Balancing environmental benefits and accessibility for national park buffer zones in China

Abstract

National parks conserve species and ecosystems while offering educational, scientific, cultural, and recreational opportunities, which should extend beyond park boundaries to benefit surrounding areas. However, existing research often neglects both the recreational value of national parks and the development of their surrounding buffer zones. Using Baishanzu National Park, China, as a case study, this research proposes an integrated framework to assess the synergies and trade-offs between environmental effects and accessibility within a 30-km buffer zone. Results show that: (1) The park produces significant cooling and air purification effects, with a cooling intensity of 2.57 °C and a PM2.5 reduction of 0.97 µg/m³, extending over 2,474.08 km² and 3,039.43 km², respectively. (2) Favorable environmental conditions are mainly concentrated in Jingning and Qingyuan counties. (3) Accessibility analysis indicates relatively stable travel times and developed road networks, yet high travel costs persist, with the best accessibility observed in central Longquan and eastern Jingning counties. (4) Spatial autocorrelation analysis reveals a significant negative relationship between environmental effects and accessibility (Binary Moran’s I = − 0.392). Synergies (high-high clusters) occur near the northern park entrance, while trade-offs (high-low/low-high clusters) are found in southeastern and western buffer zones. Quantitative comparisons suggest that improving accessibility by 21.4% in high–low areas and enhancing environmental quality by 17.8% in low–high areas could mitigate spatial mismatches. These findings provide scientific support for differentiated buffer-zone planning for national parks and inform strategies to enhance their recreational potential while promoting sustainable development.

Introduction

Globally, as human activities increasingly influence natural ecosystems, new challenges emerge, including climate change, land-use transformation, and the spread of invasive species^1,2,3. In response to the global decline in biodiversity and ecosystem services^4,5,6,7, the establishment of protected areas (PAs) plays a key role in mitigating the loss of species’ habitats, among other concerns^8,9,10. National parks, which originated in the United States, represent a significant category of PAs. These parks are typically large and largely undisturbed by human activities^11,12, with the primary goal of safeguarding wilderness areas from human encroachment. However, this model of excluding local populations has also led to various social consequences^13,14,15. Balancing ecological conservation with the well-being of local communities is a critical issue¹⁶, and the social challenges associated with national parks, such as resettlement, conflicts, and the livelihoods of local residents, must be addressed. Consequently, it is widely recognized that national parks serve not only to preserve the integrity of species and natural environments but also to provide educational, scientific, cultural, and recreational opportunities suited to their context¹⁷. As the largest developing country, China faces significant challenges to biodiversity conservation due to its rapid economic growth¹⁸. PAs in China, which account for nearly one-fifth of the country’s land area, often experience high levels of conflict due to “people-land constraints,” as most reserves encompass long-standing local communities and feature collectively owned land^19,20. Amid the conflict between economic development and ecological conservation, the Chinese government has advanced the “Two Mountains Theory” – the idea that “lucid waters and lush mountains are invaluable assets” – to promote harmonious coexistence between humans and nature^21,22. Under this guiding theory, China is establishing a national park-based system of PAs, designed to preserve natural ecosystems without compromising sustainable economic and social development²³. Land use and land cover change surrounding PAs can increase pressure on these regions, and the establishment of external buffer zones is widely regarded as an effective mitigation strategy²⁴. In China, the management approach often involves creating such peripheral zones, which are designed to serve both public welfare and economic functions, including ecological protection, education and research, and community services²⁵. However, existing research on national parks predominantly focuses on internal aspects, such as institutional systems, ecosystem services, and zoning^13,26,27, with little attention given to buffer zones and the realization of the park’s recreational value. This neglects the role of national parks in addressing local needs. Buffer zones benefit from the park’s ecological spillover effects while preserving its integrity, and should therefore be integrated into planning to support sustainable development.

Urban parks, as the most prevalent form of blue-green infrastructure, offer recreational spaces and provide essential ecological functions. Recreational services encompass a variety of outdoor activities, including dynamic pursuits such as walking, biking, and jogging, as well as static activities like sunbathing, reading, and nature observation²⁸. Ecologically, parks contribute to cooling, air pollutant reduction, carbon sequestration, and the conservation of biodiversity^29,30,31. Among these environmental benefits, the cooling capacity and air quality improvement of parks have attracted considerable attention³². The interaction between vegetation and water bodies within parks helps mitigate surface heat absorption, while processes such as adsorption and sedimentation reduce the dispersion of air pollutants, collectively creating a natural system for cooling and air purification^33,34,35. As cities expand, environmental issues such as urban heat islands and air pollution negatively impact urban residents³⁶. Temperature and air quality can influence residents’ decisions regarding outdoor recreational activities. Therefore, quantifying the cooling and air quality improvement effects of parks is essential for understanding their contribution to the surrounding environment and their role in supporting recreational functions. Current research has proposed indicators to assess the cooling effects of urban parks, such as cooling distance, area, and efficiency^37,38, as well as air purification metrics like PM_2.5 (fine particulate matter with an aerodynamic diameter < 2.5 μm) reduction distance, area, and efficiency³⁹. While existing research has predominantly focused on typical urban parks, this study shifts the focus to national parks. National parks share a fundamental characteristic with traditional parks through their commitment to public welfare, serving as vital sites for ecotourism, scientific research, and environmental education⁴⁰. We posit that the environmental mechanisms observed in urban parks are fundamentally relevant to national parks. Given their typically larger ecological spaces and more complex ecosystems, the cooling and air purification effects of national parks are likely to extend over a broader area than those of urban parks, potentially benefiting surrounding communities to a greater extent. Building upon the knowledge derived from urban studies, this research aims to quantify these environmental effects, thereby enhancing the ecological assessment systems for national parks and clarifying how they improve the quality of life for local residents—for example, by mitigating extreme heat and promoting respiratory health. This, in turn, offers a scientific basis for decision-makers to strike a balance between ecological conservation and improving local livelihoods, ultimately fostering the co-realization of the ecological and recreational functions of national parks.

