Effect of Lockdown Measures on Atmospheric Nitrogen Dioxide during SARS-CoV-2 in Spain

The disease caused by SARS-CoV-2 has affected many countries and regions. In order to contain the spread of infection, many countries have adopted lockdown measures. As a result, SARS-CoV-2 has negatively influenced economies on a global scale and has caused a significant impact on the environment. In this study, changes in the concentration of the pollutant Nitrogen Dioxide (NO2) within the lockdown period were examined as well as how these changes relate to the Spanish population. NO2 is one of the reactive nitrogen oxides gases resulting from both anthropogenic and natural processes. One major source in urban areas is the combustion of fossil fuels from vehicles and industrial plants, both of which significantly contribute to air pollution. The long-term exposure to NO2 can also cause severe health problems. Remote sensing is a useful tool to analyze spatial variability of air quality. For this purpose, Sentinel-5P images registered from January to April of 2019 and 2020 were used to analyze spatial distribution of NO2 and its evolution under the lockdown measures in Spain. The results indicate a significant correlation between the population’s activity level and the reduction of NO2 values.


Introduction
Clean air is an essential requirement for human health, and as such air pollution is a major threat to human well-being. Air pollution is the largest environmental health risk in many regions around the world. The World Health Organization estimates that air pollution kills 7 million people worldwide every year, making it necessary to monitor air pollution and improve air quality [1]. The environmental impact is more evident in areas where population density is high, being particularly severe in megacities where high population density, extensive motor vehicle use and strong industrial expansion are combined [2]. Poor air quality is not exclusive to megacities; even small cities with populations around 150,000 can have this problem [3]. Therefore, the economic development of cities with expanding industrial areas is associated with increasing population size and environmental degradation of the surrounding areas [4]. In addition, the high levels of motor vehicle activity [5] and their inappropriate use [6] cause an increase of air pollutants in city centers [7]. As a result, quality of life and human health is worsening specifically in cardiovascular, neurological, and respiratory diseases [8][9][10] and even results in higher mortality rates [11,12]. As a consequence of this, developed and developing countries are more and more attentive to urban air quality, developing guidelines, directives and standards to inform and support policymakers [13,14] to reduce the health impacts of air pollution.
Risk Report showed how climate change and environmental risks were among the top positions [48]. However, both problems are associated, the origin of new pathogens, such as SARS-CoV-2, may be explained by environmental degradation. The coronaviruses have been known about since 1930 [49]. They are transmitted from animals [50] and have been increasing in number over the last decades [51]. The degradation of natural spaces from human activity is increasing the rate of contact between wild spaces and humans, resulting in new diseases and facilitating their expansion [52][53][54]. Nevertheless, while many believe that the climate, and therefore the environment is changing, some think this is not attributable to human activity [55], which may be due to their perception of cultural values [56] or ideologically and politically motivated actors [57].
Unfortunately, the spread of the coronavirus SARS-Cov-2 has been unstoppable and has become a pandemic [58], with dramatic results in countries like Spain, Italy and the United Kingdom [59] in Europe's case. Interventions like quarantine or isolation have shown to be effective in reducing the number of SARS-CoV-2 infections [60]. In addition, some countries and regions have deemed it necessary to impose lockdown measures on economic activities and with unprecedented travel restrictions [61,62]. Under this lockdown period, changes in air pollution can provide valuable information on air quality improvement when there are restrictions on emissions. In this manuscript, using Sentinel S5P images, we analyze the spatial and temporal variation of NO 2 concentrations during Spain's SARS-Cov-2 lockdown phase which took place in March and April 2020, and how these variations are related to city-scale demographics.

Study Area
Spain has just over 47 million inhabitants as of 1 January 2020. Although unevenly distributed throughout the territory, the average population density is equal to 82 habitants per km 2 . Figure 1 shows population density using the European Environment Agency 10 × 10 km grid as a reference. There are areas with highly concentrated populations next to areas of demographic voids. The factors that explain this unbalanced population distribution in Spain are both natural and historical. Regarding natural factors areas with flat and low-lying relief, with a temperate and humid climate and access to the sea and rivers are the most populated. As far as historical factors are concerned, the distribution of the population is related to the economic structure of the country and the development of transport and communication infrastructure. As a result, the average population density in Spain is 416 inhabitants per square kilometer, with very high density areas, over 750 habitants per square kilometer, compared to others with very low densities.
