We analyzed a combination of mobility data (driving, walking, and transit), the emission for several sectors (ground transport, industry, power generation, residential, domestic, and international aviation), and NO2 concentration from the January of 2020 up to the end of August 2021 to calculate daily changes compared to the same period in 2019 in EU nations and the UK except Croatia, Cyprus, and Malta because of no data.
In the case of traffic congestion, our analysis was conducted before the confinements, and the fractional reduction afterwards was calculated for each European nation and UK. In addition, its effects on each mode (driving, walking, and transit) were evaluated, and compared with the same period in 2019.
The TomTom traffic index 25 offers traffic congestion ranges within several cities worldwide for the past ten years. The review rates cities from the maximum to the minimum congested. It is powered by actual traffic information as well as displays all the modifications on the traffic movements. In this regard, a zero-congestion degree indicates that the traffic flow is totally ‘normal’ or fluid; however, it does not indicate zero emissions as well as any automobiles. It is consequently crucial to recognize the lesser threshold of emissions whenever the congestion degree is actually zero. To accomplish this, Liu et al. 28 established a sigmoid function of day-to-day mean TomTom congestion degree (see “method”) with the mean daily vehicle through publicly accessible real-time information through an average of 60 roads in the megacity region for the city of Paris. The function developed for the city of Paris utilized for some other locations involved within the TomTom dataset.
A radar chart (Fig. 1a) presents the average monthly traffic congestion during the COVID-19 throughout EU countries and the UK during 2019, 2020, and 2021. In 2020, there was no reduction in average traffic congestion until late February. The drop was coincident with the spread of the coronavirus as well as the beginning of lockdowns with more significant reductions by -31% (March), -74% (April), -57% (May), -31% (June), -13% (July), -15% (August), -19% (September), -27% (October), -48% (November), and -40% (December) compared to the same period in 2019. In 2021, average traffic congestion in comparison with 2019 was reduced by -33% (January), -26% (February), -30% (March), -27% (April), -11% (May), -5% (June), -25% (July), and -39% (August). This was because of particular limitation guidelines (see Supplementary Tables S1 until S3) announced and by government authorities. Analysis for each EU nation & UK (see Supplementary Figs. S1 until S5) shows that traffic congestion significantly decreased in April 2020 due to the impact of COVID-19 lockdown measures, which was estimated at around -94% and -45% as the maximum and minimum decrease in Italy and Sweden, respectively. The analysis indicated that traffic congestion fluctuated during the eight months of 2021, as provided in Supplementary Figs. S1 until S5.
The distribution of traffic congestion during rush hour of morning and evening for all EU nations and UK throughout 2020 in comparison to 2019 is represented in Figure 1b. Following the lockdowns (see Supplementary Tables S1 until S3); the congestion was substantially reduced since the end of February, by -30% (for mornings) and -33% (for evenings). It reached a minimum level in April of -78% (i.e. for mornings) and -70% (i.e. for evenings). These relative reductions gradually attenuated from the first of May until July as a consequence of lifting lockdown measures from level 3 to 2 (e.g. Austria, Belgium, Malta, Portugal, Spain, UK) and from level 2 to 1 (e.g. Ireland, Romania, Slovakia, Slovenia) and those reached to the -14% (i.e. for the morning) and -11% (i.e. for the evening). Then, the ranges of mobility during morning and evening decreases to -47% and -33%, respectively, until the end of 2020.
Apple mobility changes 26, and the Le Quéré et al. 11 pointed out that more than 50% of the people worldwide decreased travel by more than 50% in April 2020, and the Google mobility 21 data point out that more than 80% of the people throughout 114 nations decreased their travel by more than 50%. Forster et al. 27 deduced that Google mobility information and emission decrease rates depending on confinement level evaluation in Le Quéré et al. agree on country-level ground transportation tendencies to around 20%.
Based on the collected data from Apple mobility trends 26, daily human mobility including three modes, i.e. driving, transit, and walking, (Fig. 2) was analyzed from the January 2020 as beginning of the spread of COVID-19 until the end of August 2021. These comparisons were carried out based on the three-level restriction order (see Supplementary Tables S1 until S3), and it was identified on levels of 0 to 3 as well as assigned the level to which standard day-to-day actions were restricted for part or entire of the people in each country. Level 0 means that no actions were in place, level 1 that policies aimed at minor categories of people suspected of carrying the virus, level 2 that policies aimed need closing (only some categories or levels, for example just public schools or just high school) 28,29, and level 3 signifies countrywide policies 11. The gray shading in Figure 2 demonstrates the third level as the highest restriction order.