One of the most important factors influencing the availability of recreational functions is accessibility⁴¹. Accessibility, defined as the ease of reaching a target location from a given point^42,43, directly influences the performance of ecological services within parks and the extent to which local communities’ benefit. In general, parks with higher accessibility are more likely to provide recreational services, whereas those with inadequate transportation infrastructure, despite offering a superior natural environment, are less likely to do so due to limited resident access⁴⁴. Established methods for measuring accessibility include the container method, gravity model-based methods, and travel cost methods⁴⁵. In recent years, the application of big data technology has further reduced data collection costs through open data. National parks and their surrounding buffer zones, as accessible natural areas, not only provide crucial ecological protection but also offer a range of benefits to nearby residents, including economic opportunities (e.g., ecotourism), enhanced social well-being (e.g., public recreational spaces), and improved psychological well-being (e.g., the therapeutic effects of nature)⁴⁶. Given these benefits, conducting accessibility research on national parks is necessary to fully understand their potential in supporting both ecological and recreational functions. However, most existing research has primarily focused on the accessibility of urban public services (e.g., city parks, hospitals), leaving a gap in studies on the accessibility of national parks as unique natural environments.

Baishanzu National Park in Zhejiang Province, China, is a pilot project of China’s national park system, renowned for its stunning natural landscapes, including Huangmaojian, the highest peak in the province. The park benefits from a favorable climate and excellent air quality, making it an ideal summer retreat for urban residents seeking relief from the heat and desiring a closer connection to nature. Given that summer represents the peak visitation period and poses the most significant challenges for environmental management, this study focuses primarily on summer conditions to develop a spatially-explicit assessment framework integrating environmental effects with accessibility analysis. Considering the park’s internal ecological integrity and the policy constraints it faces, this study applies the methodology to the 30-km external buffer zone to identify areas with high recreational potential, and on this basis proposes zoning strategies to guide the sustainable realization of recreational value. Our research objectives are: (1) to analyze the environmental impacts of the national park on the surrounding area, focusing on surface temperature and PM_2.5 levels; (2) to evaluate accessibility within the park’s buffer zone by combining dynamic travel time data with static road network information; and (3) to integrate environmental impacts and accessibility through spatial correlation modeling to identify key areas within the buffer zone with potential for recreational development. The study result can provide a scientific foundation for the dual realization of ecological and recreational functions in national parks, offering a new model for their sustainable development.

Study area and data

Study area

Baishanzu National Park, located in the southern part of Lishui City, Zhejiang Province (Fig. 1), spans Longquan, Qingyuan, and Jingning counties, covering a total area of 503.68 km². The park lies in a mid-subtropical monsoon climate zone, where summer temperatures are consistently high. For instance, in July and August 2024, the number of days with temperatures exceeding 35 °C averaged 28.5 (http://www.tianqihoubao.com/), reflecting a strong local demand for thermal relief. As a national park still in its early development stage, Baishanzu is actively exploring pathways to translate ecological assets into socioeconomic benefits while promoting coordinated regional development—making it a representative case for studying conservation-compatible development models in China.

The flexibility of development in the external buffer zone—without compromising the ecological integrity of the park’s core—makes it a suitable area for examining the spillover effects of national parks. Previous studies suggest that the cooling effect of a national park can extend to an area roughly 11 times its size³². In addition, a 30-km radius from the park generally corresponds to about 30 min of driving time, aligning with the typical tolerance of visitors for recreational travel. Therefore, this study defines a 30-km external buffer zone for analysis. This zone encompasses urban-rural transitional areas that capture both ecological and infrastructural advantages, effectively reflecting the tourism spillover effects of the national park while avoiding visitor attrition due to excessive distance.