Remote Sens. 2020, 12, x FOR PEER REVIEW 3 of 15 They are transmitted from animals [50] and have been increasing in number over the last decades [51]. The degradation of natural spaces from human activity is increasing the rate of contact between wild spaces and humans, resulting in new diseases and facilitating their expansion [52][53][54]. Nevertheless, while many believe that the climate, and therefore the environment is changing, some think this is not attributable to human activity [55], which may be due to their perception of cultural values [56] or ideologically and politically motivated actors [57]. Unfortunately, the spread of the coronavirus SARS-Cov-2 has been unstoppable and has become a pandemic [58], with dramatic results in countries like Spain, Italy and the United Kingdom [59] in Europe's case. Interventions like quarantine or isolation have shown to be effective in reducing the number of SARS-CoV-2 infections [60]. In addition, some countries and regions have deemed it necessary to impose lockdown measures on economic activities and with unprecedented travel restrictions [61,62]. Under this lockdown period, changes in air pollution can provide valuable information on air quality improvement when there are restrictions on emissions. In this manuscript, using Sentinel S5P images, we analyze the spatial and temporal variation of NO2 concentrations during Spain's SARS-Cov-2 lockdown phase which took place in March and April 2020, and how these variations are related to city-scale demographics.

Study Area
Spain has just over 47 million inhabitants as of 1 January 2020. Although unevenly distributed throughout the territory, the average population density is equal to 82 habitants per km 2 . Figure 1 shows population density using the European Environment Agency 10 × 10 km grid as a reference. There are areas with highly concentrated populations next to areas of demographic voids. The factors that explain this unbalanced population distribution in Spain are both natural and historical. Regarding natural factors areas with flat and low-lying relief, with a temperate and humid climate and access to the sea and rivers are the most populated. As far as historical factors are concerned, the distribution of the population is related to the economic structure of the country and the development of transport and communication infrastructure. As a result, the average population density in Spain is 416 inhabitants per square kilometer, with very high density areas, over 750 habitants per square kilometer, compared to others with very low densities.   Figure 2 shows the location of cities with more than 275,000 inhabitants, with the highest concentrations in Madrid, having more than 3 million inhabitants, and Barcelona, with just over 1.6 million. The relationship between the number of inhabitants in these eleven cities and the variation of NO 2 VCD under the lockdown measures is the focus of this analysis.
Remote Sens. 2020, 12, x FOR PEER REVIEW 4 of 15 Figure 2 shows the location of cities with more than 275,000 inhabitants, with the highest concentrations in Madrid, having more than 3 million inhabitants, and Barcelona, with just over 1.6 million. The relationship between the number of inhabitants in these eleven cities and the variation of NO2 VCD under the lockdown measures is the focus of this analysis.

Remote Sensing Image Collections
The TROPOMI on-board the Sentinel-5 Precursor (Sentinel-5P) was used to collect data on NO2 concentration. The Sentinel-5P mission, launched by the European Space Agency in 2017, is a loworbit polar satellite used to monitor Earth's atmosphere with a high spatio-temporal resolution using the TROPOMI. Concretely, it is a multispectral sensor that registers reflectance values at ultravioletvisible (250-500 nm), near-infrared (675-775 nm) and short-wave infrared (2305-2385 nm) wavelengths which measures concentrations of ozone, methane, formaldehyde, aerosols, carbon monoxide, nitrogen oxide and sulphur dioxide as well as cloud characteristics like cloud fraction, cloud base and pressure.
