Assessing air quality variations in relation to vegetation patterns and meteorological conditions using remote sensing

Rapid industrialization, urban expansion, and the growth of transportation networks have significantly increased air pollutant emissions, especially in urban and peri-urban areas of developing nations. This study investigates the interplay among air quality, vegetation condition, and meteorological variables across two contrasting environments: the highly urbanized area of Sandton and the semi-urban setting of Tembisa. By comparing these locations, the research assesses how varying degrees of urbanization shape the relationships between atmospheric pollutants, ecological health, and local climate dynamics. During summer (DJF), Sandton recorded relatively low NO₂ concentrations (5–7 μg/m³) alongside high Modified Soil Adjusted Vegetation Index (MSAVI) values (0.4–0.67), reflecting extensive, healthy vegetation maintained through urban greenery such as trees, irrigated gardens, and landscaped areas. Vegetation remained comparatively stable year-round, with high summer greenness and only moderate winter decline, indicating a more persistent vegetative cover than in Tembisa. Tembisa, by contrast, showed fragmented and strongly seasonal vegetation, with reduced greenness outside summer and a pronounced decline in winter. In the winter months (JJA), environmental variables in Tembisa exhibited strong interrelationships: NO₂ and land surface temperature (LST) were highly positively correlated (r = 0.98), while Modified Soil Adjusted Vegetation Index and precipitation showed a strong negative correlation (r = −0.89), suggesting that winter rainfall may negatively affect vegetation through processes such as waterlogging, erosion, or runoff. In Sandton, however, only weak winter correlations were detected among these variables, indicating that precipitation has minimal influence on vegetation, temperature, and SO₂ levels within this urban environment.

1 Introduction

Air pollution is critical environmental health risk globally, linked to more than 6.5 million deaths each year and responsible for increasing rates of respiratory and cardiovascular diseases (World Health Organization, 2016a; Rafaj et al., 2018). Rapid industrialization, urbanization, and the expansion of transportation networks have contributed to a substantial increase in air pollutant emissions, particularly in urban and peri-urban areas of developing countries. Amounts of pollutants are also produced by a variety of human activities such as coal dumps, petrochemical operations, mining, agriculture, and electricity generation (Lourens et al., 2011; Pandey et al., 2018; Altieri and Keen, 2019; Kim, 2019). Among the most harmful air pollutants are sulphur dioxide (SO₂) and nitrogen dioxide (NO₂), which are primarily produced through fossil fuel combustion in industrial processes, coal-fired power generation, vehicular traffic, and biomass burning (Lourens et al., 2011; Pandey et al., 2018; Altieri and Keen, 2019). These gases not only pose direct health threats but also contribute to the formation of secondary pollutants such as particulate matter (PM_2.5) and tropospheric ozone, further exacerbating environmental and public health impacts (Dominick et al., 2012; He et al., 2024; Ren and Xie, 2022). Furthermore, although SO₂ and NO₂ are not direct greenhouse gases, they contribute to global warming through secondary chemical reactions that generate potent GHGs (Shikwambana et al., 2020), thus necessitating the need for integrated air quality monitoring and management approaches. The development of global frameworks such as the United Nations Sustainable Development Goals (SDGs) (UN, 2015; Amegah and Agyei-Mensah, 2017) have increasingly emphasised the need of addressing air quality challenges through specific targets and indicators. In particular, Goal 3 which promotes Good Health and Wellbeing, includes Target 3.9 and indicator 3.9.1 which aim to reduce the number of mortality rate and illness from hazardous air pollution. Likewise, Goal 11 on sustainable cities and communities includes Indicator 11.6.2 which highlights the importance of monitoring and improving urban air quality (UN, 2015; Amegah and Agyei-Mensah, 2017). Thus, such targets and indicators highlights the growing recognition of air pollution and the importance of continuous air quality monitoring in improving public health, environmental sustainability, and urban resilience. This is particularly important in rapidly growing countries such as South Africa.

Atmospheric pollutant concentrations are shaped by a combination of land-use patterns, topography, seasonal cycles, meteorological variables, and vegetation cover (Chantara et al., 2012; Sabir et al., 2024; Valjarević, 2024). In urban settings, vegetation serves as an active mitigation tool by absorbing gaseous pollutants, intercepting and retaining particulate matter on leaf surfaces, releasing oxygen, and creating physical barriers that alter local airflow and pollutant dispersal pathways (Rani and Kumar, 2023; Diener and Mudu, 2021). Empirical evidence consistently links higher vegetation coverage—particularly tree canopy and green spaces—with lower ambient levels of key pollutants such as SO₂, NO₂, PM₂.₅, and CO (Rani and Kumar, 2023; Diener and Mudu, 2021). That said, the pollution-reduction capacity of vegetation is highly context-dependent, fluctuating with season, species composition, leaf area index, prevailing weather, and the spatial arrangement of green infrastructure (Diener and Mudu, 2021). Consequently, site-specific evaluations remain indispensable for informing effective policy and planning.

Meteorological parameters including rainfall, temperature, relative humidity, wind speed, and wind direction—exert strong control over both pollutant concentrations and transport directions (Chantara et al., 2012). These influences exhibit pronounced seasonal and diurnal variability: for instance, Zhang et al. (2015) documented marked winter peaks and summer minima for PM₂.₅, PM₁₀, CO, SO₂, and NO₂ in Beijing, while Xu et al. (2011) showed that trace-gas levels in the North China Plain were tightly coupled to wind regimes. Similar meteorological dependencies have been reported elsewhere (Jury, 2017; Danek et al., 2022; Shelton et al., 2022; Nakyai et al., 2025). Nevertheless, most prior work has relied on surface-based measurements to establish statistical correlations between weather variables and pollutant loadings. Given the intricate, non-linear interplay among vegetation structure, atmospheric dynamics, and emissions, a more integrated analysis of air pollution, vegetative cover, and meteorological regimes is warranted to advance mechanistic understanding. This need for integrated approaches is particularly urgent in regions where monitoring capacity remains limited, such as the African continent.

African countries are currently facing limited air quality monitoring systems due to insufficient ground-based infrastructure, in spite of increasing evidence of high greenhouse gases (GHGs). (Arowosegbe et al., 2021). In South Africa, air pollution has been identified as a significant environmental concern, particularly in industrialized regions such as the Highveld Priority Area which includes the Gauteng Province (Josipovic et al., 2010; Shikwambana et al., 2020). The trace gases are mainly emitted from coal fired power stations, petrochemical complexes, vehicular traffic, and mining operations. However, like other Africa countries, the availability of air quality data remains a major challenge with fewer accredited air pollution monitoring stations by the South African National Accreditation System (SANAS) (Scott, 2010). Moreover, operational challenges such as frequent vandalism of monitoring facilities, power outages have led to incomplete or missing temporal air pollutants measurements. These limitations compromise the spatial and temporal coverage of air quality data, making it difficult to conduct reliable monitoring assessments. Consequently, the availability of free Earth observation datasets at high spatial, temporal and spectral resolutions have provided new opportunities and alternative methods for constantly tracking air pollution changes at different scales addressing some of the data gaps. The TROPOspheric Monitoring Instrument (TROPOMI) Sentinel-5P has gained popularity in air pollution monitoring studies (Shami et al., 2022; Shikwambana et al., 2020). The Sentinel-5P TROPOMI sensor ranges from different wavelength such as shortwave infrared and ultraviolet, allowing for near-daily global measurements of trace gases like NO₂ and SO₂. Complementing this, the Sentinel-2A Multispectral Instrument (MSI) by European Space Agency (ESA) (European Space Agency, 2015) also provides high-resolution optical imagery suitable for vegetation monitoring at different spectral bands. The extent and health of vegetation in urban areas are commonly mapped using satellite-derived vegetation indices such as the Normalized Difference Vegetation Index (NDVI). The integration of satellite derived parameters for air quality and vegetation from both these sensors could improve our understanding of seasonal variations of air quality and how vegetation contributes to regulating pollution. This insight is essential for sustainable urban planning, effective public-health interventions, and informed environmental policy. While numerous studies have assessed air quality trends in major metropolitan areas globally, there is a pressing need to investigate how vegetation and meteorological conditions interacts with pollutant concentrations across distinct urban environments using high-resolution satellite data.