Fig. 1

Study area. This map displays the location of the study area. It was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

Data sources

The data used in this study primarily consist of multi-source satellite data and PM_2.5 concentration data for calculating environmental effects, as well as government point-of-interest (POI) data for each district and county and driving cost data for travel, both of which are used to calculate accessibility. All data were resampled to a spatial resolution of 30 m×30 m, with the projected coordinate system being WGS_1984_UTM_Zone_50N. Detailed data sources are provided in Table 1.

Table 1 Data sources.

Methodology

Framework

This study presents a comprehensive framework for national parks to simultaneously achieve ecological and recreational functions, based on the interdependence between ecological protection and socio-economic development (Fig. 2). From a theoretical perspective, the development of the national park can be divided into three levels. The first level focuses on the internal dynamics of the national park, with an emphasis on ecological protection through environmental conservation and sustainable management practices designed to preserve and enhance the park’s ecological integrity. The second level examines the buffer zone, a dynamic region where ecological functions and socio-economic activities intersect. This zone aims to convert the park’s ecological benefits into tangible socio-economic values, such as eco-tourism and agricultural production, thereby benefiting the surrounding areas. The outermost level encompasses urban areas, reflecting broader social and economic dynamics that are positively impacted by the national park. While the interior of the national park is typically subject to strict regulations, the buffer zone, as a transitional area between the park and its surroundings, offers relatively good natural resources along with greater development potential and flexibility, due to fewer regulatory constraints. This makes the buffer zone particularly significant for realizing the recreational value of the national park. Therefore, this study uses Baishanzu National Park in China as a case study, focusing on its 30-km buffer zone, and is structured around three core steps: (1) Environmental Effect Analysis: This component examines two critical environmental factors—Land Surface Temperature (LST) and PM_2.5 levels—quantifying their effects using indicators including park cooling/PM_2.5 reduction areas (PCA/PPRA) and intensities (PCI/PPRI). (2) Accessibility Analysis in the 30-km Buffer Zone: The analysis combines dynamic data (travel time costs from government offices in Lishui to townships within the buffer zone) and static data (distance to the road network) to assess the accessibility of these areas. (3) Spatial Autocorrelation Analysis of Environmental Effects and Accessibility: Using both global and local Moran’s I, this analysis examines the spatial autocorrelation between environmental effects and accessibility, identifying hotspots where favorable environmental and transportation conditions coincide. By comprehensively exploring the synergy and trade-offs between environmental effects and accessibility, the study aims to develop a balanced model that integrates ecological protection with regional development objectives, thereby promoting the sustainable management of national parks and their surrounding areas.

Fig. 2

The research framework.

Assessment of environmental effects for national park

LST and PM_2.5

Hot weather in Lishui typically lasts from June to September. Given that the park’s cooling effect on the surrounding area is most pronounced during the summer months, and considering data availability, this study selected Landsat 8 OLI_TIRS imagery from the United States Geological Survey (USGS) on September 27, 2021 (path 119, rows 40 & 41), with less than 0.1% cloud cover. Remote sensing imagery was preprocessed using ENVI 5.6, which involved image mosaicking and clipping, as well as radiometric calibration and atmospheric correction. On this date, the weather in Lishui was clear, with temperatures ranging from 22 °C to 34 °C and no significant weather events. The detailed LST inversion procedure is provided in the online Supplementary Methods.

The PM_2.5 data for 27 September 2021 were sourced from the China High Air Pollutant (CHAP) dataset, which is widely used in research due to its high resolution (1 km×1 km)^48,49. This dataset estimated PM_2.5 concentrations using the Spatio-Temporal Extra Tree (STET) model, which is based on long-term, high-resolution aerosol optical depths derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) Multi-Angle Implementation of Atmospheric Corrections (MAIAC) algorithm, with a spatial resolution of 1 km⁴⁷.

Park environmental effects measurement

Temperature and PM_2.5 reduction are common environmental benefits provided by parks to surrounding areas. This study quantified these effects by assessing the park’s ability to lower both temperature and PM_2.5 concentrations. Parks reduce these factors within their boundaries and in surrounding areas, with a notable distance attenuation (Fig. 3)^39,50. These effects are commonly quantified by comparing LST and PM_2.5 inside and outside park boundaries, using buffer zones of 10–20 distances and identifying the first turning points to analyze cooling and air quality improvements^51,52. In this study, we focus on a national park, which differs from urban parks in terms of its larger influence radius. Given that Baishanzu National Park spans three counties, with a maximum radius of 30-km from the park’s center, we created a 30-km buffer zone and 20 additional buffer zones spaced 1500 m apart to capture the full extent of the park’s influence. Mean LST and PM_2.5 concentrations were calculated for each buffer and modeled relative to distance from the park using a cubic polynomial function, E(x), where x is distance from the park boundary and y is either LST or PM_2.5 concentration.

