Impact of COVID-19 lockdown upon the air quality and surface urban heat island intensity over the United Arab Emirates.

The 2019 pandemic of Severe Acute Respiratory Syndrome-Corona Virus Diseases (COVID-19) has posed a substantial threat to public health and major global economic losses. The Northern Emirates of the United Arab Emirates (NEUAE) had imposed intense preventive lockdown measures. On the first of April 2020, a lockdown was implemented. It was assumed, due to lower emissions, that the air quality and Surface Urban Heat Island Intensity (SUHII) had been strengthened significantly. In this research, three parameters for Nitrogen Dioxide (NO2), Aerosol Optical Depth (AOD), and SUHII variables were examined through the NEUAE. we evaluated the percentage of the change in these parameters as revealed by satellite data for 2 cycles in 2019 (March 1st to June 30th) and 2020 (March 1st to June 30th). The core results showed that during lockdown periods, the average of NO2, AOD, and SUHII levels declined by 23.7%, 3.7%, and 19.2%, respectively, compared to the same period in 2019. Validation for results demonstrates a high agreement between the predicted and measured values. The agreement was as high as R2=0.7, R2=0.6, and R2=0.68 for NO2, AOD, and night LST, respectively, indicating significant positive linear correlations. The current study concludes that due to declining automobile and industrial emissions in the NEUAE, the lockdown initiatives substantially lowered NO2, AOD, and SUHII. In addition, the aerosols did not alter significantly since they are often linked to the natural occurrence of dust storms throughout this time of the year. The pandemic is likely to influence several policy decisions to introduce strategies to control air pollution and SUHII. Lockdown experiences may theoretically play a key role in the future as a possible solution for air pollution and SUHII abatement.

Introduction

The emergent Coronavirus disease 2019 (COVID-19) is a transmittable disorder characterized by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) (Islam et al., 2020). By massive human-to-human transmission, COVID-19 has deeply hit the world and prompted a spike in the human mortality rate and massive economic casualties around the world (Bukhari and Jameel, 2020). The number of cases of Covid-19 globally hit approximately 34 million by the first of October 2020, while the number of deaths reached 1 million (WHO, 2020a). In late 2019, the first case of COVID-19 was identified in China and since then has spread very quickly throughout the world (Q. Li et al., 2020). On 11 March 2020, the World Health Organization (WHO) announced the novel coronavirus disease as a pandemic (WHO, 2020b). The WHO COVID-19 dashboard (https://covid19.who.int/) can be used to find the specifics of global COVID-19 cases. The first COVID-19 infection was reported in the United Arab Emirates (UAE) on 29 January 2020, and the first death was recorded on 21 March 2020. Since then, as per the Ministry of Health and Prevention (https://www.mohap.gov.ae/), there has been an alarming surge in active and death cases due to COVID-19.

In people with cardiovascular and respiratory disorders, the mortality risk of COVID-19 is substantially higher (Archer et al., 2020; Isaifan, 2020; Y. Zhu et al., 2020). Cardiovascular and respiratory disorders are also closely related to air pollution. Emissions leading to respiratory health conditions from primary pollutants containing particulate matter (aerosols) and gases such as nitrogen dioxide (NO2) also have adverse environmental effects like soil and water acidification (Griffin et al., 2019; Mulenga and Siziya, 2019; Xu et al., 2020). The WHO (2020c) reports that approximately 4.2 million inhabitants die worldwide per year from factors primarily related to air pollution, which is currently increased, related to COVID-19 infected patients (Conticini et al., 2020; Wu et al., 2020). Also, recent worldwide studies on the effect of climate on the spread of the COVID-19 (Briz-Redón and Serrano-Aroca, 2020), showed that temperature and humidity are not crucial factors in the COVID-19 transmission, while precipitation, radiation, and wind speed have not been investigated in details.