This study examines the spatiotemporal patterns of air pollution, vegetation cover, and meteorological variables across the Sandton and Tembisa areas of Gauteng Province, South Africa, throughout 2024. Such an integrated analysis is crucial for elucidating seasonal fluctuations, population exposure risks, and the modulating influence of urban and peri-urban vegetation on local air quality.

2 Materials and methods

2.1 Study area description

The study focuses on Tembisa and Sandton, located in Gauteng Province, South Africa, as depicted in Figure 1. Tembisa (Figure 1a), a large township in the Ekurhuleni Metropolitan Municipality, features a diverse land cover of built-up areas, vegetation, and bare land, including informal settlements that shape its landscape (Brou et al., 2021; Morole et al., 2023). It has a subtropical highland climate with warm, partly cloudy summers, short, cold, dry, and clear winters, and approximately 58 mm of annual precipitation. Sandton (Figure 1b), characterized by residential, commercial, and some agricultural zones, is dominated by woodland and built-up areas, reflecting its status as a key commercial and residential hub. Sandton experiences a subtropical highland climate with warm, sunny summers (December-February, averaging 28 °C) and mild, cool winters (June-August, averaging 14 °C), receiving about 600 mm of rainfall annually.

Figure 1

Figure 1. Study area map of (a) Tembisa and (b) Sandton.

2.2 Data acquisition and preprocessing

2.2.1 Sentinel-5P (TROPOMI)

Launched on 13 October 2017, the Sentinel-5P satellite was designed to monitor air quality, ozone, and surface UV climate. Its TROPOspheric Monitoring Instrument (TROPOMI), a multispectral sensor, measures Earth’s radiance across ultraviolet–visible (UV–VIS, 267–499 nm), near-infrared (NIR, 661–786 nm), and shortwave infrared (SWIR, 2,300–2,389 nm) wavelengths, achieving a ground resolution of up to 5.5 km × 3.5 km. These measurements detect atmospheric pollutants such as aerosols, CO, NO₂, O₃, methane (CH₄), formaldehyde, and SO₂. Compared to earlier instruments, TROPOMI provides data with improved accuracy and spatio-temporal resolution (Theys et al., 2017; Tilstra et al., 2020). Its data is categorized into three levels: Level 0 (raw, non-public telemetry from four spectrometers), Level-1B (geo-located, radiometrically corrected radiances and solar irradiances), and Level-2 (geo-located total columns of SO₂, NO₂, CO, O₃, CH₄, formaldehyde, tropospheric O₃, O₃ profiles, and cloud/aerosol data including aerosol layer height and absorbing aerosol index). Data processing occurs in three modes: near real-time (NRT, available within 3 h), offline (OFFL, available within 12 h), and reprocessing (no time limit).

2.2.2 Sentinel-2A

The Sentinel-2A Multispectral Instrument (MSI), developed by the European Space Agency (ESA), provides high-resolution optical imagery for land monitoring and consists of 13 spectral bands (European Space Agency, 2015). It has a multi-spectral imager with resolutions of 10 m, 20 m, and 60 m depending on the spectral band. Four visible and near-infrared bands have a resolution of 10 m, while the red-edge and shortwave infrared bands have a 20 m resolution, and three bands for atmospheric correction have a 60 m resolution. Sentinel-2A MSI provides two product types: Level-1C (top of atmosphere reflectance) and Level-2A (bottom of atmosphere reflectance). In this study, Sentinel-2A imagery was accessed from the Google Earth Engine (GEE) platform using the COPERNICUS/S2_SR collection, which provides atmospherically corrected surface reflectance data. The images were acquired for summer (01 December 2023 to 28 February 2024) and winter (01 July 2024 to 31 August 2024). The region of interest was spatially clipped using a defined boundary. Cloud masking was performed to minimize atmospheric and clouds effect by selecting scenes with less than 10% cloud cover using the Scene Classification Layer (SLC) and QA60 cloud mask. Additionally, COPERNICUS/S2_CLOUD_PROBABILITY dataset was used to further refine cloud masking. Seasonal composites images were generated by calculating the median of all cloud-free observations during the defined dates of the summer and winter periods. This allowed for consistent seasonal comparison of vegetation patterns.

2.2.3 Meteorological data

MODIS (Moderate Resolution Imaging Spectroradiometer) Land Surface Temperature (LST) data, derived from NASA’s Terra and Aqua satellites, provide global measurements of the Earth’s surface temperature and emissivity. Acquired using thermal infrared bands (e.g., bands 31 and 32), MODIS LST data are generated at 1 km spatial resolution, with daily, 8-day, or monthly temporal resolutions. The data capture day and night surface temperatures, useful for studying climate, urban heat islands, and environmental processes. Products like MOD11 (Terra) and MYD11 (Aqua) are processed to account for atmospheric effects, cloud cover, and land cover types, ensuring accuracy. For more details, refer to Wan et al. (2002).

Wind speed data from the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2), produced by NASA’s Global Modeling and Assimilation Office (GMAO), provide global wind measurements at multiple heights (e.g., 2 m, 10 m, and 50 m) starting from 1980. Generated using the Goddard Earth Observing System Model (GEOS-5.12.4), MERRA-2 offers hourly and monthly mean wind speed data, including eastward (U) and northward (V) components, at a spatial resolution of approximately 0.5 ° × 0.625 ° (∼50 km). The dataset, part of collections like M2IMNXLFO (surface wind speed) and M2T1NXSLV (single-level diagnostics), is derived from assimilated satellite and conventional observations, updated monthly with a ∼3-week latency. For further details, see Gelaro et al. (2017).

Precipitation data from the Global Precipitation Measurement (GPM) mission, a joint NASA-JAXA initiative, provides global rainfall and snowfall estimates with high spatial and temporal resolution. The GPM Core Observatory, launched in 2014, uses the Dual-frequency Precipitation Radar (DPR) and GPM Microwave Imager (GMI) to measure precipitation at a spatial resolution of ∼5–10 km. Data products, such as the Integrated Multi-satellitE Retrievals for GPM (IMERG), offer near-real-time and final-run precipitation estimates at 0.1 ° × 0.1 ° (∼10 km) resolution, available in half-hourly, daily, or monthly formats. These data are critical for weather forecasting, hydrology, and climate studies. For more details, see Hou et al. (2014).