Based on the curve, we identified the park’s cooling/PM_2.5 Reduction Distance, which represents the point where significant changes or plateaus occur on the LST/PM_2.5 concentration curve. Building on this, we selected four indicators to assess the environmental effects: park cooling/PM_2.5 reduction area (PCA/PPRA) and park cooling/PM_2.5 reduction intensity (PCI/PPRI). The PCA/PPRA represents the maximum buffer zone within which park cooling/PM_2.5 reduction effects can be detected. It is determined by measuring the maximum distance from the park boundary at which cooling and PM_2.5 reduction effects are still noticeable⁵³. Operationally, this maximum distance was used to create an external buffer zone around the park polygon in ArcGIS, and the area of this zone was then computed to quantify the PCA/PPRA. Higher values indicate a more pronounced cooling/PM_2.5 reduction effect. PCI/PPRI is calculated as the difference between the initial inflection point temperature/PM_2.5 concentration and the average park temperature/PM_2.5 level⁵⁴, assessing the sensory experience of local residents. Higher values indicate a stronger perception of reduced surface temperatures and PM_2.5 concentrations. Additionally, this study normalized the spatial data of LST and PM_2.5 within the 30 km buffer zone of the park to a range of 0 to 1. A value close to 1 indicates lower LST and PM_2.5 concentrations, reflecting better environmental conditions. We assigned a weight of 0.5 to the normalized results for both LST and PM_2.5 and summed them to assess the environmental impacts within the buffer zone around the park, resulting in the spatial distribution of environmental effects within the buffer zone.

$$E(x) = ax^{3} + bx^{2} + cx + d$$

(1)

Fig. 3

Schematic diagram of cubic multiple function curve between cooling/PM_2.5 reduction of park and distance from park.

Assessment of accessibility considering road conditions

Traditional accessibility measurements typically use ArcGIS to calculate travel costs under ideal conditions⁵⁵. In reality, travel costs are influenced by traffic, leading to higher actual costs. With advances in information technology, real-time navigation data now provide a more comprehensive view of transport accessibility across time and space. Therefore, this study integrated both dynamic and static data. Dynamic data reflect travel time costs based on real-time traffic conditions⁵⁶, while static data measure distance from roads, with closer proximity generally indicating higher accessibility.

Travel time estimation

Travel time estimation is essential for measuring accessibility. This study used Amap open APIs to obtain road network and real-time traffic data, including speed and congestion levels, passively collected from user-generated location and navigation data. When users navigate through an online mapping application, their vehicle location and speed data are anonymously aggregated by the mapping service⁵⁷. Using Python, trip details such as time, distance, and route were extracted with parameters including start and end coordinates, departure time, transport mode, coordinate system, and travel preferences. Given that public transportation is underdeveloped in the study area and residents primarily travel by car, we set the mode of transport to driving. In the Amap Route option, the “Recommended” route was selected, providing a balanced choice based on an algorithmic assessment of travel time and cost. The study selected 9 governments of districts and counties in Lishui as the starting points and the geometric centers of 61 townships within the 30-km buffer zone surrounding the National Park as the endpoints. District and county governments were chosen as starting points due to their central role in regional human activity, serving as the primary departure points for most residents of Lishui. Townships in the 30-km buffer zone were selected as endpoints to represent rural and semi-rural areas directly influenced by the presence and development of the National Park. We averaged the data from different starting points that reached the same endpoint to represent the travel time from the urban area to that endpoint at a specific time. To enhance the reliability of the estimated travel times and account for the fact that urban residents mainly visit the national park perimeter on weekends, we collected data at two-hour intervals (from 8:00 AM to 6:00 PM) on weekends (Saturdays and Sundays) and computed the average for each time point.

Accessibility measurement

The study averaged drive times at each time point to represent overall travel costs from each township to the urban area within the buffer. To facilitate spatial analysis, kriging interpolation was applied to convert these data into a continuous surface. Euclidean distance was used to calculate proximity to roads, forming a static accessibility model. Drive time and road distance were then normalized to a 0–1 range, with values closer to 1 indicating better accessibility. The two normalized datasets were weighted equally (0.5 each) and summed to calculate final accessibility within the park’s 30-km buffer.

Spatial autocorrelation analysis

Spatial autocorrelation occurs when values of spatial variables, such as habitat quality, are similar in nearby locations⁵⁸. Moran’s I is commonly used to assess autocorrelation, providing insights into the distribution and interrelationships of spatial data⁵⁹. Moran’s I can be univariate or bivariate, with the bivariate version describing the spatial relationship between two distinct variables and revealing potential spatial spillover effects. GeoDa software was used to perform both global and local spatial correlation analyses of environmental effect and accessibility, examining their spatial clustering patterns around the park.

Global Spatial autocorrelation analysis

Global Moran’s I measures the overall spatial relationship between all geographical units in a study area⁶⁰. Its value ranges from − 1 to 1: positive values indicate clustering, with larger values representing stronger positive correlation, while negative values indicate dispersion, with smaller values representing stronger negative correlation. Values near 0 suggest no correlation and a random spatial distribution. Binary Moran’s I is calculated using the following formula⁶¹:

$$\:{I}_{ap}=\frac{N\sum\:_{i=1}^{N}\sum\:_{j\ne\:i}^{N}{W}_{ij}\hspace{0.17em}{z}_{i}^{a}{z}_{j}^{p}}{\left(N-1\right)\sum\:_{i=1}^{N}\sum\:_{j\ne\:i}^{N}{W}_{ij}}$$

(2)

Where, $\:{I}_{ap}$ is the bivariate global spatial correlation index, N is the total number of geographic units (grids), $\:{W}_{ij}$ is the spatial weight matrix, $\:{z}_{i}^{a}$ is the standardized environmental effects degree in grid i, and $\:{z}_{j}^{p}$ is standardization of accessibility degree in grid j.

