Modelling the Global Air Quality Conditions in Perspective of COVID-19 Stimulated Lockdown Periods Using Remote Sensing Data

COVID-19 outbreak across the world has invited forced lockdown conditions, which causes a huge economic landslide. But it has brought an opportunity for restoring the environment of its own which may cause ecosystem well-being. Focusing on the second issue, the present work has intended to explore the streams of air quality change based on some quality components and develop a multi-date air quality state (AQS) model for the world in consequence of emergency lockdown. It is very clear from the result that amid lockdown aerosol optical depth (AOD), sulfur dioxide (SO 2 ), ozone, carbon monoxide (CO), particulate matter (PM 2.5 ), and black carbon (BC) concentration level have been signicantly reduced in fully lockdown countries. AQS is considerably improved amid lockdown. Hotspots of COVID-19 were under unhealthy, very unhealthy air quality class in pre lockdown condition, but amid lockdown, these countries have been shifted to good and moderate healthy air quality classes


Introduction
On December 31, 2019, World Health Organization (WHO) rst informed about the infection of pneumonia-like unknown disease from a seafood market in Wuhan city of Hubei province, China (Neradi et al., 2020). A guideline was issued by WHO for the countries to develop their ability to detect the unknown virus-like to another coronavirus (CoV) like Severe Acute Respiratory Syndrome (SARS) and the provisional name was given like SARS-CoV-2 or novel coronavirus (2019-nCoV) (Ahmed et al., 2020;Bherwani et al., 2020) on 10 th of January, 2020. Con rmation of the spreading of novel coronavirus outside China came into the front on 13 th January 2020. The 2019-nCoV was then found to be rapidly spreading disease-carrying by man (Hui et al., 2020), and reportedly 1320 of con rmed, 1965 of suspect cases were found till 25 th January 2020, out of which 237 were poorly sick and 41 died (World Health Organization, 2020a). A few days later (30 th of January, 2020), WHO declared a public health emergency of international concern regarding the outbreak of 2019-nCoV.
This disease was given an o cial name as 2019-coronavirus disease (COVID-19) on 11 th February 2020 (Shereen et al., 2020), without referring to any particular geographical location (World Health Organization, 2020b). Later on, the rapid outbreak of this disease called for the issuance of several health emergency guidelines regarding the mass gathering, carrying of ill travelers, massive preparedness activities, be a ready campaign , and nally on 11 th March 2020, WHO characterized COVID-19 as pandemic (World Health Organization, 2020b). Meanwhile, the virus kept spreading and 3,244,586 people got infected and 229,182 people died to date (30th April 2020) with multiple epicenters around the world, such as Italy, Iran, Spain, France, USA, UK, Turkey, Belgium, and Brazil, etc (Gautam and Trivedi, 2020). This explosion of COVID-19 cases around the world led many countries to adopt dramatic measures, like restricting physical human interaction, encouraging social distancing and self-isolation, strict enforcement of quarantine, withdrawal of public transportation and restriction on private roaming, closer of markets/supermarkets, closers of educational institutes, colleges, and other institutions, embargoing on public gathering even at private places, declaration of voluntary curfew, even partial or total lockdown of the entire city, district, state or country (Bera et al., 2020;Bherwani et al., 2020;Lin et al., 2020). China rst imposed a partial lockdown on January 23, 2020, and with time several other countries from different parts of the world imposed partial or total countrywide lockdown responding to a severe outbreak of the disease (table 1).
All these aforesaid steps have been taken as preventive measures to withhold the spread of the virus and control the death rate (Saadat et al., 2020;Chinazzi et al., 2020). Such an unusual situation of the past few months has exceptionally changed the world with some unexpected consequences (Harapan et al., 2020). Among them, the lockdown effect has almost stopped all the economic activities, absolutely contributed to improving the environmental quality, which may offset the economic losses to some extent (Chakraborty and Maity, 2020;Muhammad et al., 2020). During the lockdown, the production interregnum of industries, restriction on transportation and public traveling system, and shut down of other business have resulted in a sharp and sudden drop in global carbon emission (Wang and Su, 2020;Chakraborty and Maity, 2020;Saadat et al., 2020). It has also been seen that due to production, manufacturing, and transportation break during the lockdown global demand for oil and coal has declined to its lowest level (Wang and Su, 2020;Muhammad et al., 2020). Therefore, not only the carbon emission, but also the release of some other severely pollutant e.g.
Greenhouse Gases (GHG) and substances like Nitrogen Dioxide (NO 2 ), Sulfur Dioxide (SO 2 ), Methane (CH 4 ), and Particulate Matter 2.5 (PM 2.5 ) have dramatically decreased (Bera et al., 2020;NASA, 2020). For example, since people were told to stay inside, a 25% reduction in CO 2 is observed in the emission data of China which means about 1million tons reduction in the substantial carbon emission (Wang and Su, 2020).