2.3 Derivation of vegetation indices

2.3.1 The modified Soil-Adjusted vegetation index

The Modified Soil Adjusted Vegetation Index (MSAVI) (Qi et al., 1994) is a widely utilized remote sensing index designed to monitor vegetation across varied landscapes. It was developed as an improvement over the standard Soil Adjusted Vegetation Index (SAVI), specifically to mitigate the influence of soil background effects that can distort vegetation reflectance measurements in areas with sparse vegetation cover. MSAVI incorporates a dynamic soil adjustment factor (L), enhancing the separation of vegetation and soil signals in the red and near-infrared (NIR) spectral bands (Qi et al., 1994; Kareem et al., 2023). One of its key advantages is its robust performance in conditions where high soil reflectance may interfere with vegetation detection. Research has shown that MSAVI yields high accuracy in estimating biomass and canopy density compared to traditional indices such as the Normalized Difference Vegetation Index (NDVI), especially in environments characterized by low vegetation cover or significant soil exposure (Jiang et al., 2007). Threshold values for MSAVI range from −1 to +1, where values below 0 indicate non-vegetated surfaces such as water, bare soil, or built-up areas. Values between 0.0 and 0.2 generally correspond to very low vegetation cover, 0.2 to 0.4 indicate moderate vegetation, and values above 0.4 suggest dense vegetation cover (Gaznayee et al., 2022; Kareem et al., 2023). MSAVI is computed using the equation:

$MSAVI=\frac{2*\text{NIR}+1-\sqrt{{\left(2*NIR+1\right)}^2-8\left(\text{NIR}-\text{Red}\right)}}{2}(1)$

where, NIR is reflectance in the near infrared band, and Red is the reflectance in the red band.

2.4 Extraction of air quality parameters

The Air Quality Index (AQI) is a numerical value that reflects the hourly or daily air quality in a specific area. It is calculated using Equation 2,

$AQI=\left(I_{P1}+I_{P2}+\dots +I_{Pn}\right)/n(2)$

where $I_P$ represents the sub-index for each pollutant’s concentration, and $n$ is the total number of pollutants. The sub-index is determined by Equation 2,

$I_P=\frac{I_{high}-I_{low}}{C_{high}-C_{low}}*\left(C-C_{low}\right)+I_{low}(3)$

where $C$ is the pollutant concentration, $C_{high}$ is the concentration breakpoint that is ≤ $C$, $C_{low}$ is the concentration breakpoint that is ≥ $C$. $I_{high}$ is index breakpoint corresponding to $C_{high}$ and $I_{low}$ is index breakpoint corresponding to $C_{low}$. According to Savenets (2021), the total column data from TROPOMI cannot be accurately converted to near-ground values. However, near-ground concentration can be estimated using Equation 3:

$C=\frac{C_{col}}{H}.M.A(4)$

where $C_{col}$ is the pollutant column content (mol/m²), $M$ is the molar mass (g/mol), $A$ is a constant for conversion from (g/m³) to (µg/m³), H is the height (m) near the surface where most pollutants are concentrated.

2.5 Statistical analysis and validation

Pearson’s correlation coefficient is a statistical measure used to quantify the strength and direction of the linear relationship between two continuous variables, $x_i$ and $y_i$. It is widely applied in scientific research to assess how closely data points align with a linear trend, with values ranging from −1 (perfect negative correlation) to 1 (perfect positive correlation), where 0 indicates no linear correlation. The coefficient, denoted r, is calculated as the covariance of $x_i$ and $y_i$ divided by the product of their standard deviations:

$\mathit{r}=\frac{\sum \left({\mathit{x}}_{\mathit{i}}-\overline{\mathit{x}}\right)\left({\mathit{y}}_{\mathit{i}}-\overline{\mathit{y}}\right)}{\sqrt{\sum {\left({\mathit{x}}_{\mathit{i}}-\overline{\mathit{x}}\right)}^{\mathbf{2}}\sum {\left({\mathit{y}}_{\mathit{i}}-\overline{\mathit{y}}\right)}^{\mathbf{2}}}}(5)$

where, $r$ is the correlation coefficient, variables $x_i$ and $y_i$ represent values for each respective variable $x$ and $y$, respectively, and $\overline{x}$ and $\overline{y}$ are mean values of the $x$ and $y$ variables, respectively. More information on the Pearson correlation test can be found in Benesty et al. (2009).

3 Results and analysis

3.1 Analysis of pollutant concentrations and air quality index in tembisa

3.1.1 NO₂ concentrations

Figure 2 illustrates the spatial and temporal variations in pollutant concentrations and the Air Quality Index (AQI) over Tembisa during the summer (December–February, DJF) and winter (June–August, JJA) seasons. As shown in Figure 2a, NO₂ concentrations are notably higher during the JJA season compared to the DJF season (see Figures 2a,b). During the DJF season, NO₂ concentrations are lower, averaging 4 μg/m³ in northern Tembisa, but slightly elevated at 10 μg/m³ in southern Tembisa. The elevated concentrations in the southern region are primarily attributed to the prevalence of informal settlements with limited access to municipal services, such as electricity, leading to widespread reliance on biomass burning for cooking and other domestic activities. In contrast, during the JJA season, NO₂ concentrations are uniformly distributed across Tembisa, averaging approximately 14 μg/m³. Major sources of NO₂ in the region include vehicular traffic, household fuel combustion (particularly for heating), and industrial activities (Shirinde et al., 2014). Additional contributions arise from OR Tambo International Airport, which emits low-level, concentrated NO₂, and large industries with stack, vent, and fugitive emissions (Scorgie et al., 2005). Meteorological conditions, such as temperature inversions and reduced atmospheric mixing during winter, further exacerbate NO₂ concentrations (Rodríguez-Sánchez et al., 2024).

Figure 2

Figure 2. Spatial distribution of NO₂ concentration during the (a) DJF and (b) JJA seasons. Spatial distribution of SO₂ concentration during the (c) DJF and (d) JJA seasons in Tembisa. (e) Timeseries of AQI for the for the DJF and JJA season for the period 2023-2.

3.1.2 SO₂ concentrations

Figures 2c,d depict SO₂ concentrations during the DJF and JJA seasons, respectively. Similar to NO₂, SO₂ concentrations are lower during the DJF season, with values generally ≤20 μg/m³ across Tembisa, except in the northern region, where concentrations reach approximately 50 μg/m³ due to localized industrial activities. During the JJA season, SO₂ concentrations are significantly higher, ranging from moderate (60 μg/m³) to elevated (100 μg/m³) levels. Central Tembisa exhibits the highest SO₂ concentrations, while the peripheral areas show moderate levels. Key sources of SO₂ emissions include industrial and commercial fuel combustion, vehicle exhausts, and household fuel burning (EM Municipality, 2015). As with NO₂, favourable meteorological conditions, such as stable atmospheric conditions during winter, contribute to elevated SO₂ concentrations (Rodríguez-Sánchez et al., 2024).

3.1.3 AQI analysis

The AQI, ranging from 0 to 500, quantifies air quality, with lower values indicating better air quality and higher values signifying increased health risks. The AQI categories are defined as follows: Good (0–50, minimal risk), Moderate (51–100, acceptable but with potential concern for sensitive individuals), Unhealthy for Sensitive Groups (101–150, health effects for sensitive populations), Unhealthy (151–200, health effects for the general public, with more severe impacts on sensitive groups), Very Unhealthy (201–300, emergency health warnings), and Hazardous (301–500, severe health effects for all). Figure 2e presents the AQI time series for Tembisa during the DJF (red) and JJA (blue) seasons. The DJF season is predominantly characterized by Good to Moderate AQI days, reflecting low pollutant levels facilitated by favourable meteorological conditions that promote pollutant dispersion. Infrequent instances of Unhealthy for Sensitive Groups days occur but are not representative of the season. In contrast, the JJA season exhibits a higher frequency of Moderate and Unhealthy for Sensitive Groups AQI days, driven by meteorological conditions such as temperature inversions and reduced wind speeds that limit pollutant dispersion. Additionally, isolated days of Unhealthy, Very Unhealthy, and Hazardous AQI levels are observed, indicating periodic severe air quality degradation during winter.