In Austria, four times lockdown announcements as the third level were imposed. On March 16th, 2020, the first lockdown caused a dramatic decrease in overall mobility, of -66% and -81% for driving and walking, respectively. In addition, there was a significant decline on November 17th (-45% and -61%), December 26th, 2020 (-53% and -73%), and April 1st, 2021 (-48% and -66%) for driving and walking during the lockdown announcements as the third level, respectively.
In Belgium, the mobility trends for all sectors were approximately close together except before the first lockdown as the third level, whereas the walking mode was larger than others. Although the mobility trends from the first of January until March 17th, 2020 increased, they substantially decreased after March 18th, 2020 to around -78%, -86%, and -80% for driving, transit, and walking, respectively. Similarly, these trends decreased to -42% (driving), -43% (transit), and -55% (walking) during the second lockdown on November 1st, 2020.
In the Czech Republic, restriction as the third level was imposed from March 15th ; until April 19th, 2020, where the overall mobility for driving, walking, and transit dramatically decreased to -76%, -91%, and -89%, respectively.
In Denmark, although there was a restriction at the third level after January 4th ; 2021, mobility trends for all sectors fell because of the second level restriction and reached -56% (driving), -80% (transit), and -63% (walking) on March 22nd ; 2020. Then, after the third level restriction, driving, transit, and walking decreased to -57%, -71%, and -60%, respectively. It clearly shows that transit usage had the highest decrease than other modes because of the government's stay-at-home order.
In Estonia, two lockdowns as the third level were announced. For the first one, after March 27th, 2020, the mobility trends for driving, transit, and walking declined rapidly to -48%, -73%, and -61%, respectively. One year after the first announcement, the second one started, and it decreased to -14% (driving), -34% (transit), and -14% (walking). In this country, mobility increased quickly between June and December 2020, simply because of the beginning to lift lockdown measures from level 3 to 1.
Finland was one of the EU nations without any restriction order as the third level. Although lockdown measures (level 2) began from March 16th, 2020, for three months, the mobility trends were decreased to nearly -42% (driving), -65% (transit), and -50% (walking). Based on survey outcomes conducted by Clausnitzer 30, from June 2020, more than fifty percent of the people living in the Finnish mentioned that they were buying less often from physical shops throughout the pandemic. There were no substantial distinctions among the kinds of shops; Finnish buyers visited throughout coronavirus; however, going to the department stores and shopping malls were relatively less popular compared to before. Similarly, in Sweden, although coronavirus appeared on 24 January, lockdown measures applied only until level 2 and the government permitted Sweden's bars, restaurants and schools for under 16-year-olds to keep open. In this regard, the mobility trends for driving were estimated to be higher than transit and walking.
The detailed dataset and graphs related to the mobility trends for all other EU nations (France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, UK, Norway) can be found in the supplementary section (see Supplementary Figs. S6 until S8).
The percentages of monthly traffic congestion during rush hour in morning (AM) and evening (PM) in 2020 in comparison to 2019 across EU countries and the UK can be seen in supplementary Figures S9 - S13. Regarding this evaluation, the percentages of congestion during PM times were estimated to be larger than AM times during most of the months of 2020.
Historical data for CO2 emission throughout EU nations from 1965 (see Supplementary Fig. S14) shows several fluctuations until 2019. Due to the lack of data in Estonia, Lithuania, and Latvia, the CO2 emission was provided from 1985).
The latest data (CO2 emissions data were collected from https://carbonmonitor.org/) display (Fig. 3) that the curtailment of mobility and economic activity due to COVID-19 throughout 2020 pushed down European CO2 emission by 10.66% in comparison with 2019, while emissions, during the first six months of 2021, increased by 11.14% from 2020 and remained lower by 4.36% relative to 2019, respectively.
In general, the transportation sector represents 31.30% of EU nations CO2 distribution in 2019 31. As lockdowns began after February 2020 in EU nations and the UK, based on the online data extracted from (https://carbonmonitor.org/), monthly ground transport emissions compared to 2019 (Fig. 3a) decreased by 16.43% (March), 31.51% (April), 20.36 (May), 1.04% (June), 0.11% (July), 3.82% (August), 9.79% (November), and 2.51% (December). Generally, ground transportation emissions in 2020 reduced by 5.89% (1.71 Mt CO2), and within the six months of 2021 increased by 11.38% (1.64 Mt CO2) relative to 2020 and remained 1.32% lower than (0.19 Mt CO2) 2019, respectively.