Urban Heat Island (UHI) is among the most noticeable and frequently reported urbanization climatological consequences, by which urban and suburban regions are hotter than surrounding areas (rural/nonurban) (Hu and Brunsell, 2013; Miles and Esau, 2017). The negative influence of UHI is extensively described in the literature. For example, UHI causes rises in energy demand (Alghamdi and Moore, 2015), which implicitly leads to global climate change (Alghamdi and Moore, 2015), environmental degradation (Lin et al., 2017), air pollution (L. Zhu et al., 2020), impact to human comfort and health (Schwarz et al., 2011), and a degradation of ecosystem function (Keeratikasikorn and Bonafoni, 2018). All of these also play a significant role in the rise in the rate of COVID-19 cases (H. Li et al., 2020; Mukherjee and Debnath, 2020) and leads to heat-related deaths (Cui and De Foy, 2012; Lowe, 2016). UHI is measured based on air temperature, while satellite-derived Land Surface Temperature (LST) data is used to estimate the surface UHI (SUHI). Moreover, the calculation of SUHI Intensity (SUHII) requires observations from paired LST locations in both urban and rural/nonurban areas (Cui et al., 2019). LST is described as the surface temperature of the Earth's skin, playing an important role in the interchange of heat and energy among land surfaces and the atmosphere to assess changes in the environment (Moradi et al., 2018).

Distinct mitigation initiatives such as social distancing, cluster and whole lockdowns, comprehensive travel bans, mass quarantines, etc. have been introduced globally to mitigate the COVID-19 pandemic risk. Such risk mitigation initiatives have had a significant effect on socio-political ties and economic development at local and global levels (Ranjan et al., 2020a). Nevertheless, due to the reduction of anthropogenic-based pollutants, such precautionary strategies to prevent COVID-19 transmission have significantly enhanced air quality. The most critical challenge in the 21st century is the degradation of air quality worldwide due to the different sorts of anthropogenic interventions (Mehdipour and Memarianfard, 2017; Motesaddi et al., 2017). At such a moment, lockdown episodes enforced for the COVID-19 pandemic mitigation resulted in anti-environmental activities to cease as a byproduct. As a result, the level of air quality in the various continents of the Earth significantly improved since the COVID-19 pandemic began. Tobías et al. (2020) recorded a 45%, 51%, 31%, and 19% decline in the level of PM10, NO2, SO2, and CO, respectively, over Barcelona in Spain within the lockdown span of a month. Isaifan (2020) recorded a substantial decrease in NO2 and carbon emissions (30% and 25%, respectively) in China, which correlated to industrial lockdowns. Karuppasamy et al. (2020) documented a 55% contraction in NO2 in India during the lockdown. A recent analysis in the Middle East has also shown a decrease in the number of air pollutants in Morocco of 75%, 49%, and 96% for PM10, SO2, and NO2, respectively (Otmani et al., 2020). Moreover, in Iran, Nemati et al. (2020) recorded noticeable reductions in air pollution during the pandemic. Likewise, during the COVID-19 pandemic, many other recent studies show a significant decline in the level of air pollutants such as NO2, SO2, PM10, PM2.5, CO, etc. globally (Archer et al., 2020; Collivignarelli et al., 2020; Dantas et al., 2020; Islam et al., 2020; Kaplan and Avdan, 2020; Kerimray et al., 2020; Nakada and Urban, 2020; Ranjan et al., 2020a; Wang and Su, 2020). However, while studies generally showed a reduction in the primary pollutants, some studies showed no reduction in air pollution. For example, Q. Li et al. (2020), L. Li et al. (2020) and H. Li et al. (2020) investigated the air quality changes during the COVID-19 lockdown over 41 cities in the Yangtze River Delta Region. The results showed a decrease in some of the main pollutants, but still reported high PM2.5 and increasing trend during the lockdown by over 20%. These contradictory results emphasize the need for continued detailed investigations on the topic.