3.2 Analysis of pollutant concentrations and air quality index in sandton

3.2.1 NO₂ concentrations

Figure 3 illustrates the spatial and temporal variability of pollutant concentrations and the AQI across Tembisa during the DJF and JJA seasons. During the DJF season (Figure 3a), NO₂ concentrations in Sandton are relatively low, ranging from 5 to 7 μg/m³. The eastern region of Sandton exhibits slightly elevated concentrations of 7 μg/m³ compared to the western region, where concentrations are approximately 5 μg/m³. The eastern part of Sandton borders Alexandra, a township characterized by a mix of residential, commercial, and open spaces, transitioning from high-density residential areas to informal settlements (Brou et al., 2021). Primary NO₂ emission sources in eastern Sandton include industrial activities, biomass burning, and vehicular emissions. In contrast, the central and western regions of Sandton are characterized by urban development with significant tree cover in residential areas and parks, where traffic emissions are the dominant NO₂ source. Low NO₂ concentrations during the DJF season are attributed to meteorological conditions favouring pollutant dispersion, such as unstable atmospheric conditions, strong winds, and precipitation (Jury, 2017; Lourens et al., 2011). Conversely, stable atmospheric conditions inhibit vertical pollutant dispersion, leading to elevated near-surface concentrations. Additionally, recirculation events over the Highveld can trap and sustain high pollutant levels (Freiman and Piketh, 2003). During the JJA season (Figure 3b), NO₂ concentrations increase significantly, ranging from 13 to 15 μg/m³. Higher concentrations of 15 μg/m³ are observed in the southern and eastern parts of Sandton, compared to 13 μg/m³ in the northern and western regions. Although southern Sandton’s land cover comprises a mix of natural and built-up areas, localized commercial agriculture and mining activities likely contribute to the elevated NO₂ concentrations during winter.

Figure 3

Figure 3. Spatial distribution of NO₂ concentration during the (a) DJF and (b) JJA seasons. Spatial distribution of SO₂ concentration during the (c) DJF and (d) JJA seasons in Sandton. (e) Timeseries of AQI for the DJF and JJA season for the period 2023-2.

3.2.2 SO₂ concentrations

Figures 3c,d depict SO₂ concentrations during the DJF and JJA seasons, respectively. During the DJF season, SO₂ concentrations are relatively low, ranging from 20 to 40 μg/m³. The northern part of Sandton exhibits concentrations of approximately 40 μg/m³, while other areas maintain concentrations around 20 μg/m³. Major SO₂ sources in Sandton include industrial activities and vehicular emissions, with additional contributions from long-range transport of SO₂ from coal-fired power stations in the Emalahleni region (Zunckel, 1999; Zunckel et al., 2000). During the JJA season (Figure 3d), SO₂ concentrations increase significantly, ranging from 70 to 100 μg/m³. The highest concentrations, reaching 100 μg/m³, are observed in eastern Sandton, likely due to industrial activities and informal settlements in the area. The remainder of Sandton exhibits moderate SO₂ concentrations of approximately 70 μg/m³.

3.2.3 AQI analysis

Figure 3e presents the AQI time series for Sandton during the DJF (red) and JJA (blue) seasons. Similar to Tembisa, the DJF season is predominantly characterized by Good to Moderate AQI days, reflecting low pollutant concentrations facilitated by meteorological conditions that promote dispersion, such as unstable atmospheric conditions, strong winds, and precipitation. Days classified as Unhealthy for Sensitive Groups are infrequent, and no Hazardous AQI days are recorded during this period. In contrast, the JJA season is dominated by days classified as Unhealthy for Sensitive Groups, with a few days reaching Unhealthy levels and 1 day recorded as Hazardous. Stable atmospheric conditions and recirculation events during the JJA season contribute to pollutant trapping, resulting in elevated AQI values.

3.3 Analysis of the vegetation conditions in tembisa and sandton

3.3.1 MSAVI spatial variability analysis in tembisa

Figure 4 shows the spatial variability of vegetation in Tembisa for DJF (Figure 4a) and JJA (Figure 4b). During summer, vegetation greenness is higher in the southern and southeastern parts where values range (0.6–0.67, dark green), reflecting healthy vegetation growth during the rainy season. These zones are linked to rainfall-driven vegetation dynamics, where seasonal precipitation enhances productivity, consistent with findings across southern Africa where greenness peaks in summer months (Funk et al., 2015). The central and northern sections are dominated by low values near 0.0 (orange to brown), indicating sparse vegetation and built-up areas. This aligns with previous studies that have indicated that high housing or settlement density and limited green infrastructure result in reduced vegetation cover especially in low-income urban areas (Schwarz et al., 2015; Adole et al., 2019).

Figure 4

Figure 4. Spatial variability of vegetation in Tembisa during the (a) summer (December–February) and (b) winter (June–August) seasons for 2023–2024 using the Modified Soil Adjusted Vegetation Index (MSAVI).

In winter much of the settlement shifts to negative or near-zero values (0.25–0.1, brown to light orange), showing widespread decline in vegetation. Only a few small patches remain above 0.4 (dark green), mostly along the southern edge and open spaces. This low vegetation cover could reflect the dependence of Tembisa’s vegetation on rainfall rather than irrigation or managed greenery. Thus, such seasonal vulnerability has also been noted in other African cities, where vegetation cover is highly sensitive to climatic variability and socio-economic constraints (Adole et al., 2019). Overall, Tembisa displays patchy and highly seasonal vegetation, with limited greenness compared to summer months and a sharp decline in winter.

3.3.2 MSAVI spatial variability analysis in sandton

Figure 5 illustrates the vegetation spatial variability for Sandton during DJF (Figure 5a) and JJA (Figure 5b). In DJF, most of the settlement records high MSAVI values (0.4–0.67, dark green) reflecting widespread and healthy vegetation linked to trees, irrigated gardens, and managed landscapes or green-spaces. Very few areas fall in the lower range (<0.1, brown to orange), which are restricted to dense built-up surfaces. In JJA, although greenness decreases, Sandton maintains large portions of moderate to high MSAVI values (0.2–0.5, light to dark green), particularly in the central and northern sections. Only limited patches shift to very low values (<0.0–0.1, brown to orange). This resilience reflects the role of irrigation, planned green infrastructure, and lower settlement density in sustaining vegetation during the dry season. Studies of urban vegetation dynamics similarly show that socio-economically advantaged areas exhibit reduced seasonal variability due to continuous water supply and active management of green infrastructure (Schwarz et al., 2015). Aronson et al. (2017) also highlights that higher-income neighborhoods tend to support more stable and diverse vegetation cover because of socio-economic investment in landscaping and tree planting. Overall, Sandton demonstrates strong year-round vegetation presence, with high greenness in summer and only moderate decline in winter, indicating more stable and continuous vegetation cover than Tembisa.

Figure 5

Figure 5. Spatial variability of vegetation in Sandton during the (a) summer (December–February) and (b) winter (June–August) seasons for 2023–2024 using the Modified Soil Adjusted Vegetation Index (MSAVI).

3.4 Relationship between air pollutants, vegetation indices, and meteorological parameters

3.4.1 Correlation analysis in tembisa

The results of the statistical relationships between different parameters during the DJF and JJA seasons are presented in Figures 6, 7. For simplicity of interpretations, the $r$ coefficients are grouped into negligible (±0.0 to ±0.3), weak (±0.31 to ±0.5), moderate (±0.51 to ±0.7), high (±0.71 to ±0.9), and very high (±0.91 to ±1.0), negative or positive correlations (Shikwambana et al., 2022). The parameters that are analysed are temperature (LST), MSAVI, SO₂, NO₂, precipitation (PRECIP) and WIND.

Figure 6

Figure 6. Correlation matrix showing the relationships between Land Surface Temperature (LST), Modified Soil Adjusted Vegetation Index (MSAVI), sulphur dioxide (SO₂), nitrogen dioxide (NO₂), precipitation (PRECIP), and wind speed during the (a) DJF (December–February) and (b) JJA (June–August) seasons for the year 2023–2024 in Tembisa. The figure highlights seasonal variations in the interactions among these environmental variables.

