Integration of Google Earth Engine and Aggregated Air Quality Index for Monitoring and Mapping the Spatio-Temporal Air Quality to Improve Environmental Sustainability in Arid Regions

Abstract

Egypt must present a more thorough and accurate picture of the state of the air, as this can contribute to better environmental and public health results. Hence, the goal of the current study is to map and track the spatiotemporal air quality over Egypt’s Qena Governorate using remote sensing data. The current investigation is considered a pioneering study and the first attempt to map the air quality index in the studied area. Multisource remote sensing data sets from the Google Earth Engine (GEE) were used to achieve this. The first is Sentinel-5P’s average annual satellite image data, which were gathered for four important pollutants: carbon monoxide (CO), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), and ozone (O₃) over a six year period from 2019 to 2024. The second is the MODIS aerosol optical density (AOD) product satellite image data from the GEE platform, which calculate the average annual particulate matter (PM_2.5 and PM₁₀). All mentioned pollutant images were used to calculate the air quality index (AQI) and aggregated air quality index (AAQI). Lastly, we used Landsat’s average yearly land surface temperature (LST) retrieval (OLI/TIRS). The aggregated air quality index (AAQI) was computed, and the U.S. Environmental Protection Agency’s (USEPA) air quality index (AQI) was created for each pollutant. According to the data, the AQI for CO, PM_2.5, and PM₁₀ in the research region ranged from hazardous to unhealthy; at the same time, the AQI for NO₂ varied between harmful and unhealthy for sensitive groups, with values ranging from 135 to 165. The annual average of the AQI for SO₂ throughout the studied period ranged from 29 to 339, with the categories ranging from good to hazardous. The constant AQI for ozone in the study area indicates that the ozone doses in Qena are surprisingly stable. Lastly, with a minimum value of 265 and a maximum of 489, the AAQI ranged from very unhealthy to dangerous in the current study. According to the data, the area being studied has poor air quality, which impacts the environment and public health. The results of this study have significant implications for environmental sustainability and human health and could be used in other areas.

1. Introduction

For the sake of sustainable agriculture and global food security, global efforts should be made to improve air quality [1]. The main environmental issues that many cities face are air pollution from growing populations, traffic, industry, and burning biomass in agriculture, as well as particulate matter from natural sources like dust and sand events and man-made fuel consumption in industry and transportation [2]. Furthermore, respiratory and cardiovascular disorders are among the major health issues that might result from excessive concentrations of several air contaminants [3]. Egypt is among the emerging nations that struggle with air pollution [2,4,5,6]. Particulate matter (PM_2.5, PM₁₀), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), ozone (O₃), methane (CH₄), formaldehyde (CH₂O), and carbon monoxide (CO) are among the most prevalent air pollutants [7]. The term “air quality management” describes all of the actions taken by a regulatory body to assist with shielding the environment and public health from the damaging impacts of air pollution. Setting goals for air quality, reducing emissions, creating control methods, and putting those plans into action are all parts of the process of controlling air quality. To determine if the air quality objectives are being fulfilled, it is critical to conduct ongoing evaluations (https://www.epa.gov/air-quality-management-process/air-quality-management-process-cycle) (accessed on 2 January 2025).

The use of satellite-based remote sensing has gained popularity recently as a potent instrument for monitoring and assessing air quality. Compared to conventional monitoring approaches, remote sensing techniques enable the collection of data on air pollution over both spatial and temporal patterns, thereby yielding a more thorough understanding of air quality [8]. For technical purposes, the Sentinel-5P TROPOMI tropospheric monitoring instrument is used to measure anthropogenic emissions and develop methods for reducing air pollution [9].

Atmospheric substances like NO₂, SO₂, CO, and CH₂O are of particular importance because they have historically and presently been released in significant quantities in urban and industrial regions [10]. Google created the free web-based computing platform known as Google Earth Engine (GEE) to make processing geographical information easier [11]. The algorithms, quick modeling, and visualization are supported by this cloud-based platform’s “application programming interface” (API) and associated “interactive development environment” (IDE) [12]. GEE provides a greater variety of satellite imaging types (such as Landsat series, MODIS (Moderate Resolution Imaging Spectroradiometer), Sentinel-1, -2, -3, and -5P, NOAA, Advanced Land Observing Satellite (ALOS), etc.) in addition to climate, weather, and geophysical information than its rivals [13]. In recent years, GEE has been widely used for a variety of remote sensing applications, including crop mapping [14,15,16,17], LULC extraction [18,19,20], and wetland delineation [21,22], which is a notable benefit of the spatiotemporal data over the traditional method [23]. GEE is also used for monitoring and evaluating the quality of the air [24,25,26,27]. The TROPOMI instrument is useful in numerous investigations since it was implemented with the introduction of S5P [28,29,30,31,32,33,34,35,36,37,38,39]. In one study, the TROPOMI data were used to monitor NO₂ levels over the Turkish region. In another study that focused on CO measurements, the first results of using TROPOMI in China were shown, demonstrating the instrument’s capabilities [40]. Global statistical comparisons between TROPOMI data and ground-based measurements showed a small percentage discrepancy between the two [41]. In a study conducted in Al Qurayyat City, Saudi Arabia, it was found that the measurements of the concentrations of CO and NO₂ by TROPOMI, mostly generated from automotive sources, showed a close connection, indicating a shared emission origin [42]. The TROPOMI satellite instrument, with its daily worldwide coverage and high resolution, was found to have successfully discovered a weekend effect in road transport-dominated CO emissions in an African study that focused on cities like Cairo and Algiers [43]. Significant relationships between surface urban heat island intensity (SUHII) and NO₂ in the spring and CO in the winter were revealed by another study conducted in Greater Cairo, Egypt, using the TROPOMI satellite instrument. This study emphasized the value of geospatial technology in assessing regional air pollution and guiding urban environmental management [2].