Local spatial autocorrelation analysis

Global Moran’s I captures the overall correlation between two variables, but local relationships may vary. To address this, we used the local LISA (Local Indicators of Spatial Association) index, which describes spatial heterogeneity of correlations and is useful for studying local patterns. The resulting binary LISA map illustrates the relationship between a variable at a location and the mean value of another variable at neighboring locations at a specified significance level⁶². The map shows high-high, high-low, low-high, and low-low clusters. High-high clusters indicate both variables have high values and are positively correlated, highlighting hotspots with favorable environmental and traffic conditions. The calculation process is shown below:

$$\:\text{LIS}{\text{A}}_{i}^{ap}={z}_{i}^{a}{\sum\:}_{j=1}^{N}{W}_{ij}{z}_{j}^{p}$$

(3)

Where, $\:\text{LIS}{\text{A}}_{i}^{ap}$ is the Bivariate local Moran’s I index, other variables are the same as in the Eq. (2).

Results

The cooling and PM_2.5 reduction effect of the national park

Spatially (Fig. 4), LST within the national park is relatively low, mostly ranging between 19.9 °C and 29.2 °C, while PM_2.5 concentrations range from 14.2 to 15.6 µg/m³. Both parameters show a gradual increase from the interior of the park outward. The comparative analysis of LST and PM_2.5 concentrations revealed distinct environmental gradients across the national park, buffer zone, and adjacent external areas (Fig. 5). Quantitative measurements showed average LST values of 26.69 °C (national park), 28.66 °C (buffer zone), and 29.35 °C (outside buffer zone), demonstrating the significant cooling capacity of intact ecosystems. Similarly, PM_2.5 concentrations exhibited parallel spatial trends, with mean values of 15.05 µg/m³ (park), 15.86 µg/m³ (buffer), and 16.86 µg/m³ (external zones). The results demonstrate national parks’ role as thermal regulators and particulate matter filters, improving air quality by reducing PM2.5. Buffer zones bridge the environmental gradients between the park and urban areas, highlighting national parks as crucial infrastructure for climate adaptation and air quality management in urban regions.

Fig. 4

Spatial patterns of LST and PM_2.5 within and around the national park. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

Fig. 5

Comparisons of LST and PM_2.5 concentration between the national park, buffer zone and outside buffer zone.

The study used a third-order polynomial function to model the cooling and PM_2.5 reduction effects in Baishanzu National Park, revealing nonlinear relationships between LST, PM_2.5 concentrations, and distance from the park (Fig. 6). Both LST and PM_2.5 concentrations rose initially and then decreased with distance from the park, with LST and PM_2.5 showing first turning points at 10.5 km and 13.8 km, respectively. The cooling effect diminished beyond 10.5 km, while the air-purifying effect extended up to 13.8 km, likely due to vegetation and atmospheric dispersion. The park’s cooling area (PCA) spanned 2,474.08 km², and its PM_2.5 reduction area (PPRA) extended over 3,039.43 km², with a greater impact on air quality. The cooling intensity (PCI) of 2.57 °C was more pronounced than the PM_2.5 reduction intensity (PPRI) of 0.97 µg/m³, highlighting a more noticeable temperature reduction.

Fig. 6

Cubic polynomial associations between LST/PM_2.5 and the distance from the national park.

Spatial patterns of environmental effects in national park buffer

The normalized LST and PM_2.5 values were weighted and aggregated to calculate environmental effect values within the buffer zone (Fig. 7). Values closer to 1 indicate better environmental conditions, with the eastern region (Jingning and Qingyuan counties) showing lower temperatures and PM_2.5 concentrations, making it favorable for both residents and wildlife. In contrast, the western region (Longquan and Qingyuan counties), with higher temperatures and PM_2.5 levels, faces poorer environmental conditions due to urbanization and industrial activity. According to the natural breakpoint classification (Table 2), 59.67% of the buffer zone had an environmental effect index greater than 0.80, suggesting generally favorable conditions. Areas with indices between 0.75 and 0.80 (24.76% of the zone) showed more variation in environmental quality. Regions with high environmental effect values, featuring mild temperatures and clean air, are attractive for nature-based tourism and long-term park sustainability. Conversely, areas with average conditions, marked by pollution and high temperatures, may deter eco-conscious tourists. These regions may need alternative strategies to improve environmental quality and support sustainable tourism.