The emission of NO 2 from the heavy industrial area of North and Northeast China was signi cantly reduced just after the rst week of lockdown (European Space Agency, 2020). Almost 50% drop in air pollution in New York City (USA) was observed at the same time last year after the measures were implemented to control the spread of the virus (Saadat et al., 2020).
The emission of NO 2 has reduced in different countries of Europe, like the UK, Spain, and Italy as captured by satellite images (Ficetola and Rubolini, 2020). A signi cant reduction in the concentration of PM 2.5 , PM 10 , CO 2 , and NO2 by 43%, 31%, 10%, and 18%, respectively compared to the previous year can be seen in over 22 cities from different parts of India (Sharma et al., 2020).
Several studies have already taken the initiative to show the changes in air quality due to the restriction posed to anthropogenic activities during the COVID-19 lockdown. Gupta et al., (2020) analyzed the impact of air pollution level on the Covid-19 lethality in the nine selected metropolitan cities of Asia paci c region. Muhammad et al. (2020) showed the decrease of mean tropospheric NO 2 density over China and different countries of Europe and the USA after the enforcement of lockdown compiling the environmental data released by National Aeronautics and Space Agency (NASA) and European Space Agency (ESA). Wang and Su (2020) analyzed the air quality index of China using air quality data and extracting the concentration of NO 2 , PM 2.5 , and PM 10 from the earth observation made by NASA and ESA. Tobías et al. (2020) used the recorded atmospheric pollutant data at the tra c air quality monitoring station of urban background to describe changes in pollution levels in Barcelona. The concentration of PM 2.5 , PM 10 , CO 2 , NO 2 , ozone (O 3 ), and SO 2 of Page 4/22 22 Indian cities on mid-March and April 2020 were compared with the same time of 2017 by Sharma et al. (2020) to show the improvement of air quality in different parts of India. After going through all such kinds of literature, some research gaps have been identi ed. Most of the studies show the changes in air quality and pollution levels by considering one or two indicators like NO 2 or carbon monoxide. Some other studies considered multiple indicators of air quality and pollution, but these are mainly focused on some isolated regions or within the country. Here a need for a comprehensive multi-parametric global scale study arises to evaluate the overall air quality status of the entire world. Though the existing multi-parametric study intensively investigated the component-speci c variation of air quality in pre and amid lockdown periods and analyzed the trend, there was a lack of air quality assessment by integrating the criteria of pollution and air quality degradation. In the present study, the lacks are focused with the critical air quality data regarding AOD, PM 2.5 , SO 2 , O 3 , BC, and CO derived from Modern Era Retrospective-Analysis for Research and Applications, version 2 (MERRA-2) (Song et al., 2018) and Atmospheric Infrared Sounder (AIRS)  to explore how far this incident can be able to improve the air quality state (AQS) across the world. It is also investigated in full edged or partial lockdown can bring any difference in the degree of AQS. The MERRA-2 is a NASA atmospheric reanalysis data that was launched in 1980 replacing original MERRA (Rienecker et al., 2011), and uses the data assimilation system of upgraded version of Goddard Earthobserving System Model, Version-5 (GEOS-5) (Randles et al., 2017). GEOS-5 is a weather and climate capable model that is composed of oceanic and land components and circulation of the atmosphere (Song et al., 2018). While assimilating atmospheric data GEOS-5 uses grid-points based interpolation of GIS algorithms which can combine both in-situ and remote sensing data (Buchard et al., 2016). This study particularly employed MERRA-2 reanalyzed AOD, PM 2.5 , SO 2 , O 3 , black carbon. MERRA-2 is the rst one multi-temporal reanalysis system, which contained the meteorological aerosol observations assimilated with the global assimilation system (Randles et al., 2017). GEOS-5 is radiatively combined with Goddard Chemistry, Aerosol, Radiation, and Transport (GOCART) provided aerosol module to produce companion gridded dataset of aerosol. Aerosol species are considered to be non-interactive with each other. Emissions of both dust and sea salt are dependent on the speed of surface wind whereas the sulfate and carbonaceous are principally emitted from the combustion of fossil fuel and burning of biomass, additionally standard inventories are prescribed as the sources of black carbon (Randles et al., 2017). The Emission Database for Global Atmospheric Research 4.1 (EDGAR) inventory is attached for prescribing SO 2 emission from anthropogenic sources (Buchard et al., 2014). The corresponding resolution of MERRA-2 and its outputs are on regular 0.625̊ x 0.5̊ longitude by latitude grid with 72 vertical layers extended up to 80 km from the ground.

