Spatiotemporal Evolution and Influencing Factors of Aerosol Optical Depth in Zhejiang Province: Insights from Land Use Dynamics and Transportation Networks Based on Remote Sensing

Abstract

Aerosol optical depth (AOD) serves as a critical indicator for atmospheric aerosol monitoring and air quality assessment, and quantifies the radiative attenuation caused by airborne particulate matter. This study uses MODIS remote sensing imagery together with land use transition datasets (2000–2020) and road network density metrics (2014–2020), to investigate the spatiotemporal evolution of AOD in Zhejiang Province and its synergistic correlations with urbanization patterns and transportation infrastructure. By integrating MODIS_1KM AOD product, grid-based road network density mapping, land use dynamic degree modeling, and transfer matrix analysis, this study systematically evaluates the interdependencies among aerosol loading, impervious surface expansion, and transportation network intensification. The results indicate that during the study period (2000–2020), the provincial AOD level shows a significant declining trend, with obvious spatial heterogeneity: the AOD values in eastern coastal industrial zones and urban agglomerations continue to increase, with lower values dominating southwestern forested highlands. Meanwhile, statistical analyses confirm highly positive correlations between AOD, impervious surface coverage, and road network density, emphasizing the dominant role of anthropogenic activities in aerosol accumulation. These findings provide actionable insights for enhancing land-use zoning, minimizing vehicular emissions, and developing spatially targeted air quality management strategies in rapidly urbanizing regions. This study provides a solid scientific foundation for advancing environmental sustainability by supporting policy development that balances urban expansion and air quality. It contributes to building more sustainable and resilient cities in Zhejiang Province.

1. Introduction

Aerosol optical depth (AOD) is the integral of the extinction coefficient of aerosol along the radiation propagation path in the vertical direction. It is a critical factor in describing aerosol properties, assessing aerosol content, evaluating the degree of atmospheric environmental pollution and aerosol radiation climate effects, and correcting the atmospheric effects of space remote sensing. The optical depth of aerosols is mainly influenced by atmospheric suspended solids, and different land use types have different impacts on atmospheric suspended solids, so land use is one of the influencing factors of atmospheric AOD. The longitudinal comparison of data analysis by the Ministry of Environmental Protection of China in recent years [1] indicates that although China’s air pollution has been decreasing year by year, the type of pollution has gradually evolved from soot to a mixture of soot and motor vehicle exhaust. The road network is the fundamental support for high-quality integration, and there exists a complex interrelationship between it and aerosol emissions that affects aerosol emissions in multiple aspects, such as regional land use layout [2], land development and utilization [3], residents’ travel patterns [4], and regional industrial structure [5].

Previously, great progress has been achieved in studying remote sensing of AOD. Since the mid-1970s, international research on the inversion of AOD using satellite data has been gradually conducted. Early research included the application of satellite sensors such as TOMS, OMI, POLDER, ATSR-2, MODIS, GOES, etc. [6,7,8,9,10,11,12], which focused on deriving AOD over land. NOAA has been using AVHRR Visible Channel 1 for remote sensing of AOD over the ocean since 1977 and has evolved to utilize dual-channel inversion as NOAA’s operational tool to provide weekly maps of global offshore AOD distributions. In the 21st century, the application of MODIS data has further facilitated the inversion of AOD in the ocean and land. For example, Chu et al. [13] and Remer et al. [14] developed a method to invert AOD using MODIS data, demonstrating significant application value in global and local air pollution detection. Levy et al. [15] proposed a new V5.2 algorithm based on the reflectance of the visible and mid-infrared channels of satellites, leading to higher accuracy in the inversion of aerosol optical characteristic parameters. Domestic aerosol remote sensing research began in the mid-1980s. In 1986, Zhao et al. [16] pioneered the remote sensing research of atmospheric aerosols in China by using AVHRR data to conduct remote sensing inversion of aerosols over the Bohai Sea. Since then, domestic scholars have made great achievements in the inversion of AOD using various satellite sensors and algorithms. For instance, Mao et al. [17] compared the results of MODIS aerosol products with terrestrial multi-band photometers to verify their accuracy and regional distribution reflection ability. Liu Guiqing et al. [18] investigated the relationship between the urban air pollution index and MODIS AOD in the Yangtze River Delta region, and they proposed that MODIS AOD can reflect the status of surface air pollution.