The aforementioned studies concentrate primarily on evaluating the level of air pollutants during the pandemic scenario of COVID-19. Although a considerable correlation between air pollutants and LST was identified in some research (Alseroury, 2015; Feizizadeh and Blaschke, 2013; Hashim and Sultan, 2010; Kahya et al., 2016; Weng and Yang, 2006), the SUHII variability during the pandemic scenario has not yet been investigated. Like other nations, the UAE forced the shutdown of industries, public transit, airlines, vehicles, and other anthropogenic operations in all Emirates from the 1st April to the end of June 2020 to mitigate COVID-19 spread. Additionally, the UAE imposed a curfew daily between 8 pm and 6 am for residents (https://www.wam.ae). A full or partial lockdown is hypothesized to lead to improved air quality and reduce the SUHII since the lockdown is associated with many anthropogenic activities that directly impact the level of emissions. To our knowledge, no studies have examined the impacts of COVID-19 lockdowns on air quality and SUHII throughout the UAE. This current study was undertaken to investigate the potential impacts of COVID-19 lockdown operations on NO2, Aerosol Optical Depth (AOD), and SUHII in the Northern Emirates of the UAE (NEUAE). This study also aims to validate satellite data over the study area using ground station records. Concentrating on the NEUAE, this work directly addresses the impact of lockdowns on air quality and SUHII, and is a resource for the science community and environmental protection policymakers, as well as resource in the development of an action plan to improve air quality and SUHII.

Data and methods

Study area

The United Arab Emirates (UAE) consists of seven federation Emirates (States): Abu Dhabi (ADH), and six northern Emirates, including Ras Al Khaymah (RAK), Dubai (DUB), Umm Al Qaywayn (UMQ), Sharjah (SHJ), Fujayrah (FUJ), and Ajman (AJM) (Fig. 1a). The Northern Emirates of the UAE (NEUAE) were chosen as a study area located approximately between latitudes 26° 365 to 26° 335 N and longitudes 54° 530 to 56° 240 E. Topography of the study area is mostly flat, with a mountain chain reaching an altitude of ~1800 m in the northeastern and eastern parts (Fig. 1b). The NEUAE is typically an arid region with a humid environment, located in the Arabian Peninsula's Eastern corner. It borders the Arabian Gulf to the north and the Gulf of Oman to the east (Fig. 1). The annual average air temperature is about 28 °C. It is much hotter (up to 45 °C) in the summer season (June to August) and colder (down to 10 °C) in the winter season (December to February) (https://www.ncm.ae/en/climate-reports-yearly.html?id=26). In the hot months, dust storms generally occur (Barbulescu and Nazzal, 2020). Over 80% of the yearly rainfall occurs in the winter season (FAO, 2008).

Fig. 1

Location map of the study area: (a) study area and major factories, (b) elevations and spatial distribution of meteorological and air quality stations, and (c) urban areas, and reference.

Around 71% of the UAE population is based in the NEUAE. Around 50% of the population resides in DUB, followed by 31% in SHJ, and only 1% resides in UMQ, as per the Federal Competitiveness and Statistics Authority (https://fcsa.gov.ae/ar-ae/Pages/home.aspx). Besides that, as documented by the United Nations (2019), more than 86.5% of the NEUAE people live in urban areas situated on the coastline (Fig. 1c). Furthermore, over the past two centuries, fast and pervasive economic and political change has caused increased population explosion, accelerated urbanization, higher energy consumption, and increased vehicular and industrial emissions (Alawadi et al., 2018). This all adds to a net increase in all anthropogenic activities.

Data

Between March to June 2019 and 2020, two different air pollutants NO2 and AOD, and night LST data were collected for the NEUAE. The average monthly NO2 and AOD data were acquired from Google Earth Engine (GEE). NO2 is measured by the Sentinel-5p TROPOMI (Tropospheric Monitoring Instrument) mission of Copernicus ESA. While AOD and night LST were obtained by MODIS MAIAC (MCD19A2) and MODIS Aqua (MYD11A2), respectively. Due to the finite temporal scope of Sentinel-5p data, it is worth noting that the baseline was 2019. A summary of the datasets used in this study is shown in Table 1 .