Fig. 7

Spatial patterns of environment effect within the national park buffer zone: (a) normalized LST, (b) normalized PM_2.5, and (c) environment effect based on weighted summation of indicators. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

Table 2 Statistics results of environmental effect in National park buffer zone.

Spatial patterns of accessibility in national park buffer

Figure 8 illustrated the spatial distribution and temporal variation of driving travel times within the 30-km buffer zone around the national park on weekends. Travel times remained relatively stable throughout the day, with minor fluctuations. Minor peaks were observed around 12:00 and 18:00, potentially corresponding to midday travel and return trips. The overall limited variability confirms that these regional highways are not subject to severe congestion. This stability validates our methodological decision to use an averaged travel time for a robust assessment of recreational accessibility. Travel times were categorized into 90, 120, 150, and 180 min, with two hours being the typical upper limit for same-day trips. Most residents of Lishui reached the buffer zone within 90 to 120 min, indicating challenges in accessibility due to transportation constraints. Spatially, driving times were shorter in the northern Longquan County and longer in the southern Qingyuan County, likely due to topography and underdeveloped infrastructure. Proximity to the buffer zone’s periphery did not always correlate with shorter travel times, as access depended on the highway network and toll gate locations, with central Longquan County benefiting from shorter travel times due to its proximity to a toll gate.

Fig. 8

Spatial patterns of travel time from urban areas to national park buffer zones over time. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

The normalized travel time and road network distance were weighted and aggregated to derive accessibility values within the buffer zone (Fig. 9). Values closer to 1 indicate better accessibility, with high accessibility areas concentrated in central Longquan and eastern Jingning counties due to well-developed road networks and proximity to toll stations. In contrast, southern and mountainous areas had lower accessibility due to sparse infrastructure and longer travel times. According to the natural breakpoint classification (Table 3), 30.98% of the buffer zone had an accessibility index between 0.78 and 0.86, suggesting relatively high but uneven accessibility compared to environmental conditions. The next range, 0.78–0.90, covered 26.15%, while 19.18% of the zone had an index over 0.86, and 4.98% had low accessibility. These variations suggest that areas with better accessibility are more suitable for eco-tourism and related services, while less accessible regions may face development challenges. These spatial disparities should be considered in park planning to ensure equitable development and distribution of ecological benefits.

Fig. 9

Spatial patterns of accessibility within the national park buffer zones: (a) normalized travel time, (b) normalized distance to roads, and (c) accessibility based on weighted summation of indicators. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

Table 3 Statistics results of accessibility for residents to National park buffer zone.

Spatial autocorrelation between environmental effects and accessibility

Spatial autocorrelation analyses of environmental effects and accessibility revealed a distinct pattern of spatial aggregation between the two variables (Figs. 10 and 11). The binary Moran’s I value between accessibility and environmental effects was − 0.392, indicating a strong negative correlation. The Local Indicators of Spatial Association (LISA) clustering diagram identifies five distinct categories representing the spatial relationships of the two variables⁶³. Among these, the high-high clusters—areas characterized by both high environmental effects and high accessibility—represent hotspots in the region. Conversely, low-low clusters, which are areas with low environmental effects and low accessibility, form cold spots.

The results showed that the high-high clusters were primarily concentrated near the northern and eastern entrances of the national park within Longquan County. Field surveys in these areas confirmed well-developed recreational infrastructure, including established B&B facilities and accommodation services, demonstrating their current status as active recreational hubs. These areas exhibit both excellent accessibility and significant cooling and PM2.5 reduction effects, creating ideal conditions for sustainable tourism development such as eco-tourism, green lodging, and recreational services. The synergy between environmental advantages and existing infrastructure enables these zones to effectively convert ecological benefits into economic value while preserving natural resources. The established transportation network and low travel costs further enhance their appeal, creating a sustainable model where conservation and local economic development reinforce each other. On the other hand, low-low clusters, accounting for a relatively small proportion of the total area, were predominantly found in the western part of the buffer zone in Qingyuan and Longquan County. Field observations in these areas revealed limited recreational infrastructure but identified potential for developing eco-friendly agricultural practices such as terrace farming and local product marketing. The emergence of this clustering pattern can be attributed to the influence of adjacent urban areas where temperature and air quality conditions are relatively poor. While these regions are situated far from the major population centers of Lishui City, resulting in poor transportation accessibility from that direction, their location may offer better connectivity for residents of Fujian Province. Ecological protection should therefore remain the priority, supplemented by sustainable agricultural development to provide supplemental livelihoods. Cross-jurisdictional cooperation should be enhanced to maximize regional benefits while ensuring environmental conservation.

