volatile organic compounds (which include various hydrocarbons) govern the concentration of ground-level O3 in the atmosphere. The time required to carry out these reactions varies from hours to days. Episodes of air pollution or smog occur when high concentrations of precursors (e.g,. O3>40 ppb) coincide with weather conditions favorable for O3 production such as when the air is warm and slow moving.

Figure 4 shows the O3 surface levels in the greater Athens basin before, during, and after the lockdown period together with the long-term normal ozone concentration coming from the entire data archive. The conclusion drawn from Fig. 4 is that there is no clear indication that surface O3 levels have declined during the lockdown period. Particularly, most of the observations are within the range of the one standard deviation limits or higher. It should be recalled here that O3 levels at a particular location may have originated from the emissions of nitrogen oxides and volatile organic compounds that have taken place hundreds or even thousands of kilometers away. Most likely, this transport mechanism is dominated by reducing the expected sharp decrease in O3 levels during the lockdown period.

Fig. 4
figure 4

The average daily concentration of O3 at different stations (Patision (a), Pireaus-1 (b), Marousi (c), and Ag. Paraskevi (d)) of the ground-based network located in the greater Athens area before, during, and after the national lockdown due to the coronavirus disease (COVID-19) outbreak (red dots), along with the long-term mean observed in 2013–2019 (blue dots) and the 1σ error bars. Red and green lines as in Fig. 2

Full size image

3.3.3 PM10 from Ground-Based Data

Briefly describing the PM10, we may say that it is a particulate matter (mixture of solid particles and liquid droplets) 10 μm or less in diameter, causing a decrease in visibility. It should be clarified that PM10 is a particle size classification, and its chemical properties vary depending on the particles’ sources. PM10 can be either anthropogenic or natural. Among their common sources are dust from unsealed roads, smoke from fires, sea salt, car and truck exhausts, and industry.

Figure 5 shows the temporal evolution of PM10 before, during, and after the lockdown period at several monitoring stations in the wider Athens basin along with the long-term normal PM10 for comparison purposes. Inspection of Fig. 5 shows that no significant change in the PM10 concentration was observed during the lockdown period.

Fig. 5
figure 5

The average daily concentration of PM10 at different stations (Aristotelous (a), Persiteri (b), Pireaus-1 (c), Nea Smirni (d), Marousi (e), Lykovrisi (f), and Ag. Paraskevi (g)) of the ground-based network located in the greater Athens area before, during, and after the national lockdown due to the coronavirus disease (COVID-19) outbreak (red dots), along with the long-term mean observed in 2013–2019 (blue dots) and the 1σ error bars. Red and green lines as in Fig. 2

Full size image

3.3.4 PM2.5 from Ground-Based Data

Particulate matter PM2.5 or fine particulate matter means the mass of particles per m3 of air with a size (diameter) generally less than 2.5 μm. The most important sources of PM2.5 are human-made compared to natural ones, which have a small contribution to the overall concentration. For example, in cities, an important source of PM2.5 is road vehicles. In some places, industrial emissions can also be significant, such as the use of non-smokeless fuels for heating and other domestic sources of smoke such as bonfires. Long-range transport, together with pollution from local sources, can lead to short-term high pollution episodes. In addition to these direct (i.e., primary) emissions of particles, PM2.5 can also be formed by chemical reactions of gases such as SO2 and nitrogen oxides.

The measurements of the PM2.5 in some representative monitoring stations in the Athens basin during the lockdown period do not show a significant decline (Fig. 6). It is probably an indication that the long-range transport mechanism has prevailed.

Fig. 6
figure 6

The average daily concentration of PM2.5 at different stations (Aristotelous (a), Pireaus-1 (b), Lykovrisi (c), and Ag. Paraskevi (d)) of the ground-based network located in the greater Athens area before, during, and after the national lockdown due to the coronavirus disease (COVID-19) outbreak (red dots), along with the long-term mean observed in 2013–2019 (blue dots) and the 1σ error bars. Red and green lines as in Fig. 2

Full size image

3.3.5 SO2, AOT, AAOT, from Satellite Data

SO2 is formed during the combustion of fuel containing S, such as coal and oil. It is also emitted by trains, large ships, and some diesel equipment that burns high S fuel and volcanic eruptions. It is a common air pollutant that is found especially in summer smog. SO2 and nitrogen oxides affect the environment when they react with substances in the atmosphere to form acid rain.

