key: cord-0862148-kyhhmnw2
authors: Feng, Meili; Ren, Jianfeng; He, Jun; Chan, Faith; Wu, Chaofan
title: Potency of the pandemic on air quality: An urban resilience perspective
date: 2021-09-10
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2021.150248
sha: ee24e6f026f42c415c8a20ca37230864e42deb24
doc_id: 862148
cord_uid: kyhhmnw2

Since the outbreak of COVID-19 pandemic, the lockdown policy across the globe has brought improved air quality while fighting against the coronavirus. After the closure, urban air quality was subject to emission reduction of air pollutants and rebounded to the previous level after the potency period of recession. Different response patterns exhibit divergent sensitivities of urban resilience in regard to air pollution. In this paper, we investigate the post-lockdown AQI values of 314 major cities in China to analyse their differential effects on the influence factors of urban resilience. The major findings of this paper include: 1) Cities exhibit considerable range of resilience with their AQI values which are dropped by 21.1% per day, took 3.97 days on average to reach the significantly decreased trough point, and reduced by 49.3% after the lockdown initiatives. 2) Mega cities and cities that locate as the focal points of transportation for nearby provinces, together with those with high AQI values, were more struggling to maintain a good air quality with high rebounds. 3) Urban resilience shows divergent spatial sensitivities to air pollution controls. Failing to consider multi-dimensional factors besides from geomorphological and economical activities could lead to uneven results of environmental policies. The results unveil key drivers of urban air pollution mitigation, and provide valuable insights for prediction of air quality in response to anthropogenic interference events under different macro-economic contexts. Research findings in this paper can be adopted for prevention and management of public health risks from the perspective of urban resilience and environmental management in face of disruptive outbreak events in future.

In December 2019, the outbreak of novel coronavirus started to spread globally first sight from China. Since March 11, 2020, WHO has declared the COVID-19

pandemic as a global health emergency, which needs international scientific findings and knowledge on COVID-19 to tackle this crisis. Evidence shows that the coronavirus is environmentally sensitive (Carlson et al., 2020) , and exerts propound impacts on the human health and global economies (Chakraborty and Maity, 2020; Saadat et al., 2020) . According to the COVID-19 Impact Survey launched by the UN-Habitat Data & Analytics Unit, this has been a global crisis during which 95% of COVID-19 cases have taken place in urban settlements with J o u r n a l P r e -p r o o f over 1,500 cities affected worldwide (Acuto et al., 2020) . In order to suppress the spread of COVID-19 virus via minimised transmission rates, large-scale lockdown policy was implemented worldwide with travel restriction orders and business close-down measures. Effectiveness of the lockdown policy has been widely studied and proved to be evident with reduced air pollution across the globe (Li et al., 2020b) . Changes in human mobility and living patterns have given rise to substantial improvement of air quality status globally (Chauhan and Singh, 2020; Kanniah et al., 2020; Muhammad et al., 2020; Rodríguez-Urrego and Rodríguez-Urrego, 2020) .

The efforts to control coronavirus did result in a drop of primary gaseous pollutants during the COVID-19. The dramatic perturbations of environmental quality were primarily calculated within a period during the COVID-19 pandemic (usually a few months) and compared with historical measurements (mostly from year 2019). Since early 2020, studies have been carried out on cities in India (Sharma et al., 2020; Mahato et al., 2020) ; Kazakhstan (Kerimray et al., 2020) ;

Italy (Collivignarelli et al., 2020; Fattorini and Regoli, 2020; Zoran et al., 2020) ; Brazil (Dantas et al., 2020; Nakada and Urban, 2020) ; Spain (Baldasano, 2020; Tobías et al., 2020) ; Canada (Adams, 2020) ; and the USA (Bashir et al., 2020; Berman and Ebisu, 2020; Zangari et al., 2020) .

Evidence of improved air quality was also found in Chinese cities including Wuhan (Lian et al., 2020) and other cities in China (Dutheil et al., 2020; Li et al., 2020a; Lian et al., 2020; Wang and Su, 2020; Xu et al., 2020; Zheng et al., 2020b) .

(2020) discovered that reduction of Air Quality Index (AQI) values in provinces of China was related to the number of motor vehicles in use and the percentages of secondary industries. Bouffanais and Lim (2020) found that transmission rates were particularly high in large cities.