Many people’s desire to live in an urban region is increased by its possibilities [44]. According to United Nations reports from 2014 and 2018 [45,46], two-thirds of the people on Earth will reside in cities by 2050. As a result, there will be a great deal of urbanization worldwide, which will lead to urban sprawl and expansion [44,47,48,49], as well as encroachment on small agricultural and green spaces, which destroys green cover. Environmental consequences, including rising land surface temperature (LST), urban heat island (UHI), and air pollution, are brought on by this devastation and population increase. Fuladlu and Altan [44] concluded that the mutual effects of the LST, CO, and O₃ levels in Tehran province are represented by the positive correlations between CO (r = 0.641, p < 0.01) and O₃ (r = 0.744, p < 0.01) and LST. This suggests that rising LST may be a factor in the buildup of CO and O₃ values. In addition, PM_2.5 (r = 0.457, p < 0.01) suggests a favorable association with LST.

In Casablanca, Morocco, El Kamouri et al. [50] examined the temporal and spatial patterns of air pollutants, specifically nitrogen dioxide (NO₂), and found that NO₂ concentrations peaked in the winter. The years 2021, 2022, and 2023 saw the highest concentrations, indicating a noticeable rise. The NO₂ values did, however, significantly drop from 93 µmol/m² in April 2019 to 78 µmol/m² in April 2020 and from 109 µmol/m² in October 2019 to 95 µmol/m² in October 2020. The COVID-19 lockdowns, which reduced vehicle traffic, were blamed for the notable drop in NO₂ levels that was seen. Furthermore, the city center, which is known for having a high traffic density of both private and large freight vehicles, was regularly found to have the highest concentrations. This is mostly due to the substantial amount of freight transport activity that is concentrated in this area. The COVID-19 epidemic significantly reduced traffic density in 2020, but in 2023, it rapidly increased again. As demonstrated by the growth of the vehicle fleet, this recovery is a result of both increased economic activity and increased vehicular traffic. Although satellite remote sensing can lower spatial uncertainty, its measurement results are largely restricted to the upper atmosphere, where they are highly sensitive and are subject to regional, seasonal, and weather variations. Furthermore, the distinct features of every place, such as temperature, topography, and socioeconomics, make it challenging to develop a globally applicable model for assessing air pollution [51,52].

Air pollution has a variety of negative health effects. The health of those who are sensitive and fragile may be impacted even on days when air pollution is lower. The following symptoms are highly connected with short-term exposure to air pollution: asthma, respiratory illnesses, wheezing, cough, shortness of breath, and high rates of hospitalization (a measure of morbidity). The long-term effects of air pollution include cardiovascular diseases, cardiovascular mortality, pulmonary insufficiency, and chronic asthma [53]. Many types of power plants, factories, automobiles, and other means of transportation are known to be instances of anthropogenic sources. There is a strong correlation between air pollution and atherosclerotic cardiovascular diseases, such as myocardial infarction and stroke, and fuel-burning machinery is the primary source of anthropogenic NO_x emissions [54]. Stroke and cardiovascular diseases continue to be among the leading causes of morbidity and mortality worldwide, and short exposure to NO₂ and particulates (PM_2.5, PM₁₀) may increase the risk of heart attack and stroke, increase likelihood of hospitalization, and increase death rates. To prevent this from happening in the years to come, mitigation measures must be taken to reduce the concentration of indoor air pollutants in the investigated site (that factory). One way to do this is by using suction fans to increase air flow rates in the workshop to keep pollutants far from the standards.

The World Health Organization [55,56] reported that more than 43,000 Egyptians died in 2012 from diseases related to air pollution, including acute lower respiratory, chronic obstructive pulmonary disease, stroke, ischemic heart disease (IHD), and lung cancer. According to the Global Burden of Disease 2017 results, Egypt’s population is expected to reach approximately 200 million by 2100 [57]. This development will be complicated by ongoing climate change. The research states that 2079 Egyptians died of lung cancer and 22,327 Egyptians died of IHD. Based on data from 2016 to 2017, the World Bank [58] estimates that air pollution in the Greater Cairo Metropolitan Area (GCMA) alone cost EGP 47 billion in 2017, or 1.35% of Egypt’s GDP. Every year, the number of deaths rose by 34% from 9400 to 12,569, and the number of people exposed to PM_2.5 increased by 45% from 11.9 in 1999 to 17.3 in 2017. According to Apte et al. [59], air pollution reduces Egyptians’ life expectancy by an average of two years because of sickness or disability. In Egypt, air pollution caused 90,559 premature deaths in 2019 [60] and over 12% of all deaths in 2017 (Institute for Health Metrics and Evaluation 2019). The WHO database shows that, in 2016, heart disease (57.9%), stroke (17.7%), pulmonary and lower respiratory diseases, and cancer (24.4%) were among the illnesses linked to air pollution that caused premature mortality in Egypt [55]. The leading cause of death in Egypt is noncommunicable diseases, which account for 82% of deaths and 67% of premature deaths [61].