In recent years, with the increase in the number of ground-based observation sensors and the development of sensor combinations on different platforms, further progress has been achieved in aerosol remote sensing research both domestically and internationally. For instance, Tang et al. [19] designed a collaborative inversion algorithm based on Terra and Aqua binary MODIS data, which resolved the problem of difficult separation of aerosol information in high-brightness areas and achieved simultaneous inversion of surface albedo and aerosol optical depth. Currently, Bai’s team devised a multimodal tensor completion model that integrates geographic big data and deep learning, significantly improving the accuracy of AOD missing data reconstruction in complex scenarios such as cloud pollution. While ensuring the spatiotemporal consistency of global AOD, the reconstruction efficiency of large-scale missing regions is 2.3 times higher than that of the traditional method, effectively solving the problem of blind spots caused by long-term data loss and providing reliable technical support for the dynamic monitoring of aerosols at the global scale [20]. Youn et al. introduced a spatial gap-filling method to process Himawari-8/Advanced Himawari Imager (AHI) hourly AOD data using a random forest (RF) model that integrated meteorological variables and model-based AOD data. The model has been proven to be effective as it can increase data coverage from 6% to 100% [21]. Singh et al. conducted high-resolution AOD inversion based on Landsat 8 images and successfully obtained AOD products with a spatial resolution of 30 m through the development and application of the SEMARA algorithm [22].

In 2023, the State Council issued the Action Plan for Continuous Improvement of Air Quality, aiming to improve air quality, reduce heavy pollution from weather events, solve prominent atmospheric environmental problems, and reduce the concentration of fine particulate matter (PM2.5). This study investigates the temporal and spatial variation characteristics, land use changes, and road network density changes of atmospheric AOD in Zhejiang Province in the past 20 years, analyzes the correlation between them, and provides scientific suggestions for resolving atmospheric environment problems, in order to help improve the air quality in Zhejiang Province. By quantifying the relationships between AOD changes, land use transitions and transportation network growth, this study clarifies the links between urbanization and environmental sustainability. The findings offer practical evidence to support the integration of air quality considerations into sustainable spatial planning and infrastructure development.

2. Materials and Methods

2.1. Study Area

Zhejiang Province is located on the southeast coast of China, with a longitude ranging from 118°01′ to 123°10′ E and a latitude ranging from 27°02′ to 31°11′ N (Figure 1). It belongs to the subtropical monsoon climate zone with four distinct seasons and abundant precipitation. The terrain of the province slopes from southwest to northeast, mainly including mountains, hills, plains, and basins. Zhejiang Province has a population of about 57.37 million people and a gross regional product (GDP) of RMB 5619.7 billion, the industrial output value of which accounts for more than 41% of the regional GDP. Zhejiang Province is one of the most economically developed provinces in China, with a significantly higher road network density and expansion rate than the national average. Guided by the multi-faceted transportation development goals proposed in the National New Urbanization Plan (2014–2020), the density of the road network in Zhejiang Province has rapidly increased, the level of urban–rural transportation integration has been constantly improved, and the extension of transportation infrastructure to rural areas has been accelerated.

2.2. Data

2.2.1. Atmospheric AOD Data

The atmospheric AOD data used in this study were obtained from the MODIS MOD04-1K product (https://ladsweb.modaps.eosdis.nasa.gov/) within the period of 2014 to 2020, covering the Zhejiang Province region. The MODIS sensor has 36 discrete spectral bands, with a spectral range of 0.41 to 14.5 μm, a swath width of 2330 km, and a spatial resolution of 1 km. These characteristics allow for accurate and effective retrieval of AOD over both land and ocean surfaces. The MOD04-1K product utilized in this study is a standard Level 2 and Level 3 atmospheric aerosol dataset, featuring aerosol optical characteristics with a spatial resolution of 1 km, projected using the Lambert conformal conic method. The geographical coordinates are provided at a resolution of 30 arc-seconds. The daily data are Level 2 products, while the monthly data are compiled into Level 3 products.

2.2.2. Road Network Data

This study selected the Zhejiang Province road network data from 2014 to 2020 from Open Street Map (https://www.openstreetmap.org/) as the original data for road network analysis. First, a Fishnet was established to extract the road network data, and the road network density in each Fishnet unit was calculated through the intersection and fusion operations in ArcMAp 10.8. Then, the AOD concentration layer was linked to the Fishnet grid to assign corresponding road network density values to the CO₂ concentration layer.

