Application of GIS-Based AHP Model for the Impact Assessment of COVID-19 Lockdown on Environment Quality: The Case of Kabul City, Afghanistan

Abstract

The COVID-19 pandemic has negatively impacted the industrial, financial, and social aspects of our daily life due to the implementation of lockdown to protect against the spread of the virus. In addition, the lockdown deduced by COVID-19 has promising positive impacts on air quality and environmental pollution. This study aims to monitor the effects of lockdown on environmental degradation during the pandemic in Kabul city, the capital of Afghanistan, using geospatial data and a statistical model of the Analytical Hierarchy Process (AHP). To achieve the purpose of the study, the most essential influencing factors on air quality were generated from different sources using Google Earth Engine (GEE) and GIS environment; Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Normalized Difference Moisture Index NDMI) were calculated using Sentinel-2MSI, Carbon Monoxide (CO) was obtained from Sentinel-5P TROPOMI, and land surface temperature was retrieved from MODIS data. The generated thematic layers (before COVID-19, and during a lockdown of COVID-19) were weighted and rated using the AHP analysis. The weighted layers were spatially overlayed to obtain the final output. Consequently, the environmental quality degradation maps before and during COVID-19 were generated to assess the differences over the 22 districts of Kabul city. The findings of the study show that Kabul city is covered by the very low, low, moderate, high, and very high degradation of the environment by 3.17%, 5.33%, 20.54%, 26.63%, 44.32% before COVID-19 in 201,9 respectively, while the percentages are changed to 4.37%, 8.99%. 16.55%, 37.47%, and 32.62% during the lockdown caused by COVID-19 in 2020. The changes in the percentage of environmental degradation in Kabul city particularly in high and very high zones confirm the positive impact of the lockdown of COVID-19.

Introduction

The coronavirus (COVID-19) introduced as Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-Cov-2) spread rapidly all over the world after originating in Wuhan, Hubei province, China, at the end of December 2019 (Hashim et al., 2020; Stratoulias & Nuthammachot, 2020; Wang et al., 2020). Afterward, it was declared a pandemic officially by the World Health Organization (WHO). SARS-CoV-2 is an RNA virus with a high mutation rate and can be more efficiently transmitted (Stratoulias & Nuthammachot, 2020). On January 30, 2020, WHO announced public health emergency worldwide, and in February, the pandemic outbreaks started in Iran, Italy, and other countries around the globe (Hashim et al., 2020). Consequently, the first COVID-19 case in Afghanistan was acquired from Iran when a 35-year-old male Afghan shopkeeper visited Qom, Iran, and returned to Herat, West of Afghanistan, on February 15, 2020. Officially, the first case of COVID-19 was registered on February 24, 2020 (Khudadad et al., 2021). Since then, he visited his family and friends without any precautions (Mousavi et al., 2020). After that, the virus began to spread around Kabul in late February, and a lockdown was declared in Kabul city between March 28 and April 1. The spread of the virus may be contained by maintaining proper social distancing, personal hygiene, avoiding gatherings, and visiting places like hospitals, meetings, and public transportation, which have a high risk of virus contamination (Bherwani et al., 2020; Gautam et al., 2020; WHO, 2020).

As of July 04, 2021, there have been more than 182 million confirmed cases and around 3,954,324 deaths have been reported globally (WHO, 2020). The first confirmed case of the virus was reported on 24 February 2020 in Afghanistan. The lockdown in Kabul started partially with the closing of schools and universities, followed by the total lockdown. The total number of confirmed cases as of July 04, 2021 is 120,216 with the death of 4962 in Afghanistan (WHO, 2020).

The air in the environment must be safe and clean for the survival of all living beings. Anthropogenic activities are considered a major cause of air pollution due to the emission of many harmful pollutants in high concentrations (Bherwani et al., 2020; El Ghoraiby et al., 2020; Gautam et al., 2020). According to Kaplan et al. (2019), economic development, urbanization, energy consumption, transportation and motorization, and the rapid increase in the urban population are the main causes of air pollution. Particulate matter (PM), e.g., sulfur dioxide (SO₂), nitrogen dioxide (NO₂), ozone (O₃), carbon monoxide (CO), and carbon dioxide (CO₂), are considered the biggest air pollutants in our daily life (Chen et al., 2007).