Image processing was performed with the Google Earth Engine, a cloud-based platform for geospatial analysis with high computational capabilities [63]. A total of 1637 Sentinel 5P Nitrogen Dioxide level-3 scenes of Spain from January to April of 2019 were used and 1636 scenes from the same period were used from 2020 (Table 1). Thus, Sentinel 5P Level-2 data [64] are processed to obtain a single grid per orbit, which allows the Google Earth Engine to process the data. In addition, the data are previously filtered, resulting in pixels with quality assurance values less than 75% being removed, such as cloud or partially snow-covered pixels, errors and or problematic retrievals. First, for each month and year a median image was generated to represent the time series of the images, obtaining an individual image for each time period. Thus, each pixel in the output image was equal to the median value of all the images at that location. On the NO2 median images, once masked for the geographical space of Spain, the statistics corresponding to maximum, minimum, median, and 2nd and 3rd quartile were determined. This allowed a space-time comparison of the NO2 VCD throughout the year and between years.
In addition, the variation of NO2 VCD before and after the lockdown measures was determined. Thus, taking 15 March 2020 as a time reference, the day on which the lockdown measures became effective in Spain, a median image was calculated to represent the NO2 VCD one month before and after the adoption of these measures, and the variation of this component was then determined. For the most populated cities in Spain, the average value of variation of NO2 VCD was determined in order to analyze its relationship with the number of inhabitants per city.

Remote Sensing Image Collections
The TROPOMI on-board the Sentinel-5 Precursor (Sentinel-5P) was used to collect data on NO 2 concentration. The Sentinel-5P mission, launched by the European Space Agency in 2017, is a low-orbit polar satellite used to monitor Earth's atmosphere with a high spatio-temporal resolution using the TROPOMI. Concretely, it is a multispectral sensor that registers reflectance values at ultraviolet-visible (250-500 nm), near-infrared (675-775 nm) and short-wave infrared (2305-2385 nm) wavelengths which measures concentrations of ozone, methane, formaldehyde, aerosols, carbon monoxide, nitrogen oxide and sulphur dioxide as well as cloud characteristics like cloud fraction, cloud base and pressure.
Image processing was performed with the Google Earth Engine, a cloud-based platform for geospatial analysis with high computational capabilities [63]. A total of 1637 Sentinel 5P Nitrogen Dioxide level-3 scenes of Spain from January to April of 2019 were used and 1636 scenes from the same period were used from 2020 (Table 1). Thus, Sentinel 5P Level-2 data [64] are processed to obtain a single grid per orbit, which allows the Google Earth Engine to process the data. In addition, the data are previously filtered, resulting in pixels with quality assurance values less than 75% being removed, such as cloud or partially snow-covered pixels, errors and or problematic retrievals. First, for each month and year a median image was generated to represent the time series of the images, obtaining an individual image for each time period. Thus, each pixel in the output image was equal to the median value of all the images at that location. On the NO 2 median images, once masked for the geographical space of Spain, the statistics corresponding to maximum, minimum, median, and 2nd and 3rd quartile were determined. This allowed a space-time comparison of the NO 2 VCD throughout the year and between years.
In addition, the variation of NO 2 VCD before and after the lockdown measures was determined. Thus, taking 15 March 2020 as a time reference, the day on which the lockdown measures became effective in Spain, a median image was calculated to represent the NO 2 VCD one month before and after the adoption of these measures, and the variation of this component was then determined. For the most populated cities in Spain, the average value of variation of NO 2 VCD was determined in order to analyze its relationship with the number of inhabitants per city.