2.2.3. Land Use Data

A public land use dataset released by Professors Yang Jie and Huang Xin from Wuhan University (https://zenodo.org/records/12779975, accessed on 20 February 2025) was utilized in this study, and data corresponding to Zhejiang Province were extracted for analysis.

2.3. Overall Workflow

2.3.1. Technology Roadmap

The technical roadmap of this study is presented in Figure 2 to outline the workflow of the correlation analysis.

2.3.2. Spatiotemporal Matching and Accuracy Verification of AOD Data

To verify the accuracy of atmospheric AOD products, the AOD values obtained from the extracted MODOIS AOD products were matched with the ground-measured AOD data obtained from field investigations.

Custom-developed functions in ENVI^®IDL were employed to process batch remote sensing images, extract fixed-point AOD observations from the images, and match the images and measurement data in time and space. Meanwhile, image extraction programs were mainly utilized to calculate the latitude and longitude range of each pixel in a given image and recursively calculate the pixel position of each station. Generally, MOD04_3K AOD data products contain outliers. Here, an outlier removal function was involved in our programming, and only aerosol values with outliers less than 2 pixels in eight domains were accepted. The matching program mainly realized the spatiotemporal matching of images with a time error of less than 30 min and the measurement data of the same place.

The root mean square error (RMSE), also known as standard error, is the square root of the ratio of the sum of squares of the deviation between observed and true values. The smaller the RMSE, the lower the standard deviation. In this study, the RMSE was employed to represent the error between the obtained AOD product and the ground truth data of the in situ solar photometer to verify the accuracy of the satellite AOD product. The RMSE was calculated as follows:

$$RMSE=\sqrt{{\displaystyle \frac{{\sum }_{i=1}^N(y_i-x_i{)}^2}{N}}}$$

(1)

where $y_i$ denotes the AOD value retrieved from the MOD04_3K AOD product, $x_i$ is the measured value, and $N$ represents the total number of matched data.

2.3.3. Atmospheric AOD Inversion

This study utilized the secondary development features of ENVI^® IDL (version 8.5) to perform batch remote sensing image processing. The original data was preprocessed (including removing the bow tie effect, geometric correction, band synthesis, cropping, etc.), and then AOD inversion was conducted to obtain the AOD inversion data for each year. First, the MODIS LIB data was detected by cloud, and the areas with thick clouds were eliminated. Subsequently, the dark pixel identification was performed, and the relationship between the red and blue bands of the dark pixels and the surface reflectivity of the short-wave infrared bands was determined; the search table suitable for urban areas was generated by the 6S radiation transmission model, and the table finding file was generated by the IDL program mobilizing the 6S model.

2.3.4. Road Network Density Calculation

First, in ArcMap, the fishnet grid was constructed with the boundary line of Zhejiang Province. Then, the road network data was spatially connected to achieve the intersection of the road network data and the fishing net data, and the total length of each grid was obtained through geometric calculation and group statistics; next, the area of each grid was calculated by geometry, and finally, the road network density (calculated as total length divided by grid area) was determined through field calculation.

2.3.5. Analysis of Land Use Change

The land use data of Zhejiang Province classified the land into cultivated land, forest, shrub, grassland, water body, bare land, and impervious surfaces. In this study, the land use transfer matrix and land use dynamics were utilized to analyze the changes in land use types.

(a)

Land use transition matrix

The land use transition matrix can reflect the structural change characteristics of land use types in the study area over some time. It is calculated as follows:

$$S_{ij}=\left[\begin{array}{ccccc}{S}_{11}& S_{12}& S_{13}& \dots & S_{1n}\\ S_{21}& S_{22}& S_{23}& \dots & S_{2n}\\ S_{31}& S_{32}& S_{33}& \dots & S_{3n}\\ ⋮& ⋮& ⋮& \ddots & ⋮\\ S_{n1}& S_{n2}& S_{n3}& \dots & S_{nn}\end{array}\right]$$

(2)

where $S$ represents the land use area, $n$ is the number of land use types, and $i$ and $j$ denote the land use types at the beginning and end of the study, respectively.