Figure 7

Figure 7. Correlation matrix showing the relationships between Land Surface Temperature (LST), Modified Soil Adjusted Vegetation Index (MSAVI), sulphur dioxide (SO₂), nitrogen dioxide (NO₂), precipitation (PRECIP), and wind speed during the (a) DJF (December–February) and (b) JJA (June–August) seasons for the year 2023–2024 in Sandton. The figure highlights seasonal variations in the interactions among these environmental variables.

Figure 6 illustrates the correlations among various parameters in Tembisa during the DJF and JJA seasons. In the DJF season, moderate to very high correlations are prevalent. Notably, a strong positive correlation of 0.99 is observed between SO₂ and MSAVI, indicating that higher SO₂ levels are associated with increased vegetation cover or health. Although this appears counterintuitive due to SO₂’s role as a pollutant, several possible explanations exist. Sulphur, in small quantities, can act as a micronutrient that promotes short-term plant growth. Additionally, this relationship may be influenced by external factors such as economic activity or climatic conditions affecting both emissions and vegetation. Other contributing factors could include spatial averaging effects or the presence of vegetation types that are resistant or adaptive to polluted environments. Another notable correlation is between MSAVI and wind speed, which shows a strong negative relationship of −0.99. This indicates that as wind speed increases, vegetation cover or health (as measured by MSAVI) decreases. Such a strong negative correlation suggests that higher wind activity may stress, damage, or reduce vegetation, resulting in lower vegetation indices. This highlights the potential role of wind as an environmental stressor, particularly in areas that are already vulnerable due to dry conditions, poor soil quality, or the presence of sensitive vegetation types. However, a few negligible correlations are observed in the data. For instance, the correlation between MSAVI and wind speed (0.03) suggests that vegetation in the area may be resilient to wind, or that wind speeds are generally too low to significantly impact plant health. Similarly, the weak negative correlation between SO₂ and precipitation (−0.16) may be explained by the ability of rainfall to remove SO₂ from the atmosphere through the process of wet deposition. Another weak negative correlation is observed between MSAVI and precipitation (−0.11), which, although counterintuitive, could be due to excessive rainfall over short periods leading to waterlogging, soil erosion, or runoff—conditions that can harm vegetation rather than support it.

During the JJA season, several strong correlations are observed between environmental variables. A very strong positive correlation between NO₂ and land surface temperature (LST) (0.98) suggests that higher temperatures are closely associated with increased NO₂ concentrations. This could be due to wintertime temperature inversions that trap pollutants near the surface, allowing NO₂ to accumulate, or increased emissions from heating sources on warmer days. In contrast, the strong negative correlation between MSAVI and precipitation (−0.89) indicates that increased rainfall is associated with reduced vegetation cover or health. This counterintuitive relationship may result from heavy or poorly timed rainfall causing waterlogging, erosion, or runoff that negatively affects vegetation during the winter months. Additionally, a strong negative correlation between NO₂ and wind speed (−0.97) implies that higher wind speeds help disperse NO₂, leading to lower concentrations, while low wind conditions allow pollutants to accumulate. Together, these correlations highlight how temperature, precipitation, and wind dynamics significantly influence both vegetation and air quality in winter.

3.4.2 Correlation analysis in sandton

During the summer season, several strong correlations are evident between key environmental variables. The very strong negative correlation between MSAVI and LST (−0.99) indicates that higher land surface temperatures are associated with reduced vegetation cover or health, likely due to heat stress and increased evapotranspiration affecting plant growth. Conversely, the strong positive correlation between SO₂ and MSAVI (0.98) suggests that areas with higher SO₂ emissions also exhibit increased vegetation, which could be influenced by localized factors such as irrigated or managed vegetation near emission sources, or a short-term nutrient effect of sulphur on plants. The strong positive correlation between NO₂ and precipitation (0.89) may reflect that increased rainfall is linked to higher NO₂ concentrations, possibly due to atmospheric conditions that favour pollutant accumulation or transport during rainy periods. Finally, the strong negative correlation between SO₂ and wind speed (−0.99) highlights the role of wind in dispersing SO₂ pollutants, where higher wind speeds reduce local SO₂ concentrations by promoting atmospheric mixing and dilution. Together, these relationships underscore the complex interplay of temperature, pollution, precipitation, and vegetation dynamics during summer.

On the other hand, the correlation between NO₂ and MSAVI is very weakly negative (−0.09), suggesting little to no meaningful relationship between NO₂ levels and vegetation during this period. Similarly, the weak negative correlation between SO₂ and precipitation (−0.17) implies that increased rainfall may slightly reduce sulphur dioxide concentrations, possibly through wet deposition, but the effect is minimal. Lastly, the near-zero positive correlation between precipitation and wind speed (0.02) indicates almost no relationship between these two variables during summer, suggesting that changes in wind speed do not significantly influence rainfall patterns in this context. Overall, these correlations reflect the dominant influence of temperature on vegetation health in summer, while other environmental factors show minimal direct interactions.

During the winter season, a strong negative correlation between MSAVI and LST (−0.99) indicates that higher land surface temperatures are closely linked to reduced vegetation health or cover, possibly due to temperature-related stress or other environmental factors affecting plant growth during colder months. A strong positive correlation between SO₂ and NO₂ (0.89) suggests that these two pollutants often increase together, likely originating from similar sources such as combustion processes or traffic emissions. The strong negative correlation between NO₂ and wind speed (−0.86) implies that higher wind speeds help disperse nitrogen dioxide, reducing its concentration near the surface, while calmer conditions allow pollutants to accumulate. Similarly, the negative correlation between precipitation and wind speed (−0.74) suggests that increased rainfall is associated with lower wind speeds, which may affect the transport and dispersion of pollutants and moisture in the atmosphere. Together, these relationships highlight the complex interactions between vegetation, air quality, and meteorological conditions during the winter season. The very slight negative correlation between MSAVI and precipitation (−0.07) suggests that rainfall has little to no direct impact on vegetation cover or health during this period. Similarly, the weak positive correlation between land surface temperature (LST) and precipitation (0.11) indicates a minimal tendency for higher temperatures to be associated with increased rainfall, though this relationship is not strong. Lastly, the weak negative correlation between SO₂ and precipitation (−0.18) suggests that rainfall may slightly reduce SO₂ levels through wet deposition, but this effect is relatively small. Overall, these weak correlations imply that precipitation plays a limited role in influencing vegetation, temperature, and SO₂ concentrations during winter in this context.

4 Discussion

In urban green environments, vegetation significantly contributes to air quality improvement by functioning as a natural barrier and sink for various pollutants. It also plays a key role in modifying local weather conditions, creating more comfortable and stable microclimates. Trees with high leaf area index (LAI) and dense canopy structures are especially efficient in removing airborne contaminants through a process known as phytoremediation. Past research shows that urban greenery can reduce concentrations of particulate matter by 16.5%–26.7%, NO₂ by 13.9%–36.2%, and SO₂ by as much as 34.9%, primarily via dry deposition on foliage surfaces (Gong et al., 2023). Leaf characteristics such as the presence of waxy or hairy textures enhance the capture of fine particles, and this efficiency increases with greater vegetation density and species variety. More extensive and diverse vegetation coverage is strongly linked with better AQI ratings, thus lowering public health risks associated with pollution.