Results
European countries have established differing lockdown measures in order to control the spread of the SARS-CoV-2 outbreak. These measures, imposed at varying degrees, were implemented at different moments during the second half of March 2020 and have restricted freedom of movement and outlawed public meetings. Italy and Spain were the first countries in Europe to implement these measures on 11 and 14 March 2020, respectively. Taking 15 March 2020 as a time reference, Figure

Results
European countries have established differing lockdown measures in order to control the spread of the SARS-CoV-2 outbreak. These measures, imposed at varying degrees, were implemented at different moments during the second half of March 2020 and have restricted freedom of movement and outlawed public meetings. Italy and Spain were the first countries in Europe to implement these measures on 11 and 14 March 2020, respectively. Taking 15 March 2020 as a time reference, Figure Table 2 the statistic description of NO 2 VCD in Spain for 2019 and 2020 from January to April. A wide range of NO 2 VCD values were observed (Figure 4a), especially in January and February of both years. The maximum values were reduced in March and April, especially in 2020, and declined sharply in April 2020, coinciding with mobility restrictions. On the other hand, the minimum values did not fluctuate, maintaining similarity throughout the months of both years. Median, quartile 25% and 75% values were closer to the minimum values, showing the presence of geographical areas with higher values of NO 2 VCD than the rest. Figure 4.a shows how the range of NO 2 was reduced in March and April in the two years. However, while in 2019 the distribution was very similar, this was not the case in 2020. In March 2020, the reduction in the range of NO 2 VCD was more pronounced than in the same month of the previous year. In April 2020, the reduction was much more pronounced, coinciding with mobility restrictions. Figure 4b shows in detail the evolution of the distribution of NO 2 VCD around the median values. During 2019, the median values were similar, ranging from 2.08 × 10 −5 to 2.19 × 10 −5 mol m −2 in January and February, respectively. Contrariwise, in 2020, the median values of NO 2 did not show the same stable behavior of the previous year. January 2020 showed the highest value, 2.75 × 10 −5 mol m −2 , reducing slightly in February, although it was still higher than the previous year's values. However, in the month of March 2020 there was a very pronounced reduction of NO 2 in April 2020, with a median value equal to 1.65 × 10 −5 mol m −2 , the lowest value of all the months analyzed. In addition, the interquartile range in the months of March and April 2020 was smaller and therefore the distribution of NO 2 was more homogeneous throughout the territory. Figure 4 shows the box and whisker plot and Table 2 the statistic description of NO2 VCD in Spain for 2019 and 2020 from January to April. A wide range of NO2 VCD values were observed (Figure 4a), especially in January and February of both years. The maximum values were reduced in March and April, especially in 2020, and declined sharply in April 2020, coinciding with mobility restrictions. On the other hand, the minimum values did not fluctuate, maintaining similarity throughout the months of both years. Median, quartile 25% and 75% values were closer to the minimum values, showing the presence of geographical areas with higher values of NO2 VCD than the rest. Figure 4.a shows how the range of NO2 was reduced in March and April in the two years. However, while in 2019 the distribution was very similar, this was not the case in 2020. In March 2020, the reduction in the range of NO2 VCD was more pronounced than in the same month of the previous year. In April 2020, the reduction was much more pronounced, coinciding with mobility restrictions. Figure 4b shows in detail the evolution of the distribution of NO2 VCD around the median values. During 2019, the median values were similar, ranging from 2.08 × 10 −5 to 2.19 × 10 −5 mol m −2 in January and February, respectively. Contrariwise, in 2020, the median values of NO2 did not show the same stable behavior of the previous year. January 2020 showed the highest value, 2.75 × 10 −5 mol m −2 , reducing slightly in February, although it was still higher than the previous year's values. However, in the month of March 2020 there was a very pronounced reduction of NO2 in April 2020, with a median value equal to 1.65 × 10 −5 mol m −2 , the lowest value of all the months analyzed. In addition, the interquartile range in the months of March and April 2020 was smaller and therefore the distribution of NO2 was more homogeneous throughout the territory.     