The dangers of air pollutants to humans vary depending on their types and concentrations, and because they have different determinants and units, it is challenging for the general public to predict the risk of air pollution based on concentrations beforehand. As a result, a technique is required to rate the health risks associated with various contaminants as a single number [62]. Furthermore, policymakers have suggested that air quality be shown in indices to inform and warn the public about the dangers of pollution [63]. A model called the Air Quality Index (AQI) is used to convert air pollutant concentrations into a single value that indicates the state of the air quality. In the literature, there are numerous air pollution indices with various names, such as the Pollution Standard Index (PSI), Air Quality Index (AQI), and Air Pollution Index (API) [64,65]. Based on the principle of their calculating method, there are two types of AQIs. The first kind uses the highest sub-index as the AQI. With this kind of AQI, a lot of information is eliminated and public awareness is raised [66]. On the other hand, the second kind of AQI, known as aggregated AQI, aims to gather the effects of several air contaminants. It is possible to use arithmetic to aggregate the pollutants’ indices [67,68]. Thus, the primary goal of this study is to use Google Earth Engine (GEE) and the Aggregated Air Quality Index (AAQI) to track the spatiotemporal analysis of air quality over the Qena Governorate, Egypt, throughout six consecutive years: 2019, 2020, 2021, 2022, 2023, and 2024. The current investigation is considered a pioneering study and the first attempt to map the air quality index in the studied area. This offers a more thorough and precise view of the air quality in the research region, which may contribute to better environmental and public health results.

2. Materials and Methods

2.1. Description of the Research Area

The third governorate from the southern border of Egypt, following Aswan and Luxor, is Qena, which is situated between latitudes 25°13′ to 26°15′ N and longitudes 32°00′ to 32°55′ E. Figure 1 illustrates its borders: Sohag Governorate, the New Valley Governorate to the west, the Red Sea Governorate to the east, and Luxor Governorate, which broke away from Qena in 2010, to the south. Of the approximately 9565 km² that make up the Qena Governorate, 1348.32 km² (13.50%) is located in the Nile Valley (http://qena.gov.eg/Default.aspx, accessed on 6 January 2025). As the primary geographical feature of Qena Governorate, the Nile River flows 162 km across the valley from south to north, dividing the governorate’s territory to the east and west (http://qena.gov.eg/Default.aspx, accessed on 6 January 2025). Administratively, the Qena Governorate is separated into nine local units, which are Abu Tesht, Farshout, Nag Hammadi, Deshna, Alwakf, Qena, Keft, Quos, and Nagada, as well as one new urban community and 112 continuing villages. A total of 3,224,573 people were living in the Qena Governorate, with 80.3% of them residing in rural regions and 19.7% in urban areas. The Qena Governorate is one of the most densely populated governorates. The Qena Governorate is among the governorates that have a primarily agricultural economy. Sugar cane and bananas are grown in the province along with wheat, corn, vegetables (tomatoes), palm trees, lucerne, sesame, and a few fragrant and medicinal plants. The primary crop is sugar cane; the second-most important winter crop is wheat; and the second-most important summer crop is maize (https://www.eeaa.gov.eg/Uploads/Reports/Files/20221128214734177.pdf, accessed on 6 January 2025).

The research area’s climate is classified as dry, which is typical of North Africa, and is marked by scorching summers, mild winters, and little to no precipitation. In general, the relative humidity ranges from 16% to 41%, and the average temperature falls between 16.6 °C to 32.7 °C (https://www.weather-atlas.com/en/egypt/qena-climate, accessed on 6 January 2025). According to Abdelsamie et al. [20], the study region primarily consists of the following four kinds of distinct land-uses, as shown in

Table 1. Description of different land-uses of the study area.

LULC Classes	Area
	km2	%
Urban areas	281.40	23.11
Water bodies	88.60	7.27
Cultivated soils	800.40	65.72
Bare soils	47.46	3.90

and Figure 2: cultivated soils (65.72%), bare soils (3.90%), water bodies (7.27%), and urban areas (23.1%).

2.2. Data Collection

Three different kinds of data sets were used in the current investigation, as follows:

(a)

The Sentinel-5P satellite image data for the four main pollutants—SO₂, NO₂, O₃, and CO products—were obtained annually from the GEE analysis. Since latitude and longitude define the pixels in the Sentinel-5P Level-2 output, the resulting uneven grid made it more challenging to combine several photos. As a result, Level-3 processing was used. This can be resolved by using the harpconvert tool’s bin spatial operation to resample the data to a normal spatial pixel grid (Harpconvert. Available online: http://stcorp.github.io/harp/doc/html/harpconvert.html (accessed on 13 February 2021)). For every variable in a product, the HARP bin_spatial tool projects all time samples into a standard spatial lat/lon grid. As a result, Sentinel-5 provided the CO, NO₂, SO₂, and O₃ photos. After that, cloud-containing images were filtered using cloud filters, followed by spatial and temporal filters, and average maps were extracted using average filters for the years under study (Figure S1). The contaminants listed above were used to calculate the air quality index (AQI) and aggregated air quality index (AAQI). Initial information on the concentration of different contaminants in the atmosphere was derived from datasets acquired by the European Space Agency from the Copernicus Sentinel-5 Precursor satellite project. The high-resolution spectrometer system used by the Sentinel-5 mission uses seven different spectral bands: UV-1 (270–300 nm), UV-2 (300–370 nm), VIS (370–500 nm), NIR-1 (685–710 nm), NIR-2 (745–773 nm), SWIR-1 (1590–1675 nm), and SWIR-3 (2305–2385 nm). The mission operates in the ultraviolet to shortwave infrared range. The MetOp-SG A satellite hosted the device. The yearly average concentrations of SO₂, NO₂, O₃, and CO were calculated using the Google Earth Engine (GEE) to streamline the process of acquiring data from the Copernicus Sentinel-5 Precursor satellite (simplification of data processing of net CDF files). GEE offers a robust and adaptable environment for working with a variety of remote-sensing, satellite, and other geospatial datasets, including Sentinel-5 Precursor data. It consists of a cloud-based computing platform for analyzing and processing large-scale geospatial data. The capacity of GEE to effectively process and analyze vast volumes of data without the need for pricey computer hardware or software is one of its main benefits when used for geospatial analysis. Additionally, GEE provides a collaborative platform for exchanging code, data, and analysis findings. The concentration of pollutants was estimated using the “Sentinel-5P L3” collection (e.g., COPERNICUS/S5P/OFFL/L3_NO₂ for nitrogen dioxide). Several filtering techniques (to achieve average annual and monthly values), data processing techniques (such as cropping within the limits of the study area), and additional analyses of the raster values of pollutants were employed for the collection.