Air and environmental quality changes caused by the COVID-19 lockdown dramatically have become the topic of recent research studies. The utilization of spatial and non-spatial data is widely used for environmental quality assessment (Chakraborty et al., 2022; Danek & Zaręba, 2021; Faisal et al., 2022; Wei et al., 2021; Yang et al., 2021; Zhang et al., 2022) Sentinel-5P TROPOMI from European Space Agency (ESA) have effectively been used as spatial data for the extraction of NO₂ concentration to assess the quality of the environment before, during, and after COVID-19 lockdown (Hashim et al., 2020; Sarfraz et al., 2020). Hashim et al. (2020) assessed the air quality changes deduced by the COVID-19 lockdown in Baghdad, Iraq by the analysis of NO₂, O₃, PM_2.5, and PM₁₀ concentrations. The authors extracted the NO₂ tropospheric column from the Sentinel-5P satellite. The results were achieved before the pandemic, during the partial and total lockdown, and after the lockdown. The decrease in air pollutants is concluded during the lockdown in this study.

Sentinel-5P and Landsat 8 OLI/TIRS are the satellites that have widely been used for the air quality and environmental quality assessment impacted by COVID-19 (Aman et al., 2020; Ayoobi et al., 2021; Bera et al., 2020; Ghosh et al., 2020; Maithani et al., 2020). In the literature, Sentinel-5P was utilized for the extraction of NO₂ over the selected areas pre, during, and post COVID-19 periods, while, Landsat 8 OLI/TIRS imageries were used for the determination of land surface temperature. The LST data were subjected to analysis of air and environment quality.

In addition to spatial data, nonspatial data as station-wise demonstrated in a conventional study for the analysis of pollutant concentration before, during, and after lockdown caused by COVID-19 (Aydın et al., 2020; Gautam et al., 2020; Rodríguez-Urrego & Rodríguez-Urrego, 2020; Sahoo et al., 2020; Sarkar et al., 2020).

Analytical Hierarchy Process (AHP), a robust statistical decision-making model, has been used for the susceptibility, vulnerability, and spreading the risk of COVID-19 by considering the influencing factors, e.g., percentage of the number of hospitals per population (NHP), Monetary Poverty (MP), and Vulnerable Population (VP) (Badillo Rivera et al., 2020; Mahato et al., 2020). In the case of environmental and air quality, several influencing factors based on the degree of their importance should be considered. Therefore, AHP could be a powerful statistical model to weigh each factor and decide its degree of importance.

 A diversity of studies concerning the impacts of COVID-19 in Afghanistan, particularly in the capital, have been carried out; however, the conducted studies focus on the non-spatial analysis, which deals with the social, economic, and health impacts (Khudadad et al., 2021; Lucero-Prisno et al., 2020; Mousavi et al., 2020, 2021; Mushkani & Ono, 2020; Shah et al., 2020). In this study, Google Earth Engine (GEE) was used to process and generate all thematic layers. GEE is a cloud-based platform that processes big geospatial data and enables parallelized geospatial data processing on a global scale using Google’s cloud (Gorelick et al., 2017). Cloud free platform of GEE hosts a petabyte-scale of 40 years of remotely-sensed data, e.g., Landsat series, MODIS, National Oceanographic and Atmospheric Administration Advanced Very High-Resolution Radiometer (NOAA AVHRR), Sentinel 1, 2, 3, and 5-P, and Advanced Land Observing Satellite (ALOS) data (Amani et al., 2020; Tamiminia et al., 2020). Despite hosting a large repository of raw remotely-sensed imagery, GEE provides users to have access to preprocessed, cloud-removed, and mosaicked images (Gorelick et al., 2017; Tamiminia et al., 2020). This study is the first to be conducted over Kabul city and is based on a spatial analysis carried out with the use of the AHP model by taking the most relevant influencing factors of air quality into account. The novelty of the methodology and the findings of this study will scientifically contribute to the monitoring and planning of air quality management. Consequently, the main objectives of this study are: (1) to assess the advantages and disadvantages of the lockdown caused by COVID-19 in Kabul city, the capital of Afghanistan using geospatial data and a statistical model, and (2) to generate the geospatial datasets over the two periods (before and during a lockdown of COVID-19) using Sentinel-2MSI, Sentinel-5P, and MODIS sensors, and (3) to find out the capability of GIS-based AHP model for the assessment of air quality and environmental pollution issues.