Table 1

Summary of the datasets used in this study.

Data source	Parameter	Spatial resolution	Temporal resolution	Data access link
Sentinel5p TROPOMI	NO2	3.5 × 5 km	Daily
MODIS MAIAC	AOD	1 × 1 km	Daily	https://lpdaac.usgs.gov/
MODIS/ AQUA	Night LST	1 × 1 km	8-day	https://lpdaac.usgs.gov/
SRTM	DEM	30 × 30 m	–	https://earthexplorer.usgs.gov/
NCM	NO2 &PM2.5	–	Monthly	–
DM	Tamin	–	Monthly	–

TROPOMI/Sentinel-5p data (NO2)

TROPOMI was launched on 13th October 2017 as a passive hyperspectral nadir-viewing imager aboard the Sentinel-5 Precursor satellite, which is also recognized as Sentinel-5P (Veefkind et al., 2012). Sentinel-5P is a near-polar orbiting sun-synchronous satellite positioned at an altitude of 817 km in an ascending node with an equator crossing time at 13:30 (local time) offering daily worldwide coverage. Furthermore, since July 2018, TROPOMI delivered calibrated data from its nadir-viewing spectrometer that measures reflected sunlight in the ultraviolet, visible, near-infrared, and shortwave infrared with seven bands, where the fourth band spectral range is 405–500 nm, which could be used for NO2 monitoring (Venter et al., 2020). Recent works have shown that measurements of TROPOMI are quite well associated with actual ground measures of NO2 (Griffin et al., 2019; Lorente et al., 2019). TROPOMI products included in this research are L3 offline version products. Band four's spectral and spatial resolutions are 0.55 nm and 5.5 × 3.5 km, respectively, and the signal to noise ratio is also massively enhanced (Cheng et al., 2019).

MODIS data (AOD and LST)

In a near-polar solar-synchronous circular orbit, the Moderate Resolution Imaging Spectroradiometer (MODIS) was launched onboard NASA's Aqua and Terra satellites. Local crossing times are approximately 01:30 and 13:30 for the Aqua satellite and 10:30 and 22:30 for the Terra satellite. MODIS has a 2330 km (cross-track) swath and provides near-global coverage daily. MODIS is an imaging radiometer with 36 wavebands, covering the wavelength spectrum from the visible to the thermal infrared. AOD data was possessed from the cloud-masked MCD19A2-v6 product, which is a MODIS Terra and MODIS Aqua combined AOD retrieved with the Multi-Angle Implementation Atmospheric Correction (MAIAC) algorithm (Lyapustin et al., 2018). This dataset has previously been used effectively to map ground-level PM2.5 concentrations (Wei et al., 2019). In this analysis, AOD at 550 nm (i.e., the green band) was utilized because of its superior accuracy as shown by previous studies (Lyapustin and Wang, 2018). Concerning night LST data, for 2019 and 2020, MODIS Aqua's MYD11A2-v6 performance entails an 8-day composite of LST data at night from 1 March to 30 June 2019 and 2020. In the long time series, the 8-day LST composite products will undervalue the quantity of gaps created by clouds or other unwanted circumstances, that are beneficial for SUHII studies to conduct a spatial comparison (Hu and Brunsell, 2013; Schwarz et al., 2011). Modern Time LST products (v6) have resolved past version accuracy issues. Additionally, measure errors and validation assessments in bare land and arid regions currently suggest their use in these areas (Lu et al., 2018). For the present work, MAIAC AOD and night LST data on a 1 km nadir resolution were retrieved from the NASA Land Processes Distributed Active Archive Center (Table 1).