(b)

The MODIS aerosol optical density (AOD) product’s annual satellite image data were gathered from the GEE platform. The MODIS sensors on NASA’s Aqua and Terra satellites use 36 spectral bands across a broad spectral range (from 410 to 1500 nm) to provide near-global daily readings [69]. In December 1999, MODIS was introduced on the Aqua platform, and in late May 2002, it was introduced on the Terra platform [70]. About 42 satellites with AOD recovery bands have been launched worldwide so far. MODIS AOD products are now among the most widely utilized and well-liked AOD data sources by academics because of their high retrieval accuracy, ease of use, and availability [71]. The Moderate Resolution Imaging Spectroradiometers (MODIS) on the NASA satellites Terra and Aqua provide the most widely used AOD datasets in air quality research. The acquired data demonstrated its applicability for evaluating air quality and providing further information on the spatial distribution of aerosols [72]. To determine the PM_2.5 concentration level, an increasing number of studies have employed diverse AOD data as well as alternative models and algorithms in recent years, with positive outcomes [73,74,75]. However, most AOD products used in current studies are mainly based on existing mature AOD products, such as MODIS [73,76]. There are certain drawbacks to using MODIS data to calculate PM, such as the fact that MODIS has a limited lifespan and that its anticipated operating term has already passed [77]. Additionally, efforts are being made to alleviate the constraint of recording both the maximal and minimal PM_2.5 peaks. To enhance model performance, more attention should be paid to ground data, atmospheric conditions, and physical features of the regression model [13].

The following formulas were used to calculate PM_2.5 and PM₁₀ using the AOD data [78,79]:

$${\mathrm{P}\mathrm{M}}_{2.5}=165.69\mathrm{A}\mathrm{O}\mathrm{D}+8.7597$$

(1)

$${\mathrm{P}\mathrm{M}}_{10}=195.7\mathrm{A}\mathrm{O}\mathrm{D}+14.5$$

(2)

(c)

Each year, Landsat (OLI/TIRS) sensors, which record temperature data and store them as digital numbers (DNs) within a range from 0 to 255, provide the land surface temperature (LST) [80]. Each image that was used was acquired during the study years. To convert DNs to Kelvin degrees, the following steps are performed:

(1)

Transforming DN into spectral radiance: In this study, DNs in each band of the image were transformed into physical measurements of sensor radiance Lλ, as shown in the following Equations (3) and (4):

$$L\lambda =ML\times QCal+AL-Qi$$

(3)

where:

Lλ = top of atmosphere (TOA) radiance (W·sr^−··m^−m·nm⁻¹);

ML = radiance multiplicative for band 10;

Qcal = quantized and calibrated standard product pixel values (DN);

AL = radiance add band 10;

Qi = the correction for band 10.

(2)

Radiance value conversion to brightness temperature: The following formula was used for the conversion of spectral radiance to at-sensor brightness temperature:

$$T_B={\displaystyle \frac{K_2}{In(\left(\frac{K_1}{L\lambda }\right)+1)}}-273.15$$

(4)

where:

$T_B=$ top of atmosphere brightness temperature (°C);

Lλ = top of atmosphere (TOA) radiance (W·sr⁻¹·m⁻²·nm⁻¹);

K₁ and K₂ are the constants for Landsat (OLI) band 10;

K₁ is = 774. 8853;

K₂ is = 1321. 0789.

2.3. AQI Calculations

AQI was found by the following Equation (5), which is recommended by the study in [81]:

$$I_p={\displaystyle \frac{I_{Hi}-I_{Lo}}{{BP}_{Hi}-{BP}_{Lo}}}\left(C_p-{BP}_{Lo}\right)+I_{Lo}$$

(5)

where:

I_p—the index of pollutant p;

I_Hi—the AQI value corresponding to BP_Hi;

I_Lo—the AQI value corresponding to BP_Lo;

BP_Hi—the threshold that is greater than or equal to C_p;

BP_Lo—the threshold that is less than or equal to C_p;

C_p—the rounded concentration of pollutant p.

The site’s AQI was determined by the value of the highest sub-indices.

Table 2. AQI categories and corresponding concentrations for major pollutants.

AQI	PM2.5 (ppb)	PM10 (ppb)	CO (ppm)	O3 (ppm)	SO2 (ppb)	NO2 (ppb)	AQI Category
0–50	0–9	0–54	0–4.4	0–0.054	0–35	0–53	Good
51–100	9.1–35.4	55–154	4.5–9.4	0.055–0.070	36–75	54–100	Moderate
101–150	35.5–55.4	155–254	9.5–124	0.071–0.085	76–185	101–360	Unhealthy for sensitive group
151–200	55.5–125.4	255–354	12.5–15.4	0.086–0.105	186–304	361–649	Unhealthy
201–300	125.5–225.4	355–424	15.5–30.4	0.106–0.200	305–604	650–1249	Very unhealthy
301–500	225.5–326.4	425–604	30.5–50.4	0.201–0.604	605–1004	1250–2049	Hazardous

displays the principal contaminants’ AQI categories and corresponding concentrations, as explained by the following equation [81].

$$AQI=Max(I_{CO},I_{NO2},I_{O3},I_{SO2},I_{PM2.5},I_{PM10})$$

(6)

The following Equation (7) was used to construct the aggregated air quality index (AAQI) [68,82]:

$$AAQI={\left\{{\sum }_{i=1}^{i=n}{\left({AQI}_i\right)}^p\right\}}^{{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$P$}\right.}}$$

(7)

where: AAQI—aggregated air quality index;

AQI_i—sub-index for single pollutant i.