Figure 5I.c) and April ( Figure 5I.d) in 2019, although these areas presented higher values than the rest of Spain. In 2020 ( Figure 5II), the distribution of NO 2 VCD was similar to that of 2019 in the months of January (Figure 5II.a) and February ( Figure 5II.b), with the same urban areas standing out as in 2019 due to their increase in NO 2 . In March 2020, coinciding with the middle of the month in which mobility restriction measures were adopted, a reduction in NO 2 VCD was observed throughout the country, with only the geographical area around Madrid showing higher values than the rest of Spain, although lower than in previous months. Finally, in April 2020 ( Figure 5II.d), where the mobility restrictions measures were applied throughout the month, there was a very marked reduction in NO 2 VCD throughout the totally of Spain, with strongly homogeneous behavior and hardly any variation.    Figure 5I.c) and April ( Figure 5I.d) in 2019, although these areas presented higher values than the rest of Spain. In 2020 ( Figure 5II), the distribution of NO2 VCD was similar to that of 2019 in the months of January (Figure 5II.a) and February ( Figure 5II.b), with the same urban areas standing out as in 2019 due to their increase in NO2. In March 2020, coinciding with the middle of the month in which mobility restriction measures were adopted, a reduction in NO2 VCD was observed throughout the country, with only the geographical area around Madrid showing higher values than the rest of Spain, although lower than in previous months. Finally, in April 2020 ( Figure 5II.d), where the mobility restrictions measures were applied throughout the month, there was a very marked reduction in NO2 VCD throughout the totally of Spain, with strongly homogeneous behavior and hardly any variation.    (Figure 6c), between January and April 2020. Of the three cities, Madrid presented the highest NO 2 VCD values at the beginning of 2020. In the month of January, both Madrid (Figure 6a.I) and Barcelona (Figure 6b), due to the size of their metropolitan areas and number of inhabitants, showed higher values in the center of these areas, reducing radially as the distance increases from the central area. In February, there was a reduction in NO 2 in the city of Madrid (Figure 6a.II), but not in the cities of Barcelona (Figure 6b.II) and Valencia (Figure 6c.II). These geographical areas presented higher values than the surrounding areas. In March, coinciding with the limitation of mobility and activity in the middle of the month, a reduction in NO 2 was observed in the three urban areas. In the case of Madrid (Figure 6a.III), the highest NO 2 values appeared around the city and not in the metropolitan area. A similar NO 2 spatial distribution occurred in Barcelona (Figure 6b.III). On the other hand, the city of Valencia and its metropolitan area (Figure 6c.III) present very similar NO 2 values. The effect of mobility restrictions is very evident in the month of April 2020. All analyzed urban areas showed a drastic decrease in NO 2 VCD, only Madrid (Figure 6a.IV) and, to a lesser extent, Barcelona (Figure 6b.IV) showed a very slight increase in values with respect to the surrounding areas for the same period, making it practically impossible to identify a pattern associated with the urban area. In the case of Valencia (Figure 6c.IV) this difference vanished completely. Thus, after 30 days of limitations and restrictions in mobility and activity, the values of NO 2 CDV in these urban areas were similar to those of non-urban areas.
Remote Sens. 2020, 12, x FOR PEER REVIEW 8 of 15 Figure 6 presents the evolution in detail of NO2 VCD over the three most populated cities in Spain, Madrid (Figure 6a), Barcelona ( Figure 6b) and Valencia (Figure 6c), between January and April 2020. Of the three cities, Madrid presented the highest NO2 VCD values at the beginning of 2020. In the month of January, both Madrid (Figure 6a.I) and Barcelona (Figure 6b), due to the size of their metropolitan areas and number of inhabitants, showed higher values in the center of these areas, reducing radially as the distance increases from the central area. In February, there was a reduction in NO2 in the city of Madrid (Figure 6a.II), but not in the cities of Barcelona (Figure 6b.II) and Valencia (Figure 6c.II). These geographical areas presented higher values than the surrounding areas. In March, coinciding with the limitation of mobility and activity in the middle of the month, a reduction in NO2 was observed in the three urban areas. In the case of Madrid (Figure 6a.III), the highest NO2 values appeared around the city and not in the metropolitan area. A similar NO2 spatial distribution occurred in Barcelona (Figure 6b.III). On the other hand, the city of Valencia and its metropolitan area (Figure 6c.III) present very similar NO2 values. The effect of mobility restrictions is very evident in the month of April 2020. All analyzed urban areas showed a drastic decrease in NO2 VCD, only Madrid (Figure 6a.IV) and, to a lesser extent, Barcelona (Figure 6b.IV) showed a very slight increase in values with respect to the surrounding areas for the same period, making it practically impossible to identify a pattern associated with the urban area. In the case of Valencia (Figure 6c.IV) this difference vanished completely. Thus, after 30 days of limitations and restrictions in mobility and activity, the values of NO2 CDV in these urban areas were similar to those of non-urban areas.  