(b)

Land use dynamics

This study employed a single land use dynamics as the calculation model for the magnitude and velocity of land use type change in the study area, to reflect the quantitative change of a land use type in the study area at different times and spaces. The calculation formula is given by:

$$\mathrm{K}={\displaystyle \frac{{\mathrm{U}}_{\mathrm{a}}-{\mathrm{U}}_{\mathrm{b}}}{{\mathrm{U}}_{\mathrm{a}}}}\times {\displaystyle \frac{1}{\mathrm{T}}}\times 100\%$$

(3)

where $\mathrm{K}$ denotes the dynamics of a certain land use type during the study period, i.e., the annual rate of change; ${\mathrm{U}}_{\mathrm{a}}$ and ${\mathrm{U}}_{\mathrm{b}}$ represent the areas of a certain land type at the beginning and end of the study, respectively; $\mathrm{T}$ is the length of the study period.

2.3.6. Correlation Analysis

Based on the correlation between the land use transition matrix and the degree of change, combined with the AOD data analysis, a linear regression analysis was conducted on the AOD data and the road network density to obtain the correlation results.

3. Results

3.1. MODIS AOD Authenticity Check

Linear regression analysis was conducted between the measured atmospheric AOD values and the AOD values of MOD04_3K AOD products. The correlation index between the AOD values of MOD04_3K AOD products and the AOD observations of the ground photometer varied from 0.48 to 0.79 [23], suggesting that the AOD products of MOD04_3K satellites were reliable in the study area of Zhejiang.

3.2. Variation Characteristics of Atmospheric AOD in Zhejiang Province

Figure 3 illustrates the distribution of atmospheric AOD concentrations during the five years from 2005 to 2020. A horizontal comparison indicated that the high values of atmospheric AOD appeared in the northeast and eastern coastal areas of Zhejiang Province and the central plain. The low-value areas are located in the southern, southwestern, and central mountainous areas of Zhejiang Province.

Figure 4 shows that from 2005 to 2020, the concentration of atmospheric AOD in Zhejiang Province showed a downward trend, and the decline rate continued to increase. This decline can be attributed to the implementation of stricter environmental regulations, enhanced governmental oversight, and increased public engagement in pollution control efforts. As a result, the air quality in Zhejiang Province has continued to improve, resulting in notable improvements in regional air quality management.

3.3. Characteristics of Land Use Change in Zhejiang Province

In the past 20 years, changes in land use patterns in Zhejiang Province (Figure 5) have shown a clear urbanization trend. Due to the massive cultivated land reserves in Northeast China, the land use type has mainly shifted from cultivated land to impervious surfaces. Generally, the urbanization process is rapid, and the transformation of other land use types to impervious surfaces shows a trend of radiating and spreading outward from the urban center. The overall urbanization process is slow, but in recent years, with the development of the Qiantang River, the impervious surface along the basin has developed rapidly, and it mainly consists of woodland and arable land. In addition, the urbanization development in the central and southwestern parts of Zhejiang Province is limited, and the overall land use type has not changed much.

3.4. Correlation Analysis Between Atmospheric AOD and Land Use Types in Zhejiang Province

Table 1. Annual dynamics of land use.

Rate of Change/%	2000–2005	2005–2010	2010–2015	2015–2020	Total Rate of Change
Cropland	−1.25876	−1.0571	0.610158	0.182922	−1.52278
Forest	−0.03072	−0.03837	−0.56269	−0.21897	−0.85074
Shrub	−4.84276	−5.6653	−1.67437	−4.49718	−16.6796
Grassland	9.269627	3.689774	−9.17767	−7.07506	−3.29333
Water	1.285606	0.102805	−0.72694	−1.79497	−1.1335
Barren	19.60912	13.28947	6.581028	6.304833	45.78445
Impervious	7.969365	5.386533	4.204295	2.235232	19.79542

shows the annual dynamic degree of land use in Zhejiang Province from 2000 to 2005, 2005 to 2010, 2010 to 2015, and 2015 to 2020. The cultivated land area in Zhejiang Province experienced a decrease of 1.25876% in the first five-year period and 1.0571% in the second five-year period. In the past two decades, forests and grasslands have also shown a decreasing trend. A large amount of other types of land have been converted to impervious surfaces, with an increase of 19.79542% in impervious surfaces in the past 20 years.

According to the land change matrix and the matching of land change and AOD, the Pearson correlation coefficient was utilized to measure the linear correlation between AOD and various land use types (Figure 6). It can be observed that forests and shrubs are negatively correlated with atmospheric AOD, indicating a decreasing trend in atmospheric AOD concentration as both increase.