$p$ is a parameter in the range from 1 to ∞;$p$ = 1. In earlier research [68,82,83], $p$ ranged between 2 and 3. Accordingly, in this investigation, $p$ was set at 2.5, consistent with [82,84]. The flowchart of the approach framework utilized in the current study is shown in Figure 3.

3. Results and Discussion

3.1. Mean Annual LST

LST, gaseous pollutants, and particle matter make up the complex and dynamic composition of urban air pollution, which varies daily and seasonally as a result of human activity, changing land cover, and weather patterns. It is now widely recognized that land cover and urban biophysical and thermal conditions are related [44]. Although both are related to urban land-cover, it is unclear how air pollutants, LST, and, consequently, urban heat islands (UHIs) are related. UHIs are not a sign of air pollution, even if they contribute to its growth. Urbanization has greatly increased human comfort and quality of life, but it has also brought up some unforeseen issues. Among these are urban heat islands (UHI) and urban pollution islands (UPI). Mass car use has only recently led to UPI. It is a result of sprawl or suburban life becoming more commonplace. Over the past century, there have been significant advancements in transportation technologies, leading to the expansion of compact cities into low-density suburbs that rely heavily on cars. The period of widespread motor automobiles can be blamed for many of the air pollution concerns of contemporary cities. With the exception of huge desert towns, UHI is a phenomenon unique to large cities, where the temperature in the urban core is seen to be greater than that of the surrounding rural or natural areas. Because of the effects of climate change, urban heat—dubbed an invisible hazard—has the largest potential for human mortality in several nations [85]. Around the world, heat waves caused 171,856 deaths between 1900 and 2021 [86]. In many cities throughout the world, urban warming is seen as a major health risk [87]. Furthermore, the results in [88] demonstrated the growing interest in UHI–UPI interactions in research, which is indicative of the threats they pose to the ecosystem as well as to human health and life in the modern world. Higher metropolitan temperatures also translate into higher energy consumption, primarily because of the increased need for air conditioning. Pollution levels rise when power plants burn more fossil fuels [89]. Air pollutants raise the LST and are commonly used as markers of air pollution in ambient air. The urban heat island (UHI) effect and air pollution have become pervasive environmental issues in urban areas, and the process of urbanization adds to localized urban climate changes, particularly LST and air quality. Findings of Zhang et al. [90] showed a steady decrease in air pollution and surface urban heat island (SUHI) levels, with a noticeable rise in the intensity of SUHI during the day on summer days. An analysis of interactions showed that SUHI significantly degraded the quality of the air during the day. Furthermore, LST data are frequently used for comparative analyses of urban thermal environments that are linked to residential activities, industrial production, and transportation processes. These activities typically occur in urban areas and release significant amounts of air pollutants into the atmosphere, including CO, SO₂, NO₂, PM_2.5, and PM₁₀ [91].

Anthropogenic and natural sources include, for instance, the combustion of fossil fuels, mediated biomass burning, emissions from industry and traffic, and lightning [92,93]. Additionally, motor vehicle traffic in metropolitan areas emits CO, and lowering traffic speeds may result in an increase in CO emissions. Lower speeds are thought to have the potential to produce large emissions because of things like idling, frequent gear changes, and more stop-and-go traffic, which involves accelerating and decelerating frequently and typically uses a lot of gasoline [94,95]. Particularly in the current study, which is situated in one of the poorest rural areas in Upper Egypt, where nearly 43 percent of the population lived below the poverty line, rural areas in Lower Egypt came in second with 23 percent, and urban areas in Lower Egypt had the lowest percentage of the population living below the poverty line, with nearly 4 percent (https://www.statista.com/statistics/1237070/poverty-rate-national-line-in-egypt-regions/, accessed on 5 March 2025). Therefore, by combining oxygen, CO, and unburned hydrocarbons to form carbon dioxide and water, the catalytic converter helps to reduce the emission of CO [96]. LST and urban carbon emissions (UCE) may have a positive statistical relationship, according to Hong et al. [97]. In another study conducted by Song et al. [98], there was a positive correlation between LST and total carbon emissions of 0.3–0.7. As more and more researchers focus on the effects of carbon emissions and understand the mechanism of reaction between carbon emissions and LST, they discover a positive correlation between the two [99,100]. Oderinde [101] validated that, with the exception of the freshwater swamp forest and the mangrove swamp forest, where an inverse connection was noted, a rise in CO₂ emissions resulted in an increase in surface temperature. We found that both LST and CO₂ emission rates were higher in areas with less vegetative cover. Because of the high rates of vegetation loss for urbanization and other economic activity, carbon dioxide emissions in these areas have increased. Due to its detrimental effects on several economic sectors, the country is in grave jeopardy as a result of the negative climate changes brought on by the rise in CO₂ and LST. These days, it is widely recognized how urban biophysical and thermal conditions relate to land cover. The LST of the analyzed time was generated over the study area in the current study (Figure 4). According to the data, the LST has increased over the past few years and is higher in urban and industrial areas than in other locations (Figure S1). This demonstrates how human activity affected these regions’ LST.