AIRS data
The AIRS was launched in 2002 is a high-spectral-resolution infrared sounder onboard aqua satellite consisting of a total of 2378 channels which cover 3.7-15.4 µm wavelength with the spectral resolution of that provides global observation of a 40 km vertical atmospheric pro le . AIRS measures CO concentration with a 45 km spatial resolution at nadir and 1650 km of the cross-track swath (Chahine et al., 2006). Due to having the advantage long term high spatial coverage AIRS have been used by several researchers in previous years (Fisher et al., 2010;Xiong et al., 2013;Xiong et al., 2014;Field et al., 2016;Wu et al., 2017). Over the last few years the retrieval algorithm for extracting data from the AIRS has been developed (Suskind et al., 2014) and the CO data has achieved researcher con dence for studying tropospheric CO  . Besides, the long historical records are the reason behind the use of ARIS CO data in this study.

Method for preparation of air quality state (AQS)
A global AQS model has been developed to assess the status of overall air quality. The air quality is assessed based on the concentration level of the pollutants which may harm human life due to high toxicity (Olvera-Garcia et al., 2016). For ful lling the present purpose six major air quality indicators, as well as pollutants, have been considered namely AOD, PM 2.5 , SO 2 , O 3 , black carbon and CO. Concentration of all these pollutants has been investigated based on their toxic capacity and negative impact to the good air quality. Integrated AQS maps have been classi ed into six classes denoting very good to hazardous following World Air Quality Project (WAQP) (https://waqi.info/).

Fuzzy logic and its application
A fuzzy set is a semi-quantitative method based on the training and membership weighting technique which was developed by Zadeh (1996) for the modeling of an uncertain and nonlinear complex system (Sowlat et al., 2011). In the present context, a fuzzy logic system was used to integrate causative air pollutants. As this approach works based on training and membership weighting techniques, here membership value indicates the intensity of the effect of the quantity (Agarwal et al., 2017). In such a semi-quantitative approach, the fuzzy membership value ranges from 0 to 1 where values trending toward 0 indicate a lesser degree of fuzzy relation and trending toward 1 indicate a greater degree of fuzzy relation (Saha and Pal, 2019). This approach is not bound to any universal approach of weight determination for the fuzzy membership (Li and Ma, 2007).
Therefore, any suitable and best tting approach may produce an accurate result. For AQS assessment several researchers have used different knowledge-based approaches to assign fuzzy membership value (Sowlat et al., 2011;Agarwal et al., 2017;Carbajal-Hernández et al., 2012;Olvera-Garcia et al., 2016). This membership system allows us to signify a high concentration of pollutants and determine the successive degree of AQS. Following this rule, the membership values of the pollutants (x) and were assigned from 0 to 1 and with varying degree of con dence (f(x)) and the fuzzy set can be formulated as (eq. 1)where A represents the fuzzy set, x is the element of universal set R, and f(x) signi es the membership function of fuzzy.