Because of stay-at-home regulation, though the alter within electricity usage by public/commercial buildings and households was measured as a component of the power emissions. Acquiring day-to-day emissions through this field is high uncertain compared to some other areas, given that day-to-day residential natural gas usage information is not accessible for whole nations 12. Accordingly, monthly residential emissions (residential emissions data were collected from https://carbonmonitor.org/) in 2020 compared to 2019 (Fig. 3b) decreased by 12.41% (January), 7.58% (February), 7.74% (April), 13.74% (May), 3.53% (September), 4.85% (November). Generally, residential emissions in 2020 decreased by 2.46% (0.51 Mt CO2), and within the six months of 2021, increased by 11.72% (1.50 Mt CO2) and 6.54% (0.83 Mt CO2) relative to 2020 and 2019, respectively.
Industry emissions from chemicals, steel, and some other manufactured items from fossil energy-burning and cement generation represent on average 29 percent of the worldwide CO2 emissions throughout a normal year, with a greater share of nationwide emissions in developing nations. In the current research, since data collected from Liu et al. 32 and online data from (https://carbonmonitor.org/), mere emissions via direct fuel usage as well as chemical procedure emissions through the industry field were taken into consideration, electricity associated emissions regarding the industry are generally measured with the power creation field 12. Monthly industry emissions in 2020 compared to 2019 (Fig. 3c) decreased by 1.39% (January), 4.97% (February), 12.52% (March), 29.53% (April), 20.79% (May), 11.27% (June), 7.05% (July), 6.03% (August), 5.82% (September), 3.30% (October), and 0.28% (November). Generally, industry emissions in 2020 reduced by 8.56% (1.78 Mt CO2), and in the first six months of 2021 increased by 8.76% (0.88 Mt CO2) and decreased by 5.02% (0.53 Mt CO2) relative to the 2020 and 2019, respectively.
Monthly power sector emissions in 2020 compared to 2019 (Fig. 3d) decreased by 17.10% (January), 27.01% (February), 9.47% (March), 30.52% (April), 24.77 (May), 7.85% (June), 9.83% (July), 13.60% (October), 8.89% (November). Generally, power sector emissions in 2020 decreased by 11.60% (3.56 Mt CO2), and within the eight months of 2021, increased by 15.37% (2.24 Mt CO2) and decreased by 5.13% (0.79 Mt CO2) relative to the 2020 and 2019, respectively.
Aviation sector emission is divided into two parts, including domestic and international. Figure 3e and Figure 3f show that international aviation had a significant decrease in 2020 and 2021 compared to 2019. Generally, domestic aviation emissions in 2020 reduced by 48.05% (0.26 Mt CO2), and in the first six months of 2021 increased by 2.19% (0.003 Mt CO2) and decreased by 51.91% (0.14 Mt CO2) relative to the 2020 and 2019, respectively. The broad ranges of international flight emissions indicate two considerable reductions throughout 2020, one right after April and the other within December and November coincident with lockdown actions and travel bans globally. Generally, international aviation emissions throughout 2020 reduced by 58.82% (3.67 Mt CO2), and in the first six months of 2021 decreased by 26.93% (0.36 Mt CO2) and 66.67% (1.98 Mt CO2) relative to 2020 and 2019, respectively.
Figure 4 shows the emission of NOx associated with five different sectors from 1970 up to 2015. In EU countries and UK33, the most significant contributions to the global increase in emissions come from transportation (increased by 36.60% from 1970 to 1990) and power industry (increased by 49.42% from 1970 to 1980). Then these trends declined to 3.56 Million (55.68%) and 1.35 Million (67.11%) until 2015, respectively. In the case of agriculture and building, NOx emissions were begun at 0.50 Million and 1.12 Million in 1970, and they had mildly fluctuating until 2015 and reached 0.45 Million and 0.70 Million, respectively. Lastly, for other industrial combustion, although the trend started from 3.50 Million NOx emissions, however, it dramatically decreased to 1.25 Million (64.17%) until 2015.