Study Area

Kabul, the capital of Afghanistan, is located in the east part of Afghanistan between the latitude 34° 31′ 31″ and longitude 69° 10′ 42″ (Fig. 1). Most of the trade, business, political, and administration  related to the whole country is conducted in Kabul city (Ayoobi et al., 2021). Based on UN-Habitat (2015) and Wafa et al. (2020), Kabul city has 22 districts with a total land area of 1049 km² and 396,095 number residential houses. The geomorphology of Kabul city is made up of about 56% mountains and rough terrain, and 38% flat (Ayoobi et al., 2021). During the months of February, March, and April, the rain has the highest trend. Recently, 400 mm of annual rainfall has been reported in Kabul city as the highest precipitation (Mehrad, 2020). The climate in the city is arid and semi-arid with cold winters as low as – 10 °C and hot summers as high as 40 °C (Qutbudin et al., 2019; Wafa et al., 2020). The urban development within Kabul city is considered over the valley floor of the Kabul River with an average altitude of 1800 m above sea level (Ahmadi & Kajita, 2017; Kazimee, 1977). Recently, Kabul city has been experiencing rapid population growth from 720.000 in 1978 to about 4.9 million in 2015, and an increase of 8 million is predicted by 2050 (Ahmadi & Kajita, 2017; Wafa et al., 2020).

Fig. 1.

The geographic location of the study area

The COVID-19 pandemic as a global public health threat was first reported in Herat city in the Western part of Afghanistan in the mid of February 2020. As of June 11, 2020, the pandemic spread all over the 34 provinces of Afghanistan with 22,890 confirmed cases and 426 death, while 3326 recovered cases (Lucero-Prisno et al., 2020; Mousavi et al., 2020). Since February 24, 2020, of the total number of cases, 73% have been reported only from 6 provinces (Kabul, Nangarhar, Kandahar, Herat, and Balkh) (Khudadad et al., 2021).

Materials and Methods

In this study, multispectral data for the generation of geospatial datasets from three different sensors, e.g., Sentinel 2A MSI, Sentinel-5P TROPOMI, and MODIS Terra were used. The Sentinel 2A satellite was launched on June 23, 2015, from the spaceport in Kourou, French Guiana (Corporation, 2017; Zhang et al., 2018). This satellite is the first earth observation satellite in the European Copernicus program, which carries the Multi-Spectral Instrument (MSI) (Zhang et al., 2018). The satellite contains 13 spectral bands, with a 290 km swath width, and 5 days revisit time (Corporation, 2017; Hu et al., 2019; Zylshal et al., 2017) (Table 1).

Table 1.

Description of satellites Sentinel-2A MSI, Sentinel-5P TROPOMI, and MODIS Terra

Sentinel-2A MSI			Sentinel-5P TROPOMI			MODIS Terra
Band no.	Wavelength (µm)	Spatial resolution (m)	Band no.	Wavelength (µm)	Spatial resolution	Band no.	Wavelength (µm)	Spatial resolution
B1	0.433–0.453	60	B1	0.270–0.320	7 × 3.5 km2	B1	0.620–0.670	250 m
B2	0.458–0.523	10	B2	0.310–0.500		B2	0.841–0.867	250 m
B3	0.543–0.578	10	B3	0.627–0.775		B3	0.459–0.479	250 m
B4	0.650–0.680	10	B4	2.305–2.385		B4	0.545–0.565	500 m
B5	0.698–0.713	20				B5	1.230–1.250	500 m
B6	0.733–0.748	20				B6	1.628–1.652	500 m
B7	0.765–0.785	20				B7	2.105–2.155	500 m
B8	0.785–0.900	10				B8–19	0.405–0.965	1 km
B8A	0.855–0.875	20				B20	3.660–3.840	1 km
B9	0.930–0.950	60				B21	3.929–3.989	1 km
B10	1.365–1.385	60				B22	3.929–3.989	1 km
B11	1.565–1.655	20				B23	4.020–4.080	1 km
B12	2.100–2.280	20				B24–36	4.433–14.385	1 km

The Copernicus Sentinel-5P through Tropospheric Monitoring Instruments (TROPOMI) is a part of the Sentinel Fleet mission developed by the European Space Agency (ESA) to monitor the European environment and air pollutants around the globe. The satellite was launched on October 17 of 2017, and targeted to measure the pollutants, e.g., carbon monoxide, nitrogen dioxide, methane, formaldehyde, aerosol, sulfur dioxide, and ozone (Gibson et al., 2019; Safarianzengir et al., 2020; Siddiqui et al., 2020). Sentinel-5P TROPOMI  has four spectral bands, as described in Table 1, with coverage of ~ 2600 km swath width (ESA, 2020).