Changing air quality state (AQS)
Integrated air quality parameters in the name of AQS in four months across the world are portrayed in gure 4 (a-d). The qualitative pollution state has been categorized into six classes following the World Air Quality Project (WAQP) (https://waqi.info/). These classes have indicated the possible exposure level. In the month of December 2019 and January 2020, the overall pollution level is found to be high in all the highly urbanized and industrialized nations like India, China, Bangladesh, West African countries, South American Countries, and some of the western European countries. Earlier pollution reports and studies have exhibited the same situation (Tilt, 2019;North et al., 2019;Li et al., 2019). But after implementation of lockdown in different countries to stop COVID 19, the different parts of the world have experienced an improvement in air quality (Dutheil et al., 2020;Mahato et al., 2020;Muhammad et al., 2020). As all the nations have not announced a lockdown on the same date and the nature of lockdown is either partial or full-edged, all countries have not experienced a uniform improvement in air quality from a speci c date and the degree of improvement is also not uniform. For example, in China, as the lockdown was announced at the earliest (23 January 2020 were under hazardous to unhealthy AQS category in the pre-lockdown period, and all these countries have shifted to good to very good AQS category during lockdown (February-March, 2020). With the continuation of the lockdown, the gradual improvement of AQS is noticed in all the respective countries (Table 2). In the countries, where lockdown was partial, the degree of quality improvement of air is relatively less than the countries have undergone into full edged lockdown. Tobias et al (2020), Muhammad et al. (2020) have also found a declining trend of pollution level in Barcelona (Spain) and some other major cities of Italy, China, and the USA.
3.3 Rate of change in AQS with the continuation of lockdown condition Figure 5(a-c) shows the percentage of change of AQS from December month of 2020 to successive other months. In China, AQS is improved by 10% between December 2020 to February 2020. This rate is found high (>20%) in European countries like Italy, Spain, France, Germany, Denmark, UK in between December 2020 to March 2020 ( Table 2). Extension of the hotspot to the USA with the casualty of about 54000 deaths and lockdown situations has shown quality improvement of air in the USA. Russia, Latin American countries, and counties of the Middle East are also identi ed under this category. Over the progress of time more, a number of countries have included ambient AQS (Fig. 5a-c). The rate of change is recorded higher in the countries which have experienced full-edge lockdown than countries with partial lockdown with almost the same tenure of lockdown. A wider part of the developing and underdeveloped countries has recorded qualitative degradation as an annual cycle of pollution level where lockdown has not been implemented. In some countries where lockdown is applied in some speci c cities or regions, the average AQS of the entire country does not provide any signi cant change. But this effect is clear in those areas where it is implemented.

AQS of selected cities based on lockdown implementation
As few cities were already identi ed as a hotspot of COVID-19 as per the records of very high death counts and count of infected people, this section has inspected the AQS condition with special emphasis with some less or non-affected cities. In Wuhan, the rst hotspot of this virus, AQS was improved substantially just after starting lockdown. New York, Barcelona, Venice, London, Paris, Brussels, Berlin, Sydney, Amsterdam, Sao Paulo, and other cities of European and American countries have also witnessed a remarkable improvement in AQS as the regions have implemented full-edge lockdown (Table 3). Wuhan, Sydney, and Amsterdam cities have implemented partial lockdown and consequently, the air quality improvement rate is not detected so high (Table 3). New York, Washington DC and Islamabad cities have not experienced any such discernible change in AQS as these areas have not experienced lockdown. In lockdown conditions as most of the sources of air pollution like industries, transport, agriculture farms have been regulated, the pollution level is consequently reduced. The nding of the study is not any discrete event. A few studies have already been done in environmental perspective (Saadat et al, 2020;Sharma et al, 2020;Wang and Su, 2020;Tobias et al, 2020) Aydın et al., (2020) noted that the ozone level has increased signi cantly during the Covid-19 lockdown than pre-lockdown period which has increased the sunlight penetration. They also pointed that the increase in ozone is associated with the decline in PM 2.5 concentration. Thus it can be concluded that the drastic changes in atmospheric elements and pollutants have been observed due to the Covid-19 lockdown and each of the atmospheric changes are associated with each other. Table 3 Analyzing the AQS of selected cities based on lockdown implementation Page 12/22

Conclusion
The present study has clearly explored that around the world the AQS has improved in COVID -19 persuaded lockdown period. The degree of improvement is regulated by the nature of lockdown policies. The full-edge lockdown has exerted a greater impact on the degree of AQS improvement than partial and no lockdown state.
The present study is con ned up to March 2020 but the threat of COVID-19 has been accelerating over time in different countries of the world. So, lengthening of lockdown is expected to further improve the environmental quality. Certainly, this incident has brought colossal economic failure, millions of people are facing even food scarcity and more people may face severe starvation in the coming days in consequence of this but from an environmental and ecosystem point of view, this incident has extended an opportunity to restore the decaying quality of nature. The result has clearly noted that within a very short period nature has started to restore herself. It is really good for not only environmental health but also the health and well-being of the human being. The increase of ozone will obviously provide better security to the entire ecosystem. Pollution freebreathing not only refreshes our mind and body but also gives temporary relief to the millions of people who are suffering from air pollution-related diseases. However, amid curse, some blessing is there with this lockdown compulsion. When the entire world is worried about how to control pollution and paying huge amounts of money to abate this problem, this unfortunate incident has vividly pointed out that the temporary lockdown would be a good alternative to minimize pollution levels and its consequences.

Con ict of Interest
The authors declare that they have no con ict of interest on any issue.
Zhang, X., Wang, H., Che, H.Z., Tan, S.C., Shi, G.Y. and Yao, X.P., 2020. The impact of aerosol on MODIS cloud detection and property retrieval in seriously polluted East China.     This map has been provided by the authors. Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.