SRTM data (DEM)

The SRTM DEM data was downloaded from the USGS Earth Explorer (Table 1) to pick a suitable reference location for SUHII calculation. The area's topography is predominantly flat but increases slowly from the northwest to southeast and east, reaching approximately 1830 m (Fig. 1b). The Digital Elevation Model (DEM) 30 m resolution was used for this study. The DEM model was obtained from the Shuttle Radar Topography Mission (SRTM).

Meteorological data

The air quality and meteorological records utilized in our study were collected from the Municipality of Dubai (DM) and the National Center of Meteorology (NCM) for validation. From the aforementioned sources, three parameters were obtained: NO2, PM2.5, and minimum air temperature (Tamin) for each of the years 2019 and 2020, from March to June. Tamin was chosen for the highest degree of matching with MODIS night LST overpass time, which is closer to Tamin (Alqasemi et al., 2020). Likewise, PM2.5 was selected as AOD, which is considered a proxy for PM2.5 (Fan et al., 2020; Venter et al., 2020).

Methods

Data pre-processing

Usually, the products MCD19A2 and MYD11A2 were available in the sinusoidal grid projection that was re-projected to the WGS 1984 geographic coordinate system. In addition, a mosaic of two tiles was created with MYD11A2 data. The scale factor was multiplied by both MCD19A2 and MYD11A2. Then, we subtracted 273.15 to produce a night LST in Celsius, from the reported Kelvin. In the process of producing a seamless dataset, every unrealistic LST data, values above 100 °C and/or below −50 °C, or any value outside the acceptable range was labeled as no data and ignored. Only reliable pixels and high-quality data were used as per the layer of Quality Assessment (QA). Missing values were amended using the mean value of each series. Afterward, night LST data was aggregated to a monthly period by averaging the 8-day composite data to obtain monthly average night LST. As for TROPOMI-Sentinel-5P, NO2 was resampled to 1 × 1 km to ensure conformity with a spatial resolution of MODIS data (i.e., AOD and night LST). Finally, the boundary polygon that defined the study area was used to clip all datasets.

As a preventive strategy for COVID-19, the three months of April, May, and June (AMJ) were averaged as one period applied for all datasets throughout the NEUAE. The March month was collected and processed for pre-lockdown in 2020 for NO2, AOD, and night LST. Similarly, for the identical time spans as the lockdown and pre-lockdown periods, the mean of datasets was obtained for 2019 for comparative analysis, as well as utilized for evaluating the correlation with measured data from ground stations.

SUHII calculation

The night SUHII is a distinctive feature of arid regions, as cited by previous researchers (Alahmad et al., 2020; Clinton and Gong, 2013; Lazzarini et al., 2013). Additionally, the lowest surface temperature is reported at midnight, making Aqua night data an optimal choice near the SUHII maxima. SUHII is typically measured from the LST contrasts between urban and surrounding areas (will be cited here as reference). Consequently, choosing a reliable reference is important. In different tests, however, the approaches for determining the reference differed extensively in various studies. The reference should not be influenced by urban, high altitude, vegetation, or water (Hu et al., 2019). Therefore, in this study, the reference was defined as bare land using chronological images from Google Earth Pro (GEP) with elevation lower than 50 m as determined by the 30 m DEM dataset, to prevent the cooling effect on the SUHII quantification. Finally, by subtracting the average night LST of the reference from the night LST of all pixels, the SUHII was calculated for every pixel utilizing the following equation (Eq. (1)):where LSTPX is the night LST of all pixels, and LSTRF is the mean night LST of the reference.