Urban vegetation also impacts the local climate by moderating temperatures typically lowering them by 2 °C–8 °C through the processes of evapotranspiration and shading (Wu et al., 2025). This helps to counteract the urban heat island (UHI) effect, where city temperatures can exceed those of surrounding rural areas by 5 °C–10 °C at night. Furthermore, trees and plants intercept rainfall, decreasing surface runoff by up to 30% and naturally filtering contaminants from stormwater (Smets et al., 2019). Increased moisture in the air due to evapotranspiration aids in diluting gaseous pollutants. Dense plantings can alter airflow patterns, occasionally generating localized “park breezes” that improve ventilation; however, in narrow urban canyons, this same vegetation may restrict air movement and trap pollutants if not strategically designed.

In comparison, semi-urban zones often marked by lower vegetation density within mixed residential or commercial developments demonstrate weaker connections between vegetation, air quality, and microclimate regulation. These areas serve as transitional zones between cities and rural regions and often absorb pollution spillover from nearby urban centers, particularly from traffic emissions. However, due to insufficient green cover, these areas lack the capacity to effectively mitigate pollution. As a result, pollutant levels tend to be higher than in urban parks, though sometimes lower than in city cores. The climate-modifying effects of vegetation in semi-urban areas are also diminished: limited tree cover results in only modest cooling effects, typically between 1 °C and 4 °C. Although wind plays a more prominent role in dispersing pollutants in these areas, it can also transport them from urban to semi-urban locations.

Moreover, impervious surfaces in these areas lead to greater runoff of about 15%–48% more than in greener areas often carrying pollutants into waterways (Zhou, 2019). Data also indicate positive correlations between land surface temperature (LST), and pollutant levels with densely populated areas experiencing amplified effects. Unlike in urban green spaces, limited vegetation in semi-urban zones often leads to reduced airflow and poor pollutant dispersion, especially under low-wind conditions. Seasonal factors such as winter inversions can further degrade air quality during certain times of the year. In general, the environmental benefits of vegetation in these areas are minimal, making them more susceptible to pollution buildup and unstable weather patterns.

Across both urban and semi-urban environments, climate change introduces compounding feedback loops. When vegetation becomes stressed due to extreme weather or pollution, its ability to cool the environment and filter air is diminished. This deterioration worsens air quality and climatic conditions, which in turn accelerates emissions. To improve resilience, strategies such as increasing green coverage beyond 27% (World Health Organization, 2016b) and choosing plant species that can withstand climate stressors are essential.

5 Conclusion

Here are the five key finding from the study:

1. Urban vs. Semi-Urban Contrast: Sandton maintains stable, year-round vegetation due to irrigation and managed green spaces, while Tembisa shows fragmented, highly seasonal vegetation with sharp winter decline. This difference strongly influences pollutant dispersion and microclimate regulation.

2. Seasonal Pollution Dynamics: Both NO₂ and SO₂ concentrations spike during winter in Tembisa and Sandton, driven by temperature inversions, low wind speeds, and increased combustion activities. AQI shifts from “Good/Moderate” in summer to “Unhealthy” or worse in winter.

3. Vegetation–Pollution Correlations: Unexpected strong positive correlations between SO₂ and MSAVI in summer suggest complex interactions, possibly short-term nutrient effects or spatial averaging, challenging the assumption that pollution always harms vegetation because some vegetation species such as the Cinnamomum camphora (camphor tree) are air pollution tolerant and serve as natural sinks for acidic pollutants like SO₂.

4. Meteorological Influence: Wind speed and precipitation play contrasting roles: wind disperses pollutants effectively, while precipitation sometimes correlates negatively with vegetation health (likely due to waterlogging or erosion), especially in winter.

5. Policy Implication: Findings underscore the need for integrated urban planning, expanding green infrastructure and selecting resilient species to mitigate air pollution and climate stressors in rapidly urbanizing regions.

Data availability statement

The datasets for this study can be found in the following links; https://code.earthengine.google.com/ for Sentinel and MODIS data, and https://giovanni.gsfc.nasa.gov/ for MERRA-2 data.

Author contributions

SN: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing. NN: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – review and editing. LS: Conceptualization, Formal Analysis, Methodology, Writing – review and editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Acknowledgements

We thank the European Space Agency for providing the satellite imagery used in this study.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Adole, T., Stringer, L. C., Dougill, A. J., Ross, C. W., Ji, W., Kumar, S. S., et al. (2019). Variability of vegetation cover in Sub-Saharan Africa: implications for land degradation monitoring. Remote Sens. 11 (5), 576. doi:10.3390/rs11050576

CrossRef Full Text | Google Scholar

Altieri, K. E., and Keen, S. L. (2019). Public health benefits of reducing exposure to ambient fine particulate matter in South Africa. Sci. Total Environ. 684, 610–620. doi:10.1016/j.scitotenv.2019.05.355

PubMed Abstract | CrossRef Full Text | Google Scholar

Amegah, A. K., and Agyei-Mensah, S. (2017). Urban air pollution in Sub-Saharan Africa: time for action. Environ. Pollut. 220, 738–743. doi:10.1016/j.envpol.2016.09

PubMed Abstract | CrossRef Full Text | Google Scholar

Aronson, M. F. J., Lepczyk, C. A., Evans, K. L., Goddard, M. A., Lerman, S. B., MacIvor, J. S., et al. (2017). Biodiversity in the city: key challenges for urban green space management. Front. Ecol. Environ. 15 (4), 189–196. doi:10.1002/fee.1480

CrossRef Full Text | Google Scholar

Arowosegbe, O. O., Röösli, M., Künzli, N., Saucy, A., Adebayo-Ojo, T. C., Jeebhay, M. F., et al. (2021). Comparing methods to impute missing daily ground-level PM10 concentrations between 2010–2017 in South Africa. Int. J. Environ. Res. Public Health 18 (7), 3374. doi:10.3390/ijerph18073374

PubMed Abstract | CrossRef Full Text | Google Scholar

Benesty, J., Chen, J., Huang, Y., and Cohen, I. (2009). Pearson correlation coefficient. Noise Reduction Speech Processing 2, 37–40. doi:10.1007/978-3-642-00296-0_5

CrossRef Full Text | Google Scholar

Brou, Y. T., Shikwambana, L., and Sivakumar, V. (2021). “Dynamics of climatic and vegetation parameters in urban and township areas: a case study over the city of johannesburg and alexandra township in South Africa,” in Proceedings of the Earth and environmental sciences international webinar conference (Cham: Springer International Publishing), 431–448.

CrossRef Full Text | Google Scholar

Chantara, S., Sillapapiromsuk, S., and Wiriya, W. (2012). Atmospheric pollutants in Chiang Mai (thailand) over a five-year period (2005–2009), their possible sources and relation to air mass movement. Atmos. Environ. 60, 88–98. doi:10.1016/j.atmosenv.2012.06.044

CrossRef Full Text | Google Scholar

Danek, T., Weglinska, E., and Zareba, M. (2022). The influence of meteorological factors and terrain on air pollution concentration and migration: a geostatistical case study from krakow, Poland. Sci. Rep. 12, 11050. doi:10.1038/s41598-022-15160-3

PubMed Abstract | CrossRef Full Text | Google Scholar

Diener, A., and Mudu, P. (2021). How can vegetation protect us from air pollution? A critical review on green spaces’ mitigation abilities for air-borne particles from a public health perspective — with implications for urban planning. Sci. Total Environ. 796, 148605. doi:10.1016/j.scitotenv.2021.148605

PubMed Abstract | CrossRef Full Text | Google Scholar

Dominick, D., Juahir, H., Latif, M. T., Zain, S. M., and Aris, A. Z. (2012). Spatial assessment of air quality patterns in Malaysia using multivariate analysis. Atmos. Environment 60, 172–181. doi:10.1016/j.atmosenv.2012.06.021

CrossRef Full Text | Google Scholar

EM municipality (2015). Climate change response strategy. Available online at: https://www.ekurhuleni.gov.za/wp-content/uploads/2022/05/13.-2015-Climate-Change-Response-Strategy_Final.pdf.