Additionally, the high-low clusters, concentrated in the southeastern region within Jingning and Qingyuan counties, and the low-high clusters, located primarily in the western part of the buffer zone, highlight areas that require targeted interventions. Specifically, these regions should focus on balancing ecological protection with economic development potential. In high-low clusters characterized by superior environmental quality but constrained accessibility, field surveys reveal well-preserved ecological landscapes and traditional villages that remain underdeveloped due to inadequate transportation infrastructure. These areas offer significant potential for niche eco-tourism development, particularly through the establishment of research stations and environmental education centers that would appeal to environmentally conscious tourists. Quantitatively, the average accessibility index in these areas is 0.66, which falls approximately 0.14 below the 0.80 benchmark observed in high-high clusters. This gap represents a required improvement of 21.4% from current levels. To achieve this enhancement, strategic infrastructure development including the construction of parking facilities and optimized highway interchanges would significantly improve connectivity. Such improvements would make these regions more attractive and accessible to visitors while maintaining their ecological value. Similarly, low-high clusters demonstrate good accessibility but relatively poor environmental quality, primarily consisting of townscapes. Quantitative results indicate an average environmental index of 0.73 in these areas, approximately 0.13 below the 0.86 benchmark observed in high-high clusters. This gap represents a required environmental improvement of 17.8%. To address this deficit, these areas would benefit from targeted green infrastructure interventions to enhance their ecological value while maintaining convenient transportation access. Strategic urban greening measures could effectively improve environmental conditions in these well-connected zones, creating more balanced and sustainable development.

Notably, the area covered by high-high clusters is significantly larger than that of low-low clusters, indicating that Baishanzu National Park’s buffer zone holds substantial development potential. This suggests opportunities for integrating the park with the local community for sustainable development. The key to this integration lies in capitalizing on the synergies between accessibility and environmental quality. By improving infrastructure, such as roads, transportation, and communication networks, the park and surrounding communities can better align their goals of conservation and economic growth. However, the presence of a large proportion of high-low and low-high clusters indicates that further targeted efforts are needed to address the specific challenges in these areas. For example, improving accessibility in remote, low-access areas and simultaneously enhancing environmental quality in more accessible, developed areas could help balance development and conservation efforts across the buffer zone. These interventions will be critical to ensuring that the national park and surrounding communities can thrive together in a sustainable, mutually beneficial manner.

Fig. 10

The allocation of spatial relevance between environmental effects and accessibility with field survey photos. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html) and GeoDa 1.20 (https://geodacenter.github.io), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification. All photographs are original and were taken by the author on March 16 and September 10, 2025.

Fig. 11

(a) significant level map and (b) scatterplot between accessibility and environmental effects. The map was generated by the author using ArcGIS 10.7 software (https://www.arcgis.com/index.html) and GeoDa 1.20 (https://geodacenter.github.io), based on the standard map (No. GS(2024)0650) issued by the Ministry of Natural Resources of the People’s Republic of China (https://www.tianditu.gov.cn/). The base map is used without modification.

Findings and discussion

Leveraging buffer zones as spatial bridges for multifunctional national parks

Our quantification of cooling effects and PM_2.5 reduction reveals a well-defined spatial gradient of ecological benefits extending radially from the park’s core to peripheral areas. In contrast to findings from urban park studies^32,64, Baishanzu National Park exhibits a substantially broader spatial influence, underscoring the unique role of large-scale, contiguous natural landscapes in regional climate modulation and air quality enhancement. This evidence not only confirms the significant environmental impact of extensive protected areas but, more critically, reframes the functional identity of buffer zones—from serving merely as conservation transitions to functioning as recipients of “ecological dividends.”

Consequently, this study posits that buffer zones represent pivotal spatial connectors for actualizing multifunctional objectives in national park systems. Their developmental potential derives not merely from relatively flexible land-use policies but more fundamentally from the spillover of high-quality ecological capital from core protected zones—characterized by moderated microclimates and improved air purity. This inherent advantage creates ideal conditions for low-impact nature-based tourism development, including eco-resorts and wellness retreat facilities. The systematic integration of buffer zones into comprehensive national park planning does not compromise conservation goals; rather, it facilitates the transformation of ecological advantages into sustainable livelihood options, thereby reinforcing conservation outcomes across broader spatial contexts and establishing a positive feedback loop between ecological preservation and socio-economic development.

Balancing ecological integrity with public access in national park management

To address the growing demand for recreational spaces that simultaneously provide ecological and environmental functions, this study proposes an innovative “environment-accessibility” integrated analysis framework. This framework systematically incorporates two key dimensions—environmental quality and public accessibility—into national park planning, offering a more comprehensive solution for park management. Traditional national park planning has typically focused on internal conservation objectives, often relegating recreational services and public access to secondary considerations. In contrast, our framework shifts this paradigm by emphasizing the importance of recreational value and clarifying how national parks can fulfill both ecological and social functions, thus transforming planning concepts.

One of the framework’s key contributions is its ability to quantify and visualize the spatial relationship between environmental benefits and accessibility, providing a clearer understanding of how national parks can simultaneously function as ecological sanctuaries and accessible recreational spaces. By addressing the interaction between these two dimensions, the framework not only strengthens the scientific basis for park management but also guides policy-making that seeks to balance ecological conservation with public access to nature. Our analysis reveals a spatial mismatch between the environmental effects of national parks and their accessibility, which constitutes a key bottleneck in realizing the full recreational potential of these parks. Accessibility plays a critical “control valve” role, as even areas with superior environmental conditions may have their recreational potential “locked” due to weak infrastructure. This highlights the importance of improving accessibility to ensure that national parks fulfill their role as public resources, fully unlocking their potential benefits. Therefore, the framework provides planners with a vital tool to identify these accessibility gaps and prioritize infrastructure improvements in key areas to enhance the recreational value of parks.