Figure 7 shows the variation of NO2 VCD from one month before to one month after 15 March 2020 in Spain. The highest NO2 VCD reductions are represented in red, while a severe reduction is represented in yellow. On the other hand, those areas with a low reduction are represented in green and those areas with no discernible variation are in cyan. Throughout the territory, a decrease can be observed after the lockdown measures, with some areas showing a more pronounced decrease than others. The variation of NO2 followed the same spatial distribution as the population density distribution presented in Figure 1. The city of Madrid, the most densely populated city in Spain,  Figure 7 shows the variation of NO 2 VCD from one month before to one month after 15 March 2020 in Spain. The highest NO 2 VCD reductions are represented in red, while a severe reduction is represented in yellow. On the other hand, those areas with a low reduction are represented in green and those areas with no discernible variation are in cyan. Throughout the territory, a decrease can be observed after the lockdown measures, with some areas showing a more pronounced decrease than others. The variation of NO 2 followed the same spatial distribution as the population density distribution presented in Figure 1. The city of Madrid, the most densely populated city in Spain, showed a marked reduction in NO 2 concentration, reaching values of −1.56 × 10 −4 mol m −2 . Slightly lower values were found in areas such as Barcelona, Valencia and some coastal urban areas with a lower population density than Madrid. On the other hand, all these areas were surrounded by metropolitan areas where the reduction in NO 2 was much less pronounced but also important, with values around −0.08 × 10 −4 mol m −2 . Similar values were found in cities such as Seville and its metropolitan area, Valladolid and the Ebro River corridor. The rest of the territory, with lower population densities presented a reduction between 0.04 × 10 −4 and 0.01 × 10 −4 mol m −2 , lower than metropolitan areas. Therefore, the most densely populated areas with high NO 2 concentrations showed the greatest reductions compared to those areas with low density populations. As a result, the distribution of NO 2 VCD in Spain was more homogeneous than in previous months ( Figure 5II.d).
Remote Sens. 2020, 12, x FOR PEER REVIEW 9 of 15 showed a marked reduction in NO2 concentration, reaching values of −1.56 × 10 −4 mol m −2 . Slightly lower values were found in areas such as Barcelona, Valencia and some coastal urban areas with a lower population density than Madrid. On the other hand, all these areas were surrounded by metropolitan areas where the reduction in NO2 was much less pronounced but also important, with values around −0.08 × 10 −4 mol m −2 . Similar values were found in cities such as Seville and its metropolitan area, Valladolid and the Ebro River corridor. The rest of the territory, with lower population densities presented a reduction between 0.04 × 10 −4 and 0.01 × 10 −4 mol m −2 , lower than metropolitan areas. Therefore, the most densely populated areas with high NO2 concentrations showed the greatest reductions compared to those areas with low density populations. As a result, the distribution of NO2 VCD in Spain was more homogeneous than in previous months ( Figure 5II.d).  Figure 8 shows the average value of tropospheric VCD of NO2 reduction in relation to the number of inhabitants, resulting in nine categories. In general, as the number of inhabitants increases, NO2 decreases. Cities with less than 50,000 inhabitants had least significant reduction, with an average value of −2.99 × 10 −5 mol m −2 . On the other hand, those cities with more than 600,000 inhabitants were those that presented the greatest reduction, with average values of less than −1.05 × 10 −4 mol m −2 . Considering the first three categories with the lowest number of inhabitants, the factor of increase in the reduction of tropospheric VCD of NO2 was equal to 1.28 per 50,000 inhabitants. On the other hand, it decreased slightly among the categories of 150,000 to 600,000 inhabitants, being equal to 1.03 per 50,000 inhabitants.  Figure 8 shows the average value of tropospheric VCD of NO 2 reduction in relation to the number of inhabitants, resulting in nine categories. In general, as the number of inhabitants increases, NO 2 decreases. Cities with less than 50,000 inhabitants had least significant reduction, with an average value of −2.99 × 10 −5 mol m −2 . On the other hand, those cities with more than 600,000 inhabitants were those that presented the greatest reduction, with average values of less than −1.05 × 10 −4 mol m −2 . Considering the first three categories with the lowest number of inhabitants, the factor of increase in the reduction of tropospheric VCD of NO 2 was equal to 1.28 per 50,000 inhabitants. On the other hand, it decreased slightly among the categories of 150,000 to 600,000 inhabitants, being equal to 1.03 per 50,000 inhabitants.