The Moderate Resolution Imaging Spectroradiometer (MODIS) Terra satellite was launched on December 18, 1999, and helps to improve our understanding of global environmental processes and dynamics occurring on the land, in the ocean, and in the lower atmosphere (LAADS DAAC, 2020). The satellite images the entire earth’s surface every 1–2 days, with 2330 km (cross-track) and 10 km (along-track at nadir) swath dimensions (Ba et al., 2019; LAADS DAAC, 2020).

The multispectral data were downloaded and processed for two distinct time intervals using; the lockdown deduced by COVID-19 in 2020 and before COVID-19 in 2019. March 28 to April 28 due to being a short-term lockdown within the Kabul city in 2020. Consequently, the same interval was chosen before COVID-19. The Cloud Computing Platform of Google Earth Engine (GEE) was utilized for processing and downloading all geospatial datasets. ArcGIS 10.7.1 environment was applied for the preparation and plotting of thematic layers.

Analytical Hierarchy Process (AHP) model was applied for weighting and ranking each influencing factor to assess the environmental degradation of Kabul city during the lockdown and before the COVID-19. In this model, we considered five influencing factors as the criteria, e.g., NDVI, NDWI, NDMI, CO, and LST (Land Surface Temperature), as illustrated in (Fig. 2).

Fig. 2.

Flow chart illustrating the methodology processes

Generation of Geospatial Datasets

Since the environmental quality is influenced mainly by the spatial, physical, social, and economic aspects of a city, five critical datasets, e.g., NDVI, NDWI, NDMI, LST, and CO, were considered in this study. All the geospatial layers were generated using open-source datasets. Among these, NDVI, NDWI, and NDMI are stable; however, these factors may also change during the lockdown deduced by COVID-19 within Kabul city. Considering the literature, one of the most influential  factors on the air and environmental quality is carbon monoxide. Due to the lack of real-time stations for the measurement of CO to correspond to spatiality distribution over the study area, the thematic layer of this factor was generated using satellite-derived image data. Carbon monoxide concentration during the lockdown and before the COVID-19 pandemic was extracted from Sentinel-5P through the GEE. The obtained results were exported to GIS environmental for spatial analysis and plotting. The thematic layers of CO were resampled and reclassified into five classes following the Jenk classification scheme. The environment quality is directly related to the amount of CO concentration.

LST is another factor for environmental health and quality, which depends on several atmospheric and physical factors e.g., the month of study, cloudy condition, land use/land cover (LULC) pattern, and many more (Ghosh et al., 2020). MODIS–Terra was selected for the estimation of LST using GEE. The same time for pre-COVID-19 in 2019, and during the lockdown of COVID-19 was considered by taking the average of values. The estimated raster in Kelvin was exported to the GIS environment for spatial analysis purposes and unit conversion.

Biophysical indices as the essential dealer of nature are calculated to quantify environmental phenomena e.g., greenness, water bodies distribution, moisture, and many more. In this study, NDVI, NDWI, and NDMI as biophysical indices are considered to assess the impacts of lockdown on the environment quality. These indices were calculated using Sentinel-2MSI by the GEE platform and exported to the GIS environment for further processes. NDVI indicates the degree of greenness over an area. The positive values of NDVI characterize a healthy environment condition.

Analytical Hierarchy Process (AHP)

AHP is an essential way to compute the weights of each parameter for achieving the goal in the decision of complex problems (Ahmadi et al., 2021). The model is experts’ knowledge-based and was first introduced by Saaty (Razandi et al., 2015; Saaty, 1990). Comparing the influencing parameters based on their relative importance to the target decision by a pairwise comparison matrix is considered the early stage of the AHP model (Ghosh et al., 2020; Şener et al., 2018). The other stages of this model are the computation of normalized weights, calculation of consistency ratio, and final decision of making steps (Ahmadi et al., 2021; Kumar & Krishna, 2018). In this study, five influencing factors associated with the environmental quality index are considered, which are valued based on (Saaty, 1977) scale from 1 (equal significance) to 9 (extreme significance) by making a pairwise matrix as shown in (Table 2).