3.5. Characteristics of Road Network Density Change in Zhejiang Province

Since the collection of road network data started in 2014, the analysis was conducted using the data from 2014~2020. It can be seen from the road network density map of Zhejiang Province in Figure 7 that that the overall distribution of road network density was very stable during this period, exhibiting a decreasing trend from northeast to southwest. Nevertheless, with the development and expansion of cities, whether for travel or economic development, people are increasingly reliant on denser and more convenient transportation networks. Therefore, while the distribution pattern remains basically unchanged, the average road network density in Zhejiang is showing an upward trend. Meanwhile, with the development of science and technology, the construction of the road network in Zhejiang Province is gradually extending to rural and mountainous areas.

3.6. Correlation Between Atmospheric AOD and Road Network Density in Zhejiang Province

From a qualitative analysis perspective, the change map of atmospheric AOD concentration (Figure 3) was compared with the road network density map (Figure 7). The comparison indicates that the spatial distribution of the high-value AOD area and the high-density area of the road network is basically consistent in multiple regions, implying a certain spatial coupling relationship between the two. From a quantitative analysis perspective, the correlation coefficient between atmospheric AOD and road network density is mostly about 0.5 or even higher, indicating a significant positive correlation. Figure 8 presents a scatter plot of the correlation from 2014 to 2020, showing that the correlation coefficient (r) has generally shown a slight increase in the past few years. This may be due to the increase in road network density, traffic flow, and vehicle exhaust emissions with the advancement of urbanization, which in turn increases the aerosol concentration in the atmosphere, indirectly leading to the increase of AOD concentration. Although the overall level of AOD concentration has been decreasing year by year, reflecting the effectiveness of environmental management policies, road network density, as one of the important factors affecting the level of air pollution, cannot be ignored. Therefore, when formulating pollution prevention and control measures, the synergistic effect between urban traffic structure, motor vehicle management, and air pollutant control should be comprehensively considered.

4. Discussion

Based on MODIS AOD remote sensing data, this study analyzed the temporal and spatial evolution characteristics of AOD in Zhejiang Province from 2000 to 2020, and discussed its relationship with land use type and road network density. The results indicate that the AOD in Zhejiang Province as a whole exhibits a trend of first increasing and then decreasing, with typical distribution characterized as “high in the east and low in the west”, and significantly correlated with the intensity of human activities such as construction land expansion, and traffic network density. Nevertheless, the causes of AOD change are complex and diverse, influenced by both natural factors and human activities, and the following aspects are worthy of further discussion.

4.1. Limitations and Improvement Directions of Data and Methods

Firstly, although the MODIS data has good time coverage, there is a problem of reduced inversion accuracy in complex terrain and cloudy areas. Considering the diverse terrain of Zhejiang Province and the fact that there were no ground observation stations such as AERONET for cross-validation in the study, the research results may be affected by some errors. Therefore, future research will give priority to using local AERONET or fusion processing using multi-source remote sensing data (such as VIIRS and Sentinel-5P) to improve spatial resolution and reliability.

Secondly, road network density is calculated mainly based on the spatial statistics of the total length of roads, without fully distinguishing between different road grades or considering the contribution of actual traffic flow to pollution. Therefore, it is recommended that future research should introduce indicators such as traffic flow intensity and motor vehicle exhaust emission estimation to better capture the impact of traffic factors on aerosols. In addition, because road network data is currently limited to that available after 2014, this has limited our research analysis to the period 2014–2020. Although we provide insights into the spatial correlation between road networks and AOD in 2014–2020, to analyze the long-term role of transportation networks, future studies need to integrate consistent and longer-time-span traffic data.

Additionally, the correlation model established in this study did not take into account the possible lag effect of land use change on air pollution. In the future, time series methods such as time lag term or Granger causal analysis can be employed to more scientifically evaluate the dynamic relationship between urban expansion and AOD change.

4.2. Discussion on the Internal Mechanism of AOD Change

The spatial differences of AOD are not only related to human activities but also closely correlated with physical geographical conditions. Zhejiang is located in a hilly area with complex terrain, which may result in different aerosol retention and dispersion efficiency. Meanwhile, the urban heat island effect will change the boundary layer height and convection intensity, which in turn affects the vertical diffusion of pollutants and regulates the distribution of AOD. Future research will delve into the coupling relationship between the thermal environment and AOD by integrating thermal infrared remote sensing data with meteorological parameters.