The 11 cities with more than 275,000 inhabitants in Spain, are plotted in Figure 9 which shows a negative lineal relationship between population size and NO 2 reduction with a correlation coefficient equal to 0.73 (p-value 0.00004). A negative relationship was expected between population activity and NO 2 levels, where greater activity leads to a higher level. This component is one of the most important in urban air pollution, with the burning of fossil fuels such as coal, oil and gas being one of the main sources of NO 2 . It is estimated that about 86% of nitrogen dioxide in European cities are caused by fossil fuels emitted from motor vehicles [65]. That means that as populations increase, NO 2 also increases. With the lockdown measures in effect there has been almost no vehicle traffic and in turn the concentration of NO 2 has been reduced. The 11 cities with more than 275,000 inhabitants in Spain, are plotted in Figure 9 which shows a negative lineal relationship between population size and NO2 reduction with a correlation coefficient equal to 0.73 (p-value 0.00004). A negative relationship was expected between population activity and NO2 levels, where greater activity leads to a higher level. This component is one of the most important in urban air pollution, with the burning of fossil fuels such as coal, oil and gas being one of the main sources of NO2. It is estimated that about 86% of nitrogen dioxide in European cities are caused by fossil fuels emitted from motor vehicles [65]. That means that as populations increase, NO2 also increases. With the lockdown measures in effect there has been almost no vehicle traffic and in turn the concentration of NO2 has been reduced.

Discussion
NO2 concentrations in megacities exceed recommendations from the World Health Organization [13]. To control the spread of coronavirus the large majority of people have been staying at home, maintaining social distancing practices and working remotely [66]. As expected, the direct consequence of industries and transportation systems shutting down was a sudden drop in air pollutant emissions. The lockdowns have provided researchers the opportunity to set up singular experiments based on real data and not simulations to answer the question of what would happen if individual transport based on fuel combustion were removed and only those linked to public service and supply were active. Under this scenario, around the world, many cities have looked into how air quality has improved since the lockdown measures took place. Particularly in Europe, NO2 emissions  The 11 cities with more than 275,000 inhabitants in Spain, are plotted in Figure 9 which shows a negative lineal relationship between population size and NO2 reduction with a correlation coefficient equal to 0.73 (p-value 0.00004). A negative relationship was expected between population activity and NO2 levels, where greater activity leads to a higher level. This component is one of the most important in urban air pollution, with the burning of fossil fuels such as coal, oil and gas being one of the main sources of NO2. It is estimated that about 86% of nitrogen dioxide in European cities are caused by fossil fuels emitted from motor vehicles [65]. That means that as populations increase, NO2 also increases. With the lockdown measures in effect there has been almost no vehicle traffic and in turn the concentration of NO2 has been reduced.