Table 2.

Pairwise comparison matrix of influencing parameters

Parameters	Parameters
	CO	LST	NDVI	NDWI	NDMI
CO	1.000	3.000	5.000	7.000	9.000
LST	0.333	1.000	5.000	7.000	9.000
NDVI	0.200	0.200	1.000	3.000	5.000
NDWI	0.143	0.143	0.333	1.000	3.000
NDMI	0.111	0.111	0.200	0.333	1.000
Sum	1.787	4.454	11.533	18.333	27.000

Once the comparison matrix is created, a normalized pairwise comparison for normalized weights is calculated. The normalized value of each cell is computed by dividing each cell by the total of each column, while the normalized weights are calculated by taking the average of each normalized row corresponding to a parameter as shown in (Table 3).

Table 3.

Normalized pairwise matrix and computed weights for each parameter

Parameters	Parameters
	CO	LST	NDVI	NDWI	NDMI	Weights
CO	0.5595	0.6736	0.4335	0.3818	0.3333	0.476
LST	0.1865	0.2245	0.4335	0.3818	0.3333	0.312
NDVI	0.1119	0.0449	0.0867	0.1636	0.1852	0.118
NDWI	0.0799	0.0321	0.0289	0.0545	0.1111	0.061
NDMI	0.0622	0.0249	0.0173	0.0182	0.0370	0.032
Sum	1.0000	1.0000	1.0000	1.0000	1.0000	1.000

To test the consistency of the model, after the computation of normalized weights, the consistency ratio (CR) using the following equation was calculated:

$$\mathrm{CR}=\frac{\mathrm{CI}}{\mathrm{RI}}$$

CR is the consistency ratio; RI is the random index taken from the table value proposed by (Saaty, 1990). This index depends on the number of parameters taken 1.12 in this study. CI is the consistency index which was computed using the following equation:

$$\mathrm{CI}=\frac{\lambda \max -n}{n-1}$$

where $\lambda$ is the largest eigenvalue of the matrix is determined based on the created matrix, and n is the number of parameters considered in this study. Based on (Malczewski, 1999; Saaty, 1990), the model's consistency ratio (CR) must be obtained less than 0.1. If the result of the ratio is greater than 0.1, then the pairwise matrix of the considered factors must be reviewed. In this study, the ratio was obtained at 0.0811, which is evidence of the consistency of the model.

The considered parameters were classified into five classes, and following the importance of each sub-class, they were ranked as shown in (Table 4). The importance of each class and sub-class was provided from literature and interpreted as their impacts on environmental degradation. Increasing values have a positive effect on environmental degradation and therefore, the ranks of subclasses were assigned accordingly.

Table 4.

Assigned weights and rates of all considered parameters

No	Parameters	Classes		Rates	Normalized rates	Weights
		Pre-COVID	During lockdown
1	CO	22,372.621–24,915.253	24,443.482–25,896.579	1	0.056	0.476
		24,915.253–27,143.981	25,896.579–27,083.616	2	0.111
		27,143.981–28,148.478	27,083.616–28,004.593	3	0.167
		28,148.478–29,027.413	28,004.593–28,782.307	5	0.278
		29,027.413–30,377.205	28,782.307–29,662.352	7	0.389
2	LST	0–15	0–12	1	0.056	0.312
		15–21	12–18	2	0.111
		21–25	18–21	3	0.167
		25–27	21–23	5	0.278
		27–31	23–26	7	0.389
3	NDVI	− 0.404–0.051	− 0.449–0.085	1	0.056	0.118
		0.051–0.126	0.085–0.175	2	0.111
		0.126–0.249	0.175–0.305	3	0.167
		0.249–0.441	0.305–0.494	5	0.278
		0.441–0.793	0.494–0.824	7	0.389
4	NDWI	− 0.710 to − 0.299	− 0.788 to − 0.310	1	0.056	0.062
		− 0.299 to − 0.208	− 0.310 to − 0.222	2	0.111
		− 0.208to − 0.045	− 0.222 to − 0.065	3	0.167
		− 0.045–0.313	− 0.065–0.338	5	0.278
		0.313–0.952	0.338–0.951	7	0.389
5	NDMI	− 0.696 to − 0.095	− 0.650 to − 0.079	1	0.056	0.032
		− 0.095–0.002	− 0.079–0.022	2	0.111
		0.002–0.163	0.022–0.198	3	0.167
		0.163–0.525	0.198–0.543	5	0.278
		0.525–0.951	0.543–0.952	7	0.389