Urbanization, industrialization, infrastructure expansion, and an increase in human activity all contribute to the loss of green space, which has an adverse effect on our environment by increasing air pollution and LST [102]. Fuladlu and Altan [44] evaluated the association between LST and the major air pollutants (MAP) in the Iranian province of Tehran using Sentinel imagery. They discovered that the central urban area of the city had the greatest mean concentrations of PM, SO₂, and NO₂ and that the LST has a positive connection with MAP. Ngarambe et al. [103] evaluated the relationship of urban heat island intensity (UHII) with air pollutants in Seoul city and discovered that NO₂, SO₂, CO, and PM showed a positive correlation with UHII. There was a noticeable seasonal variation in the relationship between UHII and air pollutants. Prior research has demonstrated that urbanization increases air pollution and LST in a number of urban regions [104,105]. Ulpiani [106] presented a review of the research from the last 30 years on the relationship between LST and air pollution and recommended that, going forward, it would be crucial to examine these parameters jointly rather than individually. Lai and Cheng [107] found that, in the Taiwan Strait, China, surface urban heat island (SUHI) had a negative correlation with O₃ and a positive correlation with key pollutants, such as PM, NO₂, SO₂, and CO. Mishra and Mathew [108], using MODIS images, calculated the urban pollution island intensity (UPII) over Delhi city from the aerosol optical depth (AOD) image and compared it with the SUHI intensity (SUHII). There was no discernible relationship between the two; rather, the change in land cover had a greater impact on the variation of LST. Hereher et al. [2] determined a stronger association of SUHII with CO and NO₂; although, it showed weak correlation with SO₂ over Greater Cairo, Egypt. It has been established that a wide range of factors affect both LST and air pollutants, including land use/cover, population, elevation, weather, and economic development [109]. Geographical locations have a considerable impact on the association between LST and air pollution, and the existing research is insufficient to fully explain how LST and other air pollutants are related. A city’s air pollution concentration is also influenced by the amount of vegetation and built-up area. It is also crucial to investigate the relationship between air contaminants and land cover indices. Few studies have examined the association between LST and air pollutants at specific sites, but there is a dearth of research on the relationship between air pollutants and land cover features; thus, it needs more investigation. High temperatures, low relative humidity, low wind speeds, and cloudlessness are synoptic circumstances that favor the formation of UHI and severe pollution episodes simultaneously [107]. Particularly at night, urban breezes with low-level inversion and low-level convergence exacerbate local hyperthermia and air [106]. Elevated temperatures speed up some cycles of atmospheric chemistry, producing more ground-level (tropospheric) ozone, more biogenic hydrocarbons (BVOCs), and more synthetic VOCs evaporating from car engines [110]. In regions with extreme heat, air conditioners run for longer hours both during the day and throughout the year. This causes regional power plants to emit more CO₂ and ozone precursors [111] and contributes to the amplification of UHI in ever-increasing spirals by releasing exhaust heat into the urban air. Additionally, peak power absorptions aided by UHI (such as during heat waves) require more generation capacity, which is usually provided by less sustainable power plants. Accordingly, UHIs not only help to create a dust dome above the cityscape but also help to maintain themself by combining a recurrent pattern that intensifies and prolongs pollution episodes. Conversely, warmer air encourages turbulent mixing, which facilitates the dispersal of primary pollutants to higher atmospheric layers (Sarrat [112]).

3.2. Mean Annual Pollutants Concentrations

The mean annual concentrations of each studied air pollutants are analyzed and presented hereunder:

3.2.1. AQI for Particulate Matter (PM_2.5 and PM₁₀)

The complex mixture of particles known as particulate matter (PM) can be solid, liquid, or both. They differ in size, composition, and place of origin. Region, season, day, weather, and other variables all affect the precise makeup and size distribution of PM [113]. Total suspended particulates (TSP), PM₁₀ (particles with a diameter less than 10 μm, seen near roads and dusty industries), PM_2.5 (particles with a diameter less than 2.5 μm, seen in smoke and haze), and ultrafine particulates (UFPs) (particles with a diameter less than 0.1 μm, or PM_0.1) are the four categories into which PM can be separated [113]. The AOD-PM_2.5/PM₁₀ empirical–statistical models generated by [78,79] were used in the current investigation to estimate the concentrations of PM_2.5 and PM₁₀. The yearly average PM_2.5 AQI ranged from a low of 283 to a maximum of 517 between 2019 and 2024 (Figure 5). Based on this, the area is classified as very unhealthy to hazardous (

Table 3. The minimum and maximum annual AQI of major air pollutants over the study area.

Years	PM2.5	PM10	CO	O3	SO2	NO2
2019	283–449	422–702	286–326	324–325	51–303	135–149
2020	320–505	462–774	909–956	331–333	107–339	134–149
2021	298–469	425–713	918–967	304–304	85–339	136–153
2022	320–517	463–794	858–901	304–304	29–297	133–152
2023	323–479	468–729	884–936	331–333	63–281	140–157
2024	340–503	492–766	928–983	348–349	66–295	147–165

). In contrast, the PM₁₀ AQI ranged from 422 to 794, indicating the risks associated with this pollutant (Figure 6). The primary sources of PM include burning solid waste and agricultural waste, dust and sand storms, and emissions from vehicles and industries [114]. Three sugar factories, one spinning and weaving mill, and the Aluminium Complex—one of the largest industrial facilities in the Middle East—are all located inside the governorate. Two industrial zones are also located inside the governorate; the first is located at Qaft Markaz, a group of villages in Kalaheen. The nonwoven fabric plant, the antibiotic factory, the food projects, and the plastics industries are also among the most significant undertakings. Investors are encouraged to launch their projects in the second industrial zone, Yahaw in Naggâ Hamady, which has the necessary infrastructure.