Moreover, there are significant differences in the contribution of different land use types to AOD. For instance, industrial land usually corresponds to higher pollution emission densities, while green spaces and water bodies can be purified through vegetation adsorption and water vapor regulation. Hence, it is necessary to further refine the classification of land types and quantify the impact of various land uses on AOD using geostatistical analysis methods, such as geographically weighted regression (GWR).

4.3. Spatial Autocorrelation and Multi-Study Synergistic Implications

This study confirmed that the expansion of construction land and the density of transportation networks were positively correlated with AOD, and that forest and shrub vegetation had an inhibitory effect on AOD. This was consistent with the conclusion of Zhao et al. [24] in Xiamen: it was found that the proportion of forest land alone explaining AOD change was 75 times that of construction land (45.3% vs. 0.6%), which further verified the universal value of vegetation in improving urban air quality. The high-value AOD areas found in this study are concentrated along the traffic arteries and construction land expansion areas (Figure 3, Figure 5 and Figure 7), which complements the congested road study in Jinan by Feng et al. [25,26]: the pollutant emissions of congested road sections can reach 5.7 times those of unobstructed road sections through geographically weighted regression (GWR), and the road area occupancy is the core influencing factor of local air quality.

However, the spatial dependence analysis was not fully included in this study, which may overstate statistical associations and conflate spatial co-occurrence with causality, and affect the rigor of the conclusions. Therefore, in future research, it is necessary to emphasize the potential impact of spatial autocorrelation on statistical inference, and to reduce the impact of spatial autocorrelation by introducing spatial regression models and fusing multi-source data.

4.4. Expansion Direction

In the future, the scope of research can be extended from the provincial level to the city or district and county scale, and high-resolution remote sensing images and ground monitoring data can be combined to achieve more precise identification and tracking of pollution sources. Meanwhile, the correlation between AOD and ground PM₂.₅ can be investigated, and an air quality estimation model with remote sensing ground integration can be established to provide support for real-time air quality forecasting.

5. Conclusions

Based on long-term temporal road network data and land use data from 2000 to 2020, this study retrieved the AOD in Zhejiang Province using MODIS satellite imagery and investigated the correlations between atmospheric AOD, road network density, and land use types. The following conclusions were reached:

(1)

The contribution of different land use types to atmospheric AOD varies. Forest shrubs were significantly negatively correlated with AOD, which had a prominent reducing effect on atmospheric AOD, demonstrating a strong air purification capability. Due to the ability of vegetation to absorb aerosol particles in the atmosphere, reducing the secondary dust of surface dust and thus reducing air pollution, the increase in impervious surfaces reflects urban expansion, and the emission of pollutants during the expansion process leads to an increase in atmospheric AOD.

(2)

There is a positive correlation between atmospheric AOD and road network density. As an important source of aerosol emissions, the exhaust gases emitted by motor vehicles contain aerosol precursors such as nitrogen oxides (NOx) and volatile organic compounds (VOCs). These precursors create complex photochemical reactions in the atmosphere, forming secondary aerosols. The increase in road network density, especially in the expansion of road networks in busy traffic areas, leads to an increase in traffic flow, which in turn increases the exhaust emissions of motor vehicles, resulting in a higher concentration of AOD values.

(3)

The air control in Zhejiang Province has achieved fruitful results. Despite a positive correlation between atmospheric AOD and road network density, the data analysis results indicate that the atmospheric AOD concentration in Zhejiang Province showed an overall downward trend during the study period. This suggests that Zhejiang Province has achieved positive results in air pollution control, demonstrating the effectiveness of environmental policies and measures.

(4)

The road network in Zhejiang Province is extending to rural and mountainous areas, and the transportation network is well developed, promoting economic prosperity. This development trend has played a positive role in narrowing the gap between urban and rural areas. The expansion of transportation networks has improved connectivity and accessibility between regions, promoting optimal resource allocation and balanced economic development. Convenient transportation conditions have reduced logistics costs, allowing agricultural products to enter the market more quickly and tourism resources to be better developed, thereby promoting local economic growth, narrowing the income gap between urban and rural areas, and further promoting economic prosperity.