Change rate (concentration)

It is critical to understand the change in the NO2, AOD, and SUHII concentrations throughout the lockdown period. After retrieving the data on pollutants and SUHII, the Spatiotemporal pattern of average levels of NO2, AOD, and SUHII was categorized into four groups; (i) pre-lockdown (March 2020); (ii) during the lockdown period (AMJ, 2020); (iii) the same Pre-lockdown dates (March 2019); (iv) the same 2019 lockdown dates. Furthermore, the change rates were calculated utilizing Eq. (2) to reflect the percentage of change in the study area's NO2, AOD, and SUHII levels during the lockdown period related to the past year (i.e., 2019) for the same period, which were also compared to pre-lockdown in March 2020. The change rate of variation between March and AMJ throughout 2019 and 2020 was also examined.where, X is NO2, AOD, or SUHII for the selected periods in 2019 and 2020. Also, the correlation was analyzed by calculating the regression coefficients (R2) among stations and satellite data based on the availability of the metrological and air quality station data. The flowchart for the methodological workflow is presented concisely in Fig. 2 . The sequence of data collection, pre-processing, clipping, resampling, data extractions, and SUHII measurement, change rate, and validation are described in Fig. 2. For the processing, analysis, and exhibition, the ESRI ArcGIS™ version 10.4 software framework and Microsoft Office Excel were used.

Fig. 2

Workflow flowchart adopted in the present study.

Results and discussion

Nitrogen dioxide (NO2)

A dramatic drop in NO2 concentration was recorded during the lockdown period (i.e., AMJ, 2020) in the NEUAE due to the COVID-19. Fig. 4 depicts the substantial spatiotemporal changes in NO2 levels throughout lockdown and pre-lockdown in the NEUAE. The analysis showed that the six Emirates experienced a decline in NO2 levels because of restricted transportation and factories' closure. This result is compatible with other earlier studies undertaken in various regions around the globe (Baldasano, 2020; Dantas et al., 2020; Islam et al., 2020; Kaplan and Avdan, 2020; Sharma et al., 2020). Fig. 3a indicates that RAK experienced the highest decline in NO2 levels among the six Emirates (18.6%), followed by UMQ (~18%), AJM (13%), and FUJ (11.6%) compared with the average NO2 levels throughout lockdown with the average levels immediately pre-lockdown, for other Emirates relative to DUB (7.6%) and SHJ (5.4%). In addition, our study indicates that, as shown in Fig. 3a, the average decline in NO2 concentration throughout NEUAE was substantial 12.2%. This decline should be noted that NO2 emission is closely connected to fuel combusting within the factories area (Figs. 1a & 4 ). Therefore, the limitations on these areas' operations are hypothesized to correlate to the substantial drop in NO2 level throughout the lockdown period. Additionally, within March 2019, the NO2 level was lower than March 2020, as shown in Fig. 3a and c, which may be due to higher precipitation during March 2019.

Fig. 4

Spatiotemporal distribution of average NO2 over the NEUAE: (a) March 2019, (b) average AMJ 2019, (c) pre-Lockdown (March 2020), (d) during Lockdown (AMJ 2020), (e) percentage of change between March 2019 and 2020, and (f) percentage of change between average AMJ 2019 and 2020.

Fig. 3

Average concentrations of (a) NO2, (b) AOD, and (c) SUHII in the NEUAE before and during the COVID-19 lockdown, compared to the same 2019 time period.

A substantial decrease in NO2 is visible in the study area by comparing the NO2 values during the 2020 lockdown period with the identical 2019 time period. The maximum calculated decline was for UMQ (27.8%) followed by RAK (26.8%), SHJ (~26%), FUJ (24.3%), AJM (19.8%) and DUB (18.7%). The average reduction was 23.7% with the whole study area (Fig. 3a). These results are in accordance with the results of previously published research. Islam et al. (2020) recorded identical outcomes for Bangladesh, Agarwal et al. (2020) recorded an average reduction in NO2 during the lockdown in China by 49% and in Mumbai (India) by more than 76%. In the Middle East (Morocco), NO2 levels were decreased during the lockdown phase as well (Sekmoudi et al., 2020). This decline in NO2 concentrations shows that lockdown initiatives related to the COVID-19 pandemic significantly impacted NO2 concentrations in the NEUAE as well as elsewhere.