3.2.2. AQI for Carbon Monoxide (CO)

In addition to traffic emissions, industrial output, and biomass burning, the incomplete combustion of substances including petrol, natural gas, oil, coal, and wood can contribute to the problem of CO emissions into the atmosphere [115]. The annual AQI for CO ranged from 286 to 983 ppm over the study period. The AQI increased sharply between 2019 and 2020 but then continued to rise progressively from 2020 to 2024 (Figure 7 and Table 2). Locality and population density are typically associated with high CO concentrations [116]. Furthermore, anthropogenic activities, car emissions, and the locations of numerous power plants and enterprises may all contribute to the high CO concentration [117]. The WHO states that the CO AQI might be classified as having very unhealthy or hazardous effects [118].

3.2.3. AQI for Nitrogen Dioxide (NO₂)

A major health danger in urban and industrial regions, NO₂ is a hazardous air pollutant that causes severe health effects and premature death. It is the cause of tens of deaths in Cairo each year. Figure 8 shows that the AQI for NO₂ fluctuated between unhealthy for sensitive populations and unhealthy, with values ranging from 135 to 165. Fuel combustion is the main source of NO₂ in the atmosphere. Emissions from power plants, automobiles, trucks, buses, and off-road machinery produce NO₂ (https://www.epa.gov/no2-pollution/basic-information-about-no2, accessed on 7 March 2025); this finding is consistent with the findings of Hereher et al. [2], which explains why it is more prevalent in places with a high population density. Numerous emission sources, including natural and vehicle emissions, are connected to the amount of NO₂ in the atmosphere.

3.2.4. AQI for Sulfur Dioxide (SO₂)

Concentrations of SO₂ gas in the troposphere of the research area are mostly linked to population density and industrial and power generation facilities. This trend is especially evident in areas with sizable textile, smelting, chemical, petroleum, and power generation facilities. Since high sulfur heavy fuel is mostly used in these industrial processes, SO₂ emissions are produced, making the area a major hotspot for SO₂ emissions [26]. Even while the majority of the SO₂ concentration in the study area is caused by emissions from particular industrial activities, it is interesting to see that some residential areas also have relatively high SO₂ levels. The significant use of diesel-powered trucks for building new asphalt roads and performing construction tasks is the main cause of these abnormally high SO₂ concentrations in the new urban neighborhoods [26]. As seen in Figure 9, the AQI categories varied from good to hazardous, with the annual average of the AQI for SO₂ over the examined period ranging from 29 to 339.

3.2.5. AQI for Ozone (O₃)

Tropospheric ozone is a dangerous air contaminant and a greenhouse gas. Unlike other gases and particles released into the atmosphere, ozone is produced when sunlight interacts with a variety of other air contaminants. Methane, a strong greenhouse gas, as well as nitrogen oxides, non-methane volatile organic compounds, and carbon monoxide—all dangerous air pollutants—are the primary causes of the high concentrations of this poisonous gas. These “precursors” come from a variety of sources, such as industry, transportation, energy production, livestock, and wildfires. People and the earth are harmed by high tropospheric ozone levels. The greenhouse effect, which contributes to both global warming and regional climatic consequences, is driven by tropospheric ozone. Inhaling it at ground level causes respiratory and other health problems. It interferes with the growth of staple crops like maize and wheat when it is spread throughout farms. On the other hand, the stratosphere contains the greatest concentration of ozone molecules, with over 90% of them located between 15 and 30 km above the ground. Life on Earth is protected by stratospheric ozone, which blocks off a significant portion of the sun’s ultraviolet (UV) radiation. Increased UV-B radiation (280–320 nm) can impair photosynthesis, lead to skin cancer, and impair immunity [119].

Two main pollutants—NO_x (NO and NO₂) and hydrocarbons from industry, traffic, vegetation, and biomass burning—photochemically react to produce ozone [120,121,122,123,124,125]. During the night, NO is released by nearby sources, and ground deposition destroys the ozone. Due to photochemical production from anthropogenic emissions throughout the day, urban ozone diurnal fluctuation is often characterized by higher concentrations during the day and lower concentrations at night. It has also been acknowledged that “next-day” surface ozone variations may be influenced by ozone stored in nighttime residual layers (RL) [126]. Because they influence boundary layer stability, carry pollutants at the regional level, and have the potential to alter urban air quality, nocturnal low-level jets (NLLJs) are a significant phenomenon at night [127]. The ozone doses in Qena were unexpectedly stable (Table 2 and Figure 10) during the analyzed time, as evidenced by the Ozone’s AQI that varied between 304 and 349, but in 2021 and 2022, it remained steady, with very minor variations in 2019, 2020, 2023, and 2024. The O₃ concentrations in the research region fell into the very unhealthy (201–300) to hazardous category (301–500) according to the EPA [81] standards. This could indicate that, during the brief period of heavy photochemical smog, there are more health risks and crop damage in and around greater Qena. Ozone does not reach exceptionally high concentrations in the ambient urban air of Qena as may be expected, despite the city’s high main pollutant emission rates and heavy sun irradiation.

3.3. Aggregated Air Quality Index (AAQI)

Based on how they are calculated, AQIs can be divided into two categories. The first type computes the sub-indices for all pollutants and then uses the highest sub-index as the AQI based on local air pollution regulations or USEPA requirements. With this kind of AQI, a lot of information is eliminated while public awareness is raised [66]. The second kind of AQI combines several air contaminants in an effort to measure their combined effects. The AAQI is a more comprehensive indicator of air quality and is more practical for environmental management because it includes a wide range of air quality contaminants [84]. If more than one pollutant is above the critical concentration, it might be more effective than the USEPA-AQI. According to the findings of epidemiological studies on the impact of each pollutant on human health, the Environmental Protection Agency (EPA) [81] established breakpoint concentrations. Each pollutant receives a value on a scale from 0 to 500 following the standardizing procedure (i.e., 0–50 (good), 51–100 (moderate), 101–150 (unhealthy for sensitive groups), 151–200 (unhealthy), 201–300 (very unhealthy), 301–400 (hazardous), and 401–500 (hazardous)). According to the CAA (Clean Air Act), the alert level for human health on any particular day is a value more than 100. The AAQI varied from very unhealthy to hazardous in the current investigation, with a range of 265–489 from 2019 to 2024 (

Table 4. AAQI values in the studied area.