The environmental quality index (EQI) is calculated using the following equation based on the weighted linear combination (WLC) approach in which the result is dimensionless (Ahmadi et al., 2021; Ghosh et al., 2020; Razandi et al., 2015).

$$\mathrm{EQI}=\sum _{i=1}^n{R}_i\times {\omega }_i$$

where n is the number of considered parameters, $R_i$ is the rate of the i parameter, and ${\omega }_i$ is the weights of i parameter.

Results and Discussion

The rapid population growth, anthropogenic activities, industrialization, and more specifically some other environmental phenomena like water quality, vegetation quality and distribution, air quality, climate, and land surface temperature have all harmed the quality of the environment in recent years. COVID-19 has had good benefits on the rehabilitation of environmental degradation, despite its detrimental consequences on many facets of life. (Ghosh et al., 2020). Due to the COVID-19 outbreak that was spreading over the world, especially in Kabul city, there was a temporary lockdown that limited industrial production as well as human and vehicular activity. This restriction caused notable improvements in the environment's physical characteristics, such as CO, LST, NDVI, NDWI, and NDMI. In this study, the impact of the lockdown period on the environmental quality of Kabul city was evaluated using the Multiple Criteria Decision-Making (MCDM) model of AHP.

Typical software downloads and processing of all influencing elements raise several issues, including acquisition, storage, searching, sharing, transferring, analysis, and visualization throughout two-time intervals. To get around these issues, we used GEE to provide all of the theme layers. This platform made it possible to do high-speed analysis for a specific portion of the study area using various processing tools without having to look for and download the necessary data separately. We were able to determine the biophysical layer's spectral indices and download raster-based atmospheric emissions using GEE. To convert data and handle plotting tasks, the final outputs were moved to the ArcMap environment.

Following the obtained results, changes in spatial distribution and value variations in CO concentration are observed within the two intervals. Before COVID-19 in 2019, the maximum concentration of CO is 30,377,205 µmol/m² over the central districts, while the minimum concentration is 22,372,621 µmol/m² around the city. However, during the lockdown of COVID-19 in 2020, decreasing the maximum value (max 29,662,351 µmol/m², min 24,443,482 µmol/m²) and accumulation of concentration is seen. The spatial distribution with the highest concentration before COVID-19 is seen over the central and eastern parts of Kabul city, whereas this distribution was shrunk to only the central parts during the lockdown (Fig. 3a, b). The final thematic maps of LST were converted to degrees Celsius, resampled, and reclassified into five classes (Fig. 3c, d). Although the normal trend of temperature during the time is increasing due to the LULC changes and other global phenomena, according to the obtained results of LST in this study, the amount of temperature in 2019 before COVID-19 is higher than the time during the lockdown in 2020. As seen, the maximum amount of temperature (31 °C) was decreased to 26 °C from pre-COVID-19 to during lockdown, which is clear evidence of decreasing physical activities resulting from the lockdown.

Fig. 3.

Thematic geospatial datasets during the same seasons; Pre-COVID-19 in 2019 (left) and during lockdown 2020 (right); a, b CO, c, d LST, e, f NDVI, g, h NDWI, and i, j NDMI

According to (Fig. 3e, f), the distribution of greenness by NDVI has increased from 2019 to lockdown 2020, showing the positive impacts of the lockdown. This indicates that during the lockdown, anthropogenic activities were restricted, which can be considered an essential disturbance of the greenness. The changes are seen in terms of values and spatial distribution that areas around the city center have been covered by naturalness during the lockdown. The volume of waterbody or wetness estimated by the NDWI has a fundamental role in environmental quality conditions. Positive values of NDWI correspond to the favorable condition of ecological quality. Except for physical activities, the waterbody is influenced by climatic conditions, seasonal variation, rainfall, etc. Therefore, NDWI has a notable change from pre-COVID-19, 2019, to during the lockdown of 2020. According to the result, the value has been between (− 0.710 to 0.952) in 2019, while it has decreased between (− 0.788 to 0.951) within the study area (Fig. 3g, h). Anthropogenic activities can indirectly impact the moisture content of the surface as fewer human activities result in a high amount of moisture and vice versa. NDMI defines the moisture contents on the earth’s surface. Following the estimation of this index for two intervals of the same season over the study areas, it indicates that from 2019 (Before COVID-19) to 2020 (during lockdown), the NDMI value has increased. This variation shows a positive change in environmental quality (Fig. 3i, j).