Aerosol optical depth (AOD)

The average AOD values during the lockdown period were lower than the mean AOD values in 2019 for the equivalent time frame, as shown in Fig. 5 . The analysis indicates that all six Emirates encountered a decline in AOD concentration because of restricted aerosol sources, specifically, burning biomass, emissions from factories, vehicles, heavy transport, machinery (Ranjan et al., 2020b), and dust (Khuzestani et al., 2017). Therefore, a reduction in AOD due to the restrictions of industrial and automobile movement is reasonable. The spatiotemporal variations in AOD concentrations in the NEUAE (Fig. 5) show that AJM with 5.7% and UMQ with 5.3% were the largest declines, followed by SHJ with 3.4%, FUJ and RAK, both with 3.1%, and, in Dubai, the lowest drop, 1.7%. The rate of change was 3.7% throughout the overall study area (NEUAE), as shown in Fig. 3b. Similarly, a common declining trend of AOD has been recorded in China (Fan et al., 2020; Filonchyk et al., 2020), India (Gautam, 2020; Pathakoti et al., 2020; Ranjan et al., 2020b), and South Asia (Zhang et al., 2020).

Fig. 5

Spatiotemporal distribution of average AOD over the NEUAE: (a) March 2019, (b) average AMJ 2019, (c) pre-lockdown (March 2020), (d) during lockdown (AMJ 2020), (e) percentage of change between March 2019 and 2020, and (f) percentage of change between average AMJ 2019 and 2020.

During the average AMJ, the level of AOD increased significantly compared to March (Fig. 5). This may be due to the fact that rainfall throughout these months (AMJ) is scarce, and the gusty winds ensure that dust storms become frequent (Barbulescu and Nazzal, 2020; Karagulian et al., 2019). Furthermore, according to Al Otaibi et al. (2019), AOD typically rises over the gulf nations in hot summer months. Consequently, there is generally a rise in the AOD level in AMJ when compared to March. However, the change rate between March and the average AMJ in 2020 decreased relative to the past year for all six Emirates, as shown in Fig. 3b. In 2019, the change rate over the NEUAE was 26.3%, whereas it dropped in 2020 to 3.6%. Notably, AOD emission is linked with the major industries, particularly in Dubai (Figs. 1a, 5). Furthermore, Dubai recorded the smallest reduction in AOD. The most populous and established region of the UAE is the Emirate of Dubai, which includes the most industrialized regions. Therefore, certain additional considerations may explain the relatively small reduction in Dubai.

Surface Urban Heat Island Intensity (SUHII)

SUHII specifically portrays urbanized areas and mountains, as shown in Fig. 3(a–d). A drop in SUHII levels was also recorded, like NO2 and AOD (Fig. 3f), also attributed to the partial or full nationwide lockdown. The result reveals that during the specific time frame, nighttime SUHII levels over this duration were comparatively less than the 2019 levels (Figs. 3c & 6 ). The decline was generally identified in all Emirates, varying from 12.3% to 28.6%, in which the average drop is 19.2% across the whole study area.

Fig. 6

Spatiotemporal distribution of SUHII over the NEUAE: (a) March 2019, (b) average AMJ 2019, (c) pre-Lockdown (March 2020), (d) during Lockdown (AMJ 2020), (e) percentage of change between March 2019 and 2020, and (f) percentage of change between average AMJ 2019 and 2020.

The maximum SUHII concentrations are found in FUJ (28.6%) and RAK (23%) as displayed in Fig. 3c, which may be due to the elevations and type of rocks. Fig. 6(a–d) also reveals that SUHII increases during AMJ relative to March, related to heat emission in hot months (e.g., air conditioning system). Nevertheless, the change rate in 2020 between March and AMJ is less than in 2019. SUHII change rate dropped from 18.3% in 2019 to 6.7% in 2020 over the NEUAE, as exhibited in Fig. 3c. Overall, the results indicate a reduction in SUHII values as a result of the shutdown of anthropogenic activities and sources of heat emissions such as industrial processes, power plants, flight, and transport.