NOx leads to acid deposition as well as eutrophication of water whereas it leads to more acidification for the soils. The following effects of acid deposition are often substantial, such as negative results on aquatic ecosystems inside lakes and rivers and the destruction of crops, forests, and other vegetation. Eutrophication frequently leads to extreme reductions in water quality with the following effects: reduced biodiversity, modifications in species dominance and composition, and toxicity impact. In this regard, NO2 is related to negative impacts on individual health; at excessive concentrations, it can raise susceptibility to respiratory infection, decreased lung performance, and inflammation of the airways. It also plays a role in creating tropospheric ozone and secondary particulate aerosols in the atmosphere, which are essential air pollutants because of their harmful effects on individual health and some other climate impact 34.
Research conducted by Bauwens et al.5 pointed out unprecedented NO2 column reduction throughout western Europe, South Korea, the United States, and China due to public health actions enforced to contain the COVID-19 outbreak from January 2020.
The evolution of NO2 concentration throughout EU nations and the UK, from January 2019 to eight months of 2021, is visualized in Figure 5 and Supplementary Figs. S15 and S16. The comparison of three years in Figure 5 displays reduces in surface concentrations of NO2 during 2020 and 2021 compared to the identical time in 2019. This was because of the coronavirus outbreak, the imposed interruption in vehicle congestion and the decrease in industrial action throughout the lockdown.
In Austria and Belgium, the NO2 concentration in February 2019 reached the highest level, which was around 17.39 (ppb) and 24.46 (ppb), respectively. In Bulgaria, although the mean level of NO2 concentration in 2019 (except September and November) was computed to be more than 2020, however in 2021, it significantly increased with the maximum value of 17.29 ppb (February) and minimum value of 7.87 ppb (May) which was because of the progressive lifting of confinement.
In the Czech Republic, the trend of concentration during 2019 and 2021 was estimated close together where it was around 8.5 (ppb), 10 (ppb), and 6 (ppb) for January, February, and March, respectively. However, it dramatically decreased for April and May 2021 to 4.23 (ppb) and 2.97 (ppb), respectively, which was substantially lower than the same period in 2019.
In Denmark, except September 2019, the percentages of NO2 concentration were significantly higher than in 2020 and 2021, especially for April, May, and June. The average concentration distinction in Finland as another country in the Nordic region between 2019 and the other two years (2020 and 2021) decreased around 41.21% and 33.77%, respectively.
France is actually one of the nations where the mortality case associated with the COVID-19 was abnormal. The percentages of NO2 concentration during 2020 and 2021 significantly decreased compared to 2019, which was around 20.13% and 40.30%, respectively. The detailed dataset and graphs related to the NO2 concentration for all other EU nations can be found in the supplementary section (see Supplementary Figs. S15 and S16).
NO2 is generally co-emitted with CO2 emission, so NO2 information might be utilized to calculate CO2 emissions. Prior research approximated local CO2 emissions depending on NOx concentration, as well as the NOx concentration to CO2 emission ratios through bottom-up emission ranges 35–37 or co-located satellite retrievals of CO2 and NO238.
In this regard, we initially determined the monthly values of CO2 emission for several sectors (ground transport, industry, power generation, residential, domestic, and international aviation) and NO2 concentration ground-based measurements. Figure 6, as a scatter graph, indicated a robust linear relationship between CO2 emission and NO2 concentration in 2019 (Fig. 6a), as well as a solid connection for the mentioned factors from the first of January 2020 until August 2021 (Fig. 6b). These powerful linear connections among NO2 and CO2 are evident with suitable correlation R2=0.77 and R2=0.78 for 2019 and 2020-2021, respectively. All graphs as the linear regression between CO2 emissions to NO2 concentrations in 2020, 2021 (until August), and all data from the first of January 2019 until August 2021 throughout EU nations and the UK can be found in the supplementary section (see Supplementary Fig. S17). Accordingly, a study carried out by Liu et al. 35 suggested a robust linear connection among CO2 and NOx from 2006 until 2016 in the United States (US).
Based on the approximate values from Figure 6 and Figure 7, the CO2 emission and NO2 concentration ratio (See Methods) were decreased after the first of January 2020 until August 2021 relative to the same period of 2019 simply because of human being actions produced by the coronavirus. Error bars (Figure 7) display the average ratio of 39.90, 35.06, and 33.68 with standard deviations of 2.4, 2.7, and 2.0 (±1σ) for ratios of NO2 to CO2.