Variations in the spatiality and value intensity of environmental degradation are shown when taking into account the findings of the overlay analysis utilizing the normalized weights and rates of each contributing component, as indicated in (Fig. 4). The results demonstrate that the central areas of Kabul city, such as D1, D2, D3, D4, D9, D10, D15, D16, and D18, are characterized by high and very high environmental deterioration during pre-COVID-19. However, during the lockdown, it was noted that the high and very high degradation varied in the location from the center and northern regions to the most northern and western sides. For example, before COVID-19, D7, D8, D12, and D16 are characterized by high and very high degradation, but during the lockdown, degradation has altered to high and moderate in these districts, as seen in (Figs. 4, 5). The differential results before COVID-19 and during the lockdown of COVID-19 describe the environmental degradation extended in the study area depicted in (Fig. 5). D17, D11, D5, and partially D20 are mostly characterized by high negative values which represent the negative extent of environmental degradation.

Fig. 4.

The spatial distribution and intensity of environmental quality degradation over the Kabul city; a before COVID19 and b during the lockdown

Fig. 5.

Changes in environmental degradation from pre-COVID in 2019 to during lockdown in 2020. The negative values are corresponding to increasing in degradation, while the positive values representing decreasing in degradation

Spatial statistics show that the lockdown over Kabul has had a positive impact on the high and very high environmental quality degradation, decreasing it by 0.52 and 20.19%, respectively, while having a negative impact on the very low, low, and moderate degradation, increasing it by 21.91, 6.32, and 6.88%, as shown in (Fig. 6). The reason behind this negative result is the spatial variation of degradation over the regions where around Kabul city. These regions are the highways that connect the capital to northern, southern, eastern, and western provinces and also the Kabul International Airport where their activities were not completely suspended during the COVID-19.

Fig. 6.

Diagram showing the percentage of each class coverage within the Kabul city before COVID-19 and during lockdown deduced by the COVID-19

Conclusion

This study investigated the reduction of the biophysical phenomenon and air emissions resulting from the lockdown of COVID-19 in Kabul city, the capital of Afghanistan. The multi-criteria decision-making analysis of AHP was applied by considering the five most effective influencing factors of environmental quality, namely CO, LST, NDVI, NDWI, and NDMI to highlight the environmental degradation. To find out the capability of the AHP model, two intervals of time were selected over a case study: before COVID-19 in 2019 and at the same time exactly during the lockdown of COVID-19 in 2020. It was found that the AHP model has a high capability in environmental quality index vulnerability mapping. Furthermore, evidence of the positive impacts of the lockdown due to COVID-19 on the environmental quality was revealed. The highest environmental degradation caused by atmospheric emission, land surface temperature, and the biophysical phenomenon is observed within the central and northern parts of Kabul city. In contrast, this degradation is seen to be decreased in terms of intensity and spatial distribution during the lockdown due to COVID-19. It is concluded that decreasing anthropogenic activities, factory shutdowns, and decreasing vehicle activity play an essential role in environmental quality. Statistics show that 3.17%, 5.33%, 20.54%, 26.63%, and 44.32% of regions in Kabul city were covered by very low, low, moderate, high, and very high degradation of environment respectively, before COVID-19 in 2019, while these percentages were changed during the lockdown by 4.37%, 8.99%, 16.55%, 37.47%, and 32.62%. The obtained results can be a significant clue for the environmental relevant departments for better monitoring and planning of environmental protection. Furthermore, the applied methodology can scientifically be used in other regions of the world with effective results.

Availability of Data and Materials

Not applicable.

Code Availability

Not applicable.

Declarations

Conflict of interest

The authors declare no competing interests.

Consent for Publication

All authors agree to the publication in this journal.