Apart from the reduction of anthropogenic emissions (Zheng et al., 2021) , meteorological effects could be superimposed to enhance the air pollutant changes (Salma et al., 2020) . Zheng et al. (2020a Zheng et al. ( , 2021 found that industry sector was the main driver in both the decline and rebound of CO2 emissions in China during the pandemic. The increased CO2 emissions during the lockdown from January to March in 2020 from Guangxi province was induced by the drought event compared with the status in 2019. Xu et al. (2020) also suggested an enhanced impact of COVID-19 on the AQI with low relative humidity and high temperature conditions.

While admitting the fact that pandemic has largely affected the air pollutant reduction, it is also worth while looking at the AQI reduction behaviours from the point view of responsive time.

Berman and Ebisu (2020) indicated that there was particular clear reduction of PM2.5 in urban counties with early lockdown measures. These response patterns show divergent sensitivities to mitigate the lockdown-induced air pollutions. Cities are complex and multi-faceted. Defining and measuring the urban resilience as post-stress phenomenon is challenging with continuously changing environment. To further refer this to the ability of urban system to restore or rapidly recover from disturbances, urban resilience (Meerow et al., 2016) offers a new paradigm for tackling the issue here after the radical measures for the COVID-19 pandemic.

The COVID-19 lockdown is a forced experiment that urban environment is faced with sustainable challenges to optimised capability of resilience to withstand and refrain from unexpected circumstances (Acuto et al., 2020) . Although devastating to human societies, the pandemic has set up an unique chance to examine the adaptive natural urban resilience while the J o u r n a l P r e -p r o o f human activities were minimized during the lockdown periods (Cariolet et al., 2018; Cheshmehzangi, 2020; Duh et al., 2008) . To examine the potency and efficacy of the lockdown policy on the 'bounce-back ability' of urban air quality status, we regard the lockdown policy here as an attempt by the governmental authorities to influence the level of transmission by controlling human activities in various fields and consequently the changes in environment. In history, there are a number of cases where governmental authorities issued policies in restricting human activities to reduce air pollutants during special events, e.g. Beijing Olympic in 2008 (Wang et al., 2010) and In this paper, through analysing the improvement of air quality status, the delayed effects of the lockdown policy are analysed as well as the urban resilience such as the responsive rates of cities. We focus on the post-epidemic performance of AQI of 314 major cities in China. We analysed their differential effects of AQI at both temporal and spatial scales in order to identify the influence factors for urban resilience. The results can unveil the major fault lines and fragilities in current urban systems, as well as provide valuable insights for the development of resilient cities under different macro-economic contexts (Leach et al., 2021).

In previous studies, the underlying relationship between the lockdown policy and air pollution is simply assumed as a cause-effect situation that the air quality is improved due to complicated factors induced by the lockdown initiative. Although it is possible to determine the overall relationship and differences before and after the lockdown, improvement of air quality varies in each city, and the performance gain is largely affected by the feature of 'inherent urban resilience' in each city. We analyse the 'revival' of air quality after the implementation of lockdown measures through the lens of urban resilience, and evaluate it from the perspective of how much and how fast the urban air quality recovers. These responsive features were also analysed with consideration of meteorological effects under different climatic conditions in each city.

The official lockdown date for each province and city was collected from government J o u r n a l P r e -p r o o f announcements (http://www.nhc.gov.cn/), and summarized in Table 1 . As the mean residence time of particle pollutants and submicron particles is up to 100 to 1000 hours in the absence of precipitation (Bolin et al., 1974; Esmen and Corn, 1971) , we use 30 days as the post-lockdown period to analyse the urban air quality changes in this study. Extensive air quality measurements in 314 prefecture cities in China were analysed covering 31 provincial-level administrative regions (See Figure. 1 for the locations of the cities).

Daily average AQI values were collected from January 1, 2020 to May 31, 2020 using the open source API from Envicloud platform (www.envicloud.cn). Meteorological monitoring data including wind speed, air temperature, and rainfall were collected at daily scale for 148 out of the 314 cities during the same period. Detailed information of the datasets can be found in supplementary material Table S1 . 

Being inspired by the classical definition of environmental resilience (Holling, 1973) , the concept of urban resilience has been applied to the risk management of urban systems such as flood, earthquake, sea water, air pollutions and so on. It refers to the ability of urban systems to maintain or rapidly restore the required functions in the event of disturbances, the ability to adapt to change, and the ability to change systems rapidly that limits current or future adaptability (Chelleri, 2012; Meerow and Newell, 2019) . As a highly adaptive complex system, the urban resilience through disturbance-rebuild perspective is critical but not limited to disasters or destructions. The ability of an urban system in face of disturbance is reflected on both adaptation and development. It is hard to compare the reduction of air pollution only, as it is associated with multiple factors such as historical pollutant levels, energy consumption, population, GDP and meteorological conditions. In this case, the effectiveness of lockdown policy on improved air quality is an excellent chance of experiment that reveals the feature of urban resilience with regards to the recovery after external human activities are minimized. To delve into the systematic behavior of an urban system responses, it is worthwhile looking at both sides of the characteristics.