Discussion
NO2 concentrations in megacities exceed recommendations from the World Health Organization [13]. To control the spread of coronavirus the large majority of people have been staying at home, maintaining social distancing practices and working remotely [66]. As expected, the direct consequence of industries and transportation systems shutting down was a sudden drop in air pollutant emissions. The lockdowns have provided researchers the opportunity to set up singular experiments based on real data and not simulations to answer the question of what would happen if individual transport based on fuel combustion were removed and only those linked to public service and supply were active. Under this scenario, around the world, many cities have looked into how air quality has improved since the lockdown measures took place. Particularly in Europe, NO2 emissions

NO 2 concentrations in megacities exceed recommendations from the World Health
Organization [13]. To control the spread of coronavirus the large majority of people have been staying at home, maintaining social distancing practices and working remotely [66]. As expected, the direct consequence of industries and transportation systems shutting down was a sudden drop in air pollutant emissions. The lockdowns have provided researchers the opportunity to set up singular experiments based on real data and not simulations to answer the question of what would happen if individual transport based on fuel combustion were removed and only those linked to public service and supply were active. Under this scenario, around the world, many cities have looked into how air quality has improved since the lockdown measures took place. Particularly in Europe, NO 2 emissions were highly reduced over northern Italy, Spain and the United Kingdom [67]. It is well known there is a positive association between NO 2 concentration and population size [68]. As in other studies in other countries [69], in Spain, NO 2 concentrations are located around urban areas, being higher as population size increases, indicating anthropogenic sources, mainly produced by vehicles. The situation caused by the SARS-CoV-2 pandemic has led, in the case of Spain, to an 80% reduction in traffic and a reduction of fuel sales by 83% [70].
Urban planners, engineers and policymakers should take the studies that have capitalized on the unique conditions presented by the pandemic into account to promote new strategies to reduce air pollution and consolidate existing ones. To this end, it is necessary to consider how the expansion of urban areas and the complex use of land, transport patterns and socio-economic development directly factor into our living conditions. The expansion of urban areas has led to an increase in residential areas on the outskirts of cities, resulting in disproportionate distances between residential housing and places of work, generating an imbalance in transport with a high dependence on private vehicles [71]. To date, there is not a sufficient number of studies taking into account the integration of transport systems and urban planning to reduce air pollution [72,73].
Regardless of the positive impact on the reduction of air pollutants, climate effects are still present today and should not be understood as a substitute for climate change. In this way, we would like to express that in this manuscript we characterized the changes produced on air quality during the lockdown. We have not tried to attribute specifically nor quantify the effects of the lockdown since other factors may have influenced the changes, such as meteorology and regional and long-range transport of pollutants. An in-depth analysis is required to obtain this information accurately.
In addition to the impact of lockdown measures on air quality, future work should aim to study the impact of the absence of tourists on the appearance of beaches and water quality, or the reduction of noise pollution. Moreover, the massive use of personal protective equipment such as masks or gloves has increased worldwide, and probably recycling and waste management policies should be analyzed and redesigned.

Conclusions
Lockdown measures due to SARS-CoV-2 have provoked a singular and unique opportunity to evaluate the contribution and impact of human activity on the environment. In this manuscript, the effect on air pollution due to the pandemic response in Spain has been shown by evaluating and analyzing the concentration of tropospheric NO 2 from 1 January to 30 April in 2019 and 2020. In this study, data from the TROPOMI on-board the Sentinel-5P satellite platform were used to analyze the spatial-temporal variation in Spain and its relationship with population size and lockdown measures.
The satellite scenes showed a high concentration of NO 2 in the city of Madrid, which has the largest number of inhabitants in Spain. Other hotspots with high concentrations of NO 2 also appeared over cities with a large number of inhabitants. Previous lockdown measures, the relationship between the population density map of Spain and the NO 2 distribution followed the same pattern. Furthermore, as a result of the concentration of the population in very specific points within the Spanish territory and the source of NO 2 , mainly related to vehicle traffic, the distribution of this pollutant presented a wide range of values in the air, clearly differentiating between areas with a high population density and those without.
Just two weeks of lockdown measures and the mobility restrictions were reflected in a reduction of NO 2 , mainly in those areas with larger populations. One month later, the NO 2 reduction was more evident, and not only in the populated areas, adopting a homogeneous distribution throughout the territory. The comparison between the NO 2 concentration values before and after the lockdown measures shows a strong relationship with the number of inhabitants.
These results should be taken into account by governments and policymakers to develop effective NO 2 emissions reduction and air pollution prevention policies. These policies should be based on adopting local measures within a global project.

Conflicts of Interest:
The authors declare no conflict of interest.