Google Scholar

European Space Agency (2015). Sentinel-2 user handbook, ESA standard document. Paris, France: European Space Agency.

Google Scholar

Freiman, M. T., and Piketh, S. J. (2003). Air transport into and out of the industrial Highveld region of South Africa. J. Appl. Meteorology Climatol. 42 (7), 994–1002. doi:10.1175/1520-0450(2003)042<0994:ATIAOO>2.0

CrossRef Full Text | Google Scholar

Funk, C., Verdin, A., Michaelsen, J., Peterson, P., Pedreros, D., and Husak, G. (2015). A global satellite-assisted precipitation climatology. Earth Syst. Sci. Data 7, 275–287.

CrossRef Full Text | Google Scholar

Gaznayee, H. A. A., Al-Quraishi, A. M. F., Mahdi, K., Messina, J. P., Zaki, S. H., Razvanchy, H. A. S., et al. (2022). Drought severity and frequency analysis aided by spectral and meteorological indices in the Kurdistan region of Iraq. Water 14 (19), 3024. doi:10.3390/w14193024

CrossRef Full Text | Google Scholar

Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A., Takacs, L., et al. (2017). The modern-era retrospective analysis for research and applications, version 2 (MERRA-2). J. Clim. 30 (15), 5419–5454. doi:10.1175/JCLI-D-16-0758.1

PubMed Abstract | CrossRef Full Text | Google Scholar

Gong, C., Xian, C., Wu, T., Liu, J., and Ouyang, Z. (2023). Role of urban vegetation in air phytoremediation: differences between scientific research and environmental management perspectives. Urban Sustain. 3 (1), 24. doi:10.1038/s42949-023-00105-0

CrossRef Full Text | Google Scholar

He, C., Liu, J., Zhou, Y., Zhou, J., Zhang, L., Wang, Y., et al. (2024). Synergistic PM2. 5 and O3 control to address the emerging global PM2. 5-O3 compound pollution challenges. Eco-Environment & Health 3 (3), 325–337. doi:10.1016/j.eehl.2024.04.004

PubMed Abstract | CrossRef Full Text | Google Scholar

Health Organization, W. (2016a). Ambient air pollution: a global assessment of exposure and burden of disease. Clean. Air J. 26, 6. doi:10.17159/2410-972x/2016/v26n2a4

CrossRef Full Text | Google Scholar

Hou, A. Y., Kakar, R. K., Neeck, S., Azarbarzin, A. A., Kummerow, C. D., Kojima, M., et al. (2014). The global precipitation measurement mission. Bull. Am. Meteorological Soc. 95 (5), 701–722. doi:10.1175/BAMS-D-13-00164.1

CrossRef Full Text | Google Scholar

Jiang, Z., Huete, A. R., Li, J., and Qi, J. (2007). Interpretation of the modified soil-adjusted vegetation index isolines in red–NIR reflectance space. J. Appl. Remote Sens. 1 (1), 01350.

Google Scholar

Josipovic, M., Annegarn, H. J., Kneen, M. A., Pienaar, J. J., and Piketh, S. J. (2010). Concentrations, distributions and critical level exceedance assessment of SO₂, NO₂ and O₃ in South Africa. Environ. Monit. Assess. 171, 181–196. doi:10.1007/s10661-009-1270-5

PubMed Abstract | CrossRef Full Text | Google Scholar

Jury, M. R. (2017). Statistics and meteorology of air pollution episodes over the South African Highveld based on satellite–model datasets. J. Appl. Meteor. Climatol. 56, 1583–1594. doi:10.1175/JAMC-D-16-0354.1

CrossRef Full Text | Google Scholar

Kareem, H. H., Attaee, M. H., and Omran, Z. A. (2023). Evaluation the soil-adjusted vegetation indices SAVI and MSAVI for Bristol city, United Kingdom using landsat 8-OLI through geospatial technology. Ecol. Eng. & Environ. Technol. 24, 89–97. doi:10.12912/27197050/169749

CrossRef Full Text | Google Scholar

Khoder, M. I. (2002). Atmospheric conversion of sulfur dioxide to particulate sulfate and nitrogen dioxide to particulate nitrate and gaseous nitric acid in an urban area. Chemosphere 49 (6), 675–684. doi:10.1016/S0045-6535(02)00391-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Kim, S. (2019). Decomposition analysis of greenhouse gas emissions in Korea’s transportation sector. Sustainability 11, 1986. doi:10.3390/su11071986

CrossRef Full Text | Google Scholar

Lin, C.-A., Chen, Y.-C., Liu, C.-Y., Chen, W.-T., Seinfeld, J. H., and Chou, C. C.-K. (2019). Satellite-derived correlation of SO2, NO2, and aerosol optical depth with meteorological conditions over east Asia from 2005 to 2015. Remote Sens. 11 (15), 1738. doi:10.3390/rs11151738

CrossRef Full Text | Google Scholar

Lourens, A. S., Beukes, J. P., van Zyl, P. G., Fourie, G. D., Grey, J. W., Pienaar, J. J., et al. (2011). Spatial and temporal assessment of gaseous pollutants in the Highveld of South Africa. South Afr. J. Sci. 107 (1/2), 8. doi:10.4102/sajs.v107i1/2.269

CrossRef Full Text | Google Scholar

Morole, M. S., Yessoufou, K. Y., Bufo, S. A., Rampedi, I. T., and Modley, L. S. (2023). The impact of informal settlements on the surface water quality: a baseline analysis of the kaalspruit in tembisa township, South Africa. Urban Water J. 20 (10), 1756–1765. doi:10.1080/1573062X.2022.2099289

CrossRef Full Text | Google Scholar

Naboureh, A., Li, A., Bian, J., Lei, G., Nan, X., Zhang, Z., et al. (2024). Green space coverage versus air pollution: a cloud-based remote sensing data analysis in sichuan, Western China. Int. J. Digital Earth 17 (1), 2383454. doi:10.1080/17538947.2024.2383454

CrossRef Full Text | Google Scholar

Nakyai, T., Santasnachok, M., Thetkathuek, A., and Phatrabuddha, N. (2025). Influence of meteorological factors on air pollution and health risks: a comparative analysis of industrial and urban areas in Chonburi province, Thailand. Environ. Adv. 19, 100608. doi:10.1016/j.envadv.2024.100608

CrossRef Full Text | Google Scholar

Pandey, B., Gautam, M., and Agrawal, M. (2018). “Greenhouse gas emissions from coal mining activities and their possible mitigation strategies,” in Environmental carbon footprints (Butterworth-Heinemann), 259–294. doi:10.1016/B978-0-12-812849-7.00010-6

CrossRef Full Text | Google Scholar

Qi, J., Chehbouni, A., Huete, A. R., Kerr, Y. H., and Sorooshian, S. (1994). A modified soil adjusted vegetation index. Remote Sensing Environment 48 (2), 119–126. doi:10.1016/0034-4257(94)90134-1

CrossRef Full Text | Google Scholar

Rafaj, P., Kiesewetter, G., Gül, T., Schöpp, W., Cofala, J., Klimont, Z., et al. (2018). Outlook for clean air in the context of sustainable development goals. Glob. Environ. Change 53, 1–11. doi:10.1016/j.gloenvcha.2018.08.008

CrossRef Full Text | Google Scholar

Rani, A., and Kumar, M. (2023). Seasonal changes in air pollutants and their relation to vegetation over the megacity Delhi national capital region. Environ. Sci. Proc. 27 (1), 16. doi:10.3390/ecas2023-15119