A quantitative approach is applied for measuring and mapping the urban resilience related to air pollution.

[ To analyse the changes of air quality in each city after the lockdown initiative, we start from identifying the trend and abrupt change points of the time series data on air quality in each city. Firstly, the modified non-parametric trend test proposed by Hamed and Rao (1998) (2)

The urban resilience is characterized by the duration of the time span and the improvement of air quality. An urban system with better resilience by nature implies that a larger drop in 'absorption' to the changes, and smaller rebound to an unlikeable 'recovery' of the pollution level.

Adapted from the one-number measurement approach developed by Han and Goetz (2015), the Resilience Index (RI) is thus interpreted as the ratio of drop and rebound, as defined in Eqn. (3) and ( 

To further analyse the spatial difference of air quality improvements in response to the lockdown policy, we analysed the correlation of AQI values with meteorological factors including air temperature, precipitation and wind speed to evaluate the influence of meteorological effects on air pollution variations. Furthermore, we focus on the inherent features of an urban system and analysed the underlying regional disparities of these lagged effects while including geomorphic characteristics of elevation and GPS coordinates of each city. For the 314 cities, the resilience indexes are analysed using the spectral clustering method (Yu and Shi, 2003) . We take the GPS coordinates, the lockdown time, the trough time and the rebound time point with their corresponding AQI values as the input feature vectors i f , = 1, 2,.., iN , where = 314 N denotes the number of cities used in this study. We apply the spectral clustering to find the regional disparities of the resilience patterns for all cities. More specifically, we first build the neighbourhood matrix is a predefined parameter depending on the empirical knowledge of the data. According to (Webb, 2003) , the k value is set to the double number of aimed classes for classification. In this paper,

we set k to 24 as the climatic regions in China is set to 12 categories according to Zheng et al. (2013) . For more information on the spectral clustering algorithm, readers are referred to Yu and Shi (2003) and Webb (2003) .

In this way, we could map the clusters of the lockdown initiative's potency and efficacy to identify the hotspot of resilient cities. Results are analysed with Matlab ® R2020b (Mathworks, Inc., Natick, MA, USA) and visualised in the Geographic Information System (ArcGIS 10.3.2). To delve into the meteorological effects on air pollution variations, the correlation analysis between AQI values and wind speed, air temperature and precipitation is carried out respectively, as shown in Figure 4 . Figure 4 shows that precipitation is not closely correlated with the air pollution levels as most correlation coefficients are close to 0. For cities with high pollution levels, AQI is mostly negatively correlated with air temperature, which indicates that AQI will decrease with an increased air temperature. For cities with very low air pollution levels (i.e. AQI below 50), the increasing air temperature will lead to an increased AQI values. Similar trends are also found in the correlation study between AQI and wind speed, as shown in Figure 4 

Results of the resilience index have shown significant difference among 314 cities across

China. Different response patterns in each city indicate divergent sensitivities to mitigation of traffic-related human impacts, as well as the capabilities for recovery. For each city, we evaluate its urban resilience feature as the ratio of drop and rebound to delve into the integrated characteristics of recovery and resistance.

[ 

This paper is developed based upon the state-of-art that the COVID-19 induced pandemic has largely influenced air quality with reduced air pollution as induced in the background.

Extensive studies have been carried out with consideration of multiple coupling effects from anthropogenic emissions to climatic and meteorological effects. In this paper, the study is based on fact of improved air qualities and focuses on the post-epidemic performance of air pollution improvement from the lens of urban resilience. It analyses the delayed effects of the lockdown policy by looking at the responsive rate from temporal perspectives beyond the actual values of reduced air pollutants. This provides a benefit as the threshold of lockdown time is aligned and the post lockdown air pollution is measured for its potency behaviours at the same temporal scale. The developed resilience index is unit-less but represents the urban system behaviors. The method we have developed and demonstrated here will support efforts to evaluate and cope with changes in air pollution levels in near real time, while it usually takes a much longer time for climatic and meteorological effects. This approach is potentially applicable to other global cities where anthropogenic emissions are dominant and were evident declines during the lockdown measure.