Validation

In this section, the validations of derived data from satellites with the measured data from actual ground stations are investigated. The comparisons between derived NO2, AOD, and night LST plotted against the measured data NO2, PM2.5, and minimum air temperature (Tamin), respectively, are presented in Fig. 7 . Based on the availability of ground station data, the validation of the NO2 and PM2.5 data were made using thirteen and five air quality monitoring stations measurements, respectively, from March to June 2019. Tamin data covers the whole period (March–June) in 2019 and 2020 at four metrological stations (Fig. 1b). Night Aqua LST data is validated with Tamin because the Aqua satellite crosses the equator at night, close to the Tamin; therefore, this is the most appropriate temperature measurement for validation purposes. In addition, as mentioned before, the AOD is considered a proxy for (PM2.5); thus, AOD is plotted versus PM2.5.

Fig. 7

The scatter plots for validation of the derived data from satellites and measured data from ground-measuring stations; (a) measured NO2 and derived NO2, (b) measured PM2.5 and AOD, and (c) minimum air temperature (Tamin) and night LST.

Our comparisons show that the TROPOMI Sentinel-5P NO2 is highly correlated with the air quality monitoring stations data with R2 = 0.70, as shown in Fig. 7a. Likewise, Fig. 7(b) shows the scatter plot of measured PM2.5 versus MODIS MAIAC AOD. The statistical analyses showed a high coefficient of determination (R2= 0.60). Similarly, a high correlation was found between the MODIS Aqua night LST and the Tamin from station measurements (R2 = 0.68), as displayed in Fig. 7c. Overall, a good agreement is found between satellite data and actual measured data; hence, the satellite observations are a valuable resource for studying air pollution and SUHII over large geographic regions.

Conclusions

The impact of anthropogenic activities lockdown due to the COVID-19 pandemic on air quality and SUHII in the NEUAE was studied by examining NO2, AOD, and SUHII levels and evaluating variations in spatial distribution. To demonstrate how restrictive anthropogenic activities throughout the COVID-19 lockdown minimized the air pollutants and SUHII in NEUAE, satellite data of different parameters were used and compared between 2019 and 2020. As predicted, the current investigation discovered that NO2, AOD, and SUHII concentrations across the NEUAE decreased during the pandemic lockdown. The largest average drop was in NO2 (23.7%) followed by SUHII (19.2%) and AOD (3.7%) throughout the lockdown period compared with the same period in 2019.

This study showed that the satellite derived measurements of the selected air pollutants and SUHII data are highly correlated with the actual measured data. Therefore, satellite data is a significant and reliable resource for researching air quality and SUHII because of the spatial coverage and cost-effectiveness of the data, especially for developing countries like the UAE. We conclude that our study has established a benchmark paradigm that will potentially assist the authorities concerned with the potential management of air quality and SUHII in the UAE by decision-makers, particularly on industrial and vehicle pollution restrictions. The drawbacks of this work are that some datasets from ground stations were incomplete. Additionally, the baseline was only a single year (i.e., 2019) due to the limited temporal scale of TROPOMI/ Sentinel-5P data, and therefore the outcomes may change slightly when providing extra datasets. Further study is recommended to examine the correlation of air pollutants and SUHII with COVID-19 cases in the UAE.

CRediT authorship contribution statement

Abduldaem S. Alqasemi: Conceptualization, Methodology, Software, Validation, Formal analysis, Data curation, Writing- Original draft preparation and editing, Visualization, Investigation. Mohamed E. Hereherb: Conceptualization, Supervision, Writing- Reviewing and Editing. Gordana Kaplan: Conceptualization, Formal analysis, Software, Writing- Reviewing and Editing. Ayad M. Fadhil Al-Quraishi: Supervision, Writing- Reviewing and Editing. Hakim Saibi: Supervision, Writing- Reviewing and Editing. All authors have read and approved the final version of the manuscript for publication.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.