Satellite observations monitor aerosol optical thickness (AOT) and absorption AOT (AAOT) in two and three dimensions to understand the evolution of natural and anthropogenic aerosol plumes. These observations are based on the fact that the atmospheric particles alter the way the atmosphere reflects/absorbs visible and infrared light. Figure 7 shows the satellite observations conducted by OMI instrument of SO2 AOT, and AAOT over the Athens basin during the lockdown period, where no significant change was observed compared to their respective long-term variability.

Fig. 7
figure 7

Time series of the daily mean area averaged SO2 planetary boundary layer density (a), AOT (b), and AAOT (c) in the greater Athens basin before, during, and after national lockdown due to coronavirus disease (COVID-19) outbreak. The symbols and lines are as in Fig. 3

Full size image

3.4 The Impact on Air Pollution of the COVID-19 Lockdown: the Case of Thessaloniki

3.4.1 Ground-Based Data

Figure 8 shows the level of NO2, O3, PM10, and PM2.5 concentrations during the lockdown period in the city of Thessaloniki, as derived from the observations recorded at Agia Sofia station which belongs to the National Network of Air Pollution Monitoring Stations. For comparative purposes, Fig. 8 also shows the long-term variability of these pollutants. Remarkably, there is no particular difference in air quality despite the restrictive lockdown measures.

Fig. 8
figure 8

The daily mean concentrations of NO2 (a), O3 (b), PM10 (c), and PM2.5 (d) in Ag. Sofia station located in Thessaloniki before, during, and after national lockdown due to coronavirus disease (COVID-19) outbreak. The symbols and lines are as in Fig. 2

Full size image

3.4.2 Satellite Data

Given that satellite data has greater spatial coverage, it would be interesting to analyze satellite observations for the geographical coordinates of Thessaloniki city. The results obtained are shown in Fig. 9, where the NO2, SO2, AOT, and AAOT patterns are displayed for the lockdown period compared to their long-term mean pattern. From Fig. 9 is obvious that the data values during the lockdown period belong into the variability interval of their long-term mean, thus declaring that there was no decrease during the lockdown period.

Fig. 9
figure 9

Time series of the daily mean area averaged NO2 tropospheric column density (a), SO2 planetary boundary layer density (b), AOT (c), and AAOT (d) in the greater Thessaloniki area before, during, and after national lockdown due to coronavirus disease (COVID-19) outbreak. The symbols and lines are as in Fig. 3

Full size image

3.5 The Impact of the COVID-19 Lockdown on Air Pollution: the Case of Volos and Larisa

Figure 10 presents PM10 and PM2.5 data recorded by the ground-based instrumentation of the Volos and Larisa stations that belong to the National Network of Air Pollution Monitoring Stations and are operated by the local authorities. An inspection of this figure reveals that there is no statistical difference between the values collected during the lockdown period and their long-term mean. It is clear evidence that the combination of meteorological regime prevailed, and the long-range transport of particulate matter during the lockdown period resulted in the coverage of any reduction signal due to limited anthropogenic activity.

Fig. 10
figure 10

PM10 (a) and PM2.5 (b) daily mean concentrations in Volos and PM10 daily mean concentration in Larisa (c) before, during, and after national lockdown due to the occurrence of coronavirus disease (COVID-19). The symbols and lines are as in Fig. 2

Full size image

4 Conclusions

The purpose of this study was to investigate how much cleaner the air was above Greece due to the coronavirus shutdown, using observations made from both ground-based and satellite-borne instruments. The results obtained can be summarized as follows:

In most cases, there has been no statistically significant reduction in air pollution in Greece, if the air pollution levels are compared to the long-term mean pollution records of the last 7 years (for most of the cases presented). This may be attributed to the meteorological conditions that prevailed during the coronavirus shutdown period which contributed significantly. For example, meteorological conditions are governed by non-linear dynamics, obeying scaling phenomena affecting air pollution in the same way (Efstathiou and Varotsos 2010, 2012; Efstathiou et al. 2011; Varotsos et al. 2013)

The lockdown introduced, objectively slowed the spread of Covid-19, regardless of whether it also caused significantly reduced air pollution. In countries with significantly reduced air pollution, the question arisen is how many lives have been saved as a result of reduced pollution? In the case of China, about 12,125 deaths were prevented during China’s countrywide ban on traffic mobility between February 10 and March 14. According to a recently published study, this is higher than the lives lost to the pandemic - 4,633 as of May 4 (Chen et al., 2020).

In any case, because exposure to air pollution is known to damage the heart and lungs, it increases the vulnerability to more serious coronavirus effects. For example, a small increase in long-term exposure to fine particulate matter (PM2.5) leads to a large increase in the coronavirus death rate (e.g., just 1 μg/m3 to particulates corresponded to a 15% increase in COVID-19 deaths).