An exploration of the urban resilience features in regard to traffic-related air pollution controls has become a new paradigm for air quality management. Results of the urban resilience features should be referenceable in determining the goals of future traffic-related emission reduction policies. Comprehensive urban resilience framework, which addresses the multiplicity J o u r n a l P r e -p r o o f of city management requirements, can be explored in the further studies. Extending from the COVID-19 stress to other similar cases, lessons can also be referenced for the future urban management policies through the resilience perspectives. Under the challenge of achieving Sustainable Development Goals in the present and post-COVID19 era, this study provides good arguments to address the urban resilience on air pollution, which may exert influence to our stakeholders and decision makers to create a long-term strategic in the urban context (i.e. via Urban Master Plans). It is interesting to reflect on how to deliver sustainable environmental management strategies for different cities and achieve better urban resilience especially in megacities. Results also showed the spatial hotspots of the most evident cities that received the most beneficiary from the lockdown policy. These findings are vitally pushing forward on improving future urban resilience dealing with the urban air pollution in Chinese cities especially on the response and recovery processes by our findings.

Seeing COVID-19 through an urban lens

Air pollution in Ontario, Canada during the COVID-19 State of Emergency

COVID-19 lockdown effects on air quality by NO2 in the cities of Barcelona and Madrid (Spain)

Correlation between environmental pollution indicators and COVID-19 pandemic: A brief study in Californian context

Changes in US air pollution during the COVID-19 pandemic

Residence time of atmospheric pollutants as dependent on source characteristics, atmospheric diffussion processes and sink mechanisms

Cities try to predict superspreading hotspots for COVID-19

Assessing the resilience of urban areas to traffic-related air pollution: Application in Greater Paris

Misconceptions about weather and seasonality must not misguide COVID-19 response

COVID-19 outbreak: Migration, effects on society, global environment and prevention. Science of The Total Environment

Sponge City" in China-A breakthrough of planning and flood risk management in the urban context

Decline in PM2.5 concentrations over major cities around the world associated with COVID-19

From the «Resilient City» to Urban Resilience. A review essay on understanding and integrating the resilience perspective for urban systems

Characteristics and source attribution of PM2.5 during 2016 G20 Summit in Hangzhou: Efficacy of radical measures to reduce source emissions

Introduction: The City During Outbreak Events

Lockdown for COVID-2019 in Milan: What are the effects on air quality? Science of quality of the megacity Delhi

Urban resilience for whom, what, when, where, and why? Urban Geography

Defining urban resilience: A review

COVID-19 pandemic and environmental pollution: a blessing in disguise?

COVID-19 pandemic: Impacts on the air quality during the partial lockdown in São Paulo state

Association between changes in air pollution levels during the Beijing Olympics and biomarkers of inflammation and thrombosis in healthy young adults

Air quality during the COVID-19: PM2.5 analysis in the 50 most polluted capital cities in the world

Environmental perspective of COVID-19

What can we learn about urban air quality with regard to the first outbreak of the COVID-19 pandemic? A case study from central

Effect of restricted emissions J o u r n a l P r e -p r o o f Journal Pre-proof during COVID-19 on air quality in India

Climate change regionalization in China

Changes in air quality during the lockdown in Barcelona (Spain) one month into the SARS-CoV-2 epidemic

A preliminary assessment of the impact of COVID-19 on environment-A case study of China

Air quality during the 2008 Beijing Olympics: secondary pollutants and regional impact

Changes in air quality related to the control of coronavirus in China: Implications for traffic and industrial emissions

Statistical pattern recognition

Possible environmental effects on the spread of COVID-19 in China

The effect of emission control on the submicron particulate matter size distribution in Hangzhou during the 2016 G20 Summit

Multiclass spectral clustering

The influence of serial correlation on the Mann-Whitney test for detecting a shift in median

Air quality changes in New York City during the COVID-19 pandemic

Satellite-based estimates of decline and rebound in China's CO2 emissions during COVID-19 pandemic

Changes in China's anthropogenic emissions and air quality during the COVID-19 pandemic in 2020

Significant changes in the chemical compositions and sources of PM2.5 in Wuhan since the city lockdown as COVID-19

The climate regionalization in China for 1981-2010

Assessing the relationship between surface levels of PM2.5 and PM10 particulate matter impact on COVID-19 in

The authors would like to thank the anonymous reviewers for their valuable suggestions for this paper. We would also like to send our gratitude for the support by the National Natural