CrossRef Full Text | Google Scholar

Ren, J., and Xie, S. (2022). Diagnosing ozone-NOx-VOC sensitivity and revealing causes of ozone increases in China based on 2013–2021 satellite retrievals. Atmos. Chem. Phys. Discuss. 2022, 1–22. doi:10.5194/acp-22-15035-2022

CrossRef Full Text | Google Scholar

Rodríguez-Sánchez, A., Santiago, J. L., Vivanco, M. G., Sanchez, B., Rivas, E., Martilli, A., et al. (2024). How do meteorological conditions impact the effectiveness of various traffic measures on NOx concentrations in a real hot-spot? Sci. Total Environ. 954, 176667. doi:10.1016/j.scitotenv.2024.176667

PubMed Abstract | CrossRef Full Text | Google Scholar

Sabir, M. A., Nawaz, M. F., Khan, T. H., Zulfiqar, U., Haider, F. U., Rehman, A., et al. (2024). Investigating seasonal air quality variations consequent to the urban vegetation in the metropolis of faisalabad, Pakistan. Sci. Reports 14 (1), 452. doi:10.1038/s41598-023-47512-y

PubMed Abstract | CrossRef Full Text | Google Scholar

Savenets, M. (2021). Air pollution in Ukraine: a view from the Sentinel-5P satellite. Q. J. Hung. Meteorological Serv. 125 (2), 271–290. doi:10.28974/idojaras.2021.2.6

CrossRef Full Text | Google Scholar

Schwarz, K., Fragkias, M., Boone, C. G., Zhou, W., McHale, M., Grove, J. M., et al. (2015). Trees grow on money: urban tree canopy cover and environmental justice. PLoS ONE 10 (4), e0122051. doi:10.1371/journal.pone.0122051

PubMed Abstract | CrossRef Full Text | Google Scholar

Scorgie, Y., Fischer, T., and Watson, R. (2005). Air quality management plan for the ekurhuleni metropolitan municipality. Available online at: https://www.ekurhuleni.gov.za.

Google Scholar

Scott, G. M. (2010). Development of a methodology for the delineation of air quality management areas in South Africa. Westville, South Africa: University of KwaZulu-Natal.

Google Scholar

Shami, S., Ranjgar, B., Bian, J., Khoshlahjeh Azar, M., Moghimi, A., Amani, M., et al. (2022). Trends of CO and NO2 pollutants in iran during COVID-19 pandemic using timeseries Sentinel-5 images in google earth engine. Pollut. 2 (2), 156–171.

CrossRef Full Text | Google Scholar

Shelton, S., Liyanage, G., Jayasekara, S., Pushpawela, B., Rathnayake, U., Jayasundara, A., et al. (2022). Seasonal variability of air pollutants and their relationships to meteorological parameters in an urban environment. Adv. Meteorology 2022 (1), 5628911–5628918. doi:10.1155/2022/562891

CrossRef Full Text | Google Scholar

Smets, V., Wirion, C., Bauwens, W., Hermy, M., Somers, B., and Verbeiren, B. (2019). The importance of city trees for reducing net rainfall: comparing measurements and simulations. Hydrol. Earth Syst. Sci. 23, 3865–3884.

CrossRef Full Text | Google Scholar

Shikwambana, L., Mhangara, P., and Mbatha, N. (2020). Trend analysis and first time observations of sulphur dioxide and nitrogen dioxide in South Africa using TROPOMI/Sentinel-5 P data. Int. J. Appl. Earth Observation Geoinformation 91, 102130. doi:10.1016/j.jag.2020.102130

CrossRef Full Text | Google Scholar

Shikwambana, L., Kganyago, M., and Xulu, S. (2022). Analysis of wildfires and associated emissions during the recent strong ENSO phases in Southern Africa using multi-source remotely-derived products. Geocarto Int. 37 (27), 16654–16670. doi:10.1080/10106049.2022.2113449

CrossRef Full Text | Google Scholar

Shirinde, J., Wichmann, J., and Voyi, K. (2014). Association between wheeze and selected air pollution sources in an air pollution priority area in South Africa: a cross-sectional study. Environ. Health 13 (1), 32. doi:10.1186/1476-069X-13-32

PubMed Abstract | CrossRef Full Text | Google Scholar

Theys, N., De Smedt, I., Yu, H., Danckaert, T., van Gent, J., Hörmann, C., et al. (2017). Sulfur dioxide retrievals from TROPOMI onboard Sentinel-5 precursor: algorithm theoretical basis. Atmos. Meas. Tech. 10 (1), 119–153. doi:10.5194/amt-10-119-2017

CrossRef Full Text | Google Scholar

Tilstra, L. G., de Graaf, M., Wang, P., and Stammes, P. (2020). In-orbit Earth reflectance validation of TROPOMI on board the Sentinel-5 precursor satellite. Atmos. Meas. Tech. Discuss. 2020, 1–27. doi:10.5194/amt-13-4479-2020

CrossRef Full Text | Google Scholar

UN (2015). Department of economic and social affairssustainable development. Available online at: https://sdgs.un.org/goals.

Google Scholar

Valjarević, A. (2024). Long-term remote sensing-based methods for monitoring air pollution and cloud cover in the Balkan countries. Environ. Sci. Pollut. Res. 31, 27155–27171. doi:10.1007/s11356-024-32982-y

PubMed Abstract | CrossRef Full Text | Google Scholar

Wan, Z., Zhang, Y., Zhang, Q., and Li, Z. L. (2002). Validation of the land-surface temperature products retrieved from terra moderate resolution imaging spectroradiometer data. Remote Sensing Environ. 83 (1-2), 163–180. doi:10.1016/S0034-4257(02)00093-7

CrossRef Full Text | Google Scholar

World Health Organization (2016b). Urban green spaces and health. Available online at: https://iris.who.int/bitstream/handle/10665/345751/WHO-EURO-2016-3352-43111-60341-eng.pdf?sequence=3 (Accessed September 30, 2025).

Google Scholar

Wu, Z., Shi, Y., Ren, L., and Hang, J. (2025). Scaled outdoor experiments to assess impacts of tree evapotranspiration and shading on microclimates and energy fluxes in 2D street canyons. EGU General Assem. doi:10.5194/egusphere-egu25-5456

CrossRef Full Text | Google Scholar

Xu, W. Y., Zhao, C. S., Ran, L., Deng, Z. Z., Liu, P. F., Ma, N., et al. (2011). Characteristics of pollutants and their correlation to meteorological conditions at a suburban site in the north China plain. Atmos. Chem. Phys. 11 (9), 4353–4369. doi:10.5194/acp-11-4353-2011

CrossRef Full Text | Google Scholar

Zhang, H., Wang, Y., Hu, J., Ying, Q., and Hu, X. M. (2015). Relationships between meteorological parameters and criteria air pollutants in three megacities in China. Environ. Research 140, 242–254. doi:10.1016/j.envres.2015.04.004

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhou, L. (2019). Correlations of stormwater runoff and quality: urban pavement and property value by land use at the parcel level in a small sized American city. Water 11 (11), 2369. doi:10.3390/w11112369

CrossRef Full Text | Google Scholar

Zunckel, M. (1999). Dry deposition of sulphur over eastern South Africa. Atmos. Environ. 33 (21), 3515–3529. doi:10.1016/S1352-2310(98)00226-X

CrossRef Full Text | Google Scholar

Zunckel, M., Robertson, L., Tyson, P. D., and Rodhe, H. (2000). Modelled transport and deposition of sulphur over Southern Africa. Atmos. Environ. 34 (17), 2797–2808. doi:10.1016/S1352-2310(99)00495-1

CrossRef Full Text | Google Scholar