Years	AAQI	Category
2019	265–350	very unhealthy to hazardous
2020	388–477	hazardous
2021	377–475	hazardous
2022	369–470	hazardous
2023	378–475	hazardous
2024	397–489	hazardous

and Figure 11). Table 4 and Figure 11 clearly show that the measured pollutants were above the EPA [81] limits; therefore, it might make sense to utilize the AAQI. However, even if some of the pollutants had good levels, it might not be acceptable for them to contribute to the air quality index’s rise. Many air pollutants associated with industrial production, residential activities, and transport processes are usually located in urban areas that emit large amounts of air pollutants, such as fine particulate matter (PM_2.5, particulate matter < 2.5 µm), coarse particulate matter (PM₁₀), SO₂, NO₂, and CO into the atmosphere. Consequently, there is a significant relationship between AAQI and air pollutants.

Given the abundance of industries in the research area, including those producing pepper, sugar, gas, and aluminum, industrial activity may be the reason for the noticeable rise in the AAQI (Figure 12 and Figure 13). Furthermore, it is anticipated that these measures have a major impact on the quantity of air pollutants from vehicles and industries given the closure of industries and limitations on community activities, such as intercity transportation and urban traffic sites, because of vehicle emissions [128]. Fossil fuel combustion produces nitrogen oxides, which are a reliable indicator of air pollution from vehicles. It has been demonstrated that transportation accounts for over 50% and 23% of the total nitrogen oxides (NO_x) in both industrialized and developing nations [129]. Tropospheric ozone, which is a major health risk because it can cause inflammation and increase airway hyperreactivity, is created when this pollutant combines with volatile organic chemicals and the sun’s UV rays [130].

4. Limitations and Uncertainties of the Current Study

Since remotely sensed data offer worldwide coverage with acceptable spatial resolution, they can be used to supplement data from chemical transport models and ground measurements from air quality stations. This allows for the monitoring of areas where no observations are available [131,132]. However, there are certain drawbacks to remotely sensed data. While monitoring stations offer hourly data, the temporal resolution is lower than ground measurements, with the best resolution possible being daily for measurements across the entire globe. Additional limitations include the fact that remote sensing does not directly measure pollutant concentrations because it measures the integrated concentration of the pollutant from ground level to the top of the atmosphere, which must be converted using ground data to determine actual ground concentrations [131,133]. Data may also not be available for all satellite passages due to the significant impact of clouds on the quality of satellite retrievals [14]. In order to improve outcomes, machine learning techniques like neural networks and the random forest model are suggested for further study. Weather, land cover, and temporal data should also be integrated [134].

5. Conclusions

One of the main environmental stressors that has detrimental effects on ecosystem balance and human health is air pollution. To provide relevant data regarding air quality across wide areas, remote sensing imagery has recently been used as a more economical and time-efficient alternative to traditional monitoring approaches. This study is considered an investigation since it is the first attempt to evaluate the air quality index using GEE and aggregated air quality index for tracking and charting the spatiotemporal air quality in the studied area. In this context, the Sentinel-5P satellite provided high-resolution images of atmospheric pollutants including NO₂, O₃, CO, and SO₂. Additionally, both PM_2.5 and PM₁₀ were calculated using MODIS data. Lastly, Landsat (OLI/TIRS) was used to retrieve the LST. LST data are frequently used to compare urban thermal environments that are linked to residential activities, industrial production, and transportation processes. These activities typically occur in urban areas and release significant amounts of air pollutants into the atmosphere, including CO, SO₂, NO₂, PM_2.5, particulate matter < 2.5 µm, and PM₁₀. Because industry, urbanization, and traffic are the main sources of air pollution in the research region, the data obtained demonstrated significant variance in the concentrations of all pollutants under study, except NO₂, SO₂, CO, and O₃, which had steady concentrations in 2021 and 2022 and only minor variations in 2019 2020 2023, and 2024. Additionally, the findings showed that transportation and industry are the main causes of the high air pollution emissions in the studied area. Furthermore, frequent problems with air pollution are brought on by the unchecked urban growth of the urban centers. The findings demonstrated how well satellites detected air contaminants in the study area because of their nearly constant worldwide coverage. Furthermore, the superior geographical dispersion of the aerosols provided by the satellite data is thought to be challenging to measure from ground measurements, particularly during storm occurrences. Additionally, there are not enough ground measurements because a comprehensive ground monitoring network is not practical for evaluating and tracking surface air quality. Ad hoc monitoring stations are placed in regions that are subjectively thought to be of interest to land use and population or in traffic hot spots. Ad hoc monitoring station placement, however, could result in poorly informed choices for long-term exposure studies. Therefore, regional patterns can benefit from the interpolation of pollution data gathered by regulatory air quality monitoring stations; yet, city-level data collection is made difficult by the sparse arrangement of air quality monitoring networks. By using models that assist in accounting for the fluctuation of air pollution and land use characteristics within cities, the locations of these sites could be improved. Satellite data can be used as a stand-in to measure the surface level of air quality, even though measurements of air contaminants using this data have some degree of inaccuracy. Understanding the uncertainties related to satellite data is so essential. We strongly advise users to focus on the density of data for a particular region and the quality flags. This is a crucial topic that health authorities can utilize to preserve human health and enhance environmental sustainability.