Additionally, the framework offers an operational tool for regional spatial planning, advocating for a more refined and gradual zoning strategy in national park management. It emphasizes the gradual relationship between protection and utilization, rather than relying on the traditional binary classification of “protected” and “non-protected” areas. By recognizing the gradient between conservation and use, park management can adopt strategies that both preserve the integrity of these areas and promote the development of ecotourism, recreational activities, and community services. This integrated approach challenges the conventional “fortress conservation” model, which views parks as isolated islands with strict internal protection mechanisms. Instead, it calls for a broader perspective, positioning national parks as part of a larger regional green infrastructure network. This shift enables park management to align internal conservation efforts with the needs of surrounding communities, thereby achieving a balance between ecological protection and socio-economic development. Focusing on regional connectivity, national parks can become pivotal hubs for environmental services, not only safeguarding biodiversity but also promoting sustainable development in adjacent areas.

Strategic management of national park buffer zones for sustainable development

This study used spatial autocorrelation analysis to identify four clustering patterns (high-high, high-low, low-high, and low-low) in the buffer zone, providing valuable policy insights for national park planning and management. Based on these findings, planners should develop more specific strategies for effective management.

The areas characterized by the synergy between environment and accessibility are the high-high clusters, where superior environmental quality and convenient transportation intersect. These regions are best suited for sustainable tourism development, offering ideal conditions for nature-based recreational facilities such as campsites, observation decks, eco-lodges, and resorts. From a strategic perspective, this offers the opportunity to leverage the park’s brand value while simultaneously promoting local economic development through tourism. Moreover, tourism initiatives in these areas could also serve as platforms for raising public awareness on the importance of biodiversity conservation and the value of natural ecosystems, which can help foster a culture of environmental stewardship among visitors and local communities alike. However, careful planning is required to ensure that tourism development does not surpass the region’s ecological carrying capacity, which could ultimately undermine the very environmental quality that attracts visitors in the first place. In this sense, the sustainable tourism model in high-high clusters must strike a delicate balance between economic benefits and conservation goals.

The areas where environmental quality and accessibility directly exhibit antagonistic effects are the high-low and low-high clusters, which reflect a trade-off between environmental conditions and accessibility. High-low clusters, characterized by favorable environmental conditions but poor accessibility, represent both a challenge and an opportunity for park management. While these regions offer high-quality ecosystems, their remoteness limits their current potential to contribute to human well-being, particularly in terms of ecosystem services. The key strategy in these areas should focus on enhancing transportation accessibility. While improving infrastructure (e.g., roads, parking areas, and public transportation access) is essential to increase the accessibility of these regions, such interventions must be carefully managed to prevent negative ecological consequences such as habitat fragmentation or the overexploitation of resources. Thus, any infrastructure improvements should be strategically planned to ensure that they do not compromise the ecological integrity of the area. Targeted improvements in accessibility can also facilitate greater public engagement and participation in conservation activities, which may lead to stronger local support for long-term sustainability efforts. Additionally, the promotion of eco-tourism or educational programs could help raise awareness of the region’s environmental value while maintaining ecological balance. Low-high clusters, which exhibit good transportation access but poor environmental conditions, present a more challenging scenario. These areas have suffered from intense human activity and unregulated development, resulting in degraded environmental quality. As such, the management approach in high-low clusters should be dominated by ecological restoration efforts. Ecological restoration in these regions is crucial to halt further environmental decline, which could otherwise impact the park’s internal ecosystem, particularly in areas with direct ecological linkages to the park. Restoration activities could include habitat rehabilitation, reforestation, and the installation of green infrastructure to enhance ecological resilience. Additionally, stricter land-use regulations and zoning measures should be introduced to limit further degradation and facilitate the restoration of ecological functions. This proactive approach can help mitigate the negative edge effects on the park, which are often exacerbated by human-induced pressures in nearby regions. Ultimately, the successful restoration of these areas could help bridge the ecological divide between the park and surrounding communities, fostering a more resilient landscape.

Limitations and future research prospects

While this study provides valuable insights into the recreational value of national park buffer zones, it is important to acknowledge several limitations. First, due to data availability constraints, the analysis relies on static LST and PM_2.5 datasets, which capture spatial variation but do not account for temporal dynamics. Future research should incorporate multi-year time series data to investigate how clustering patterns evolve over time, particularly in the context of climate change and urban expansion pressures. This longitudinal analysis could reveal significant trends in environmental quality and accessibility changes over time. A second limitation is the relatively low spatial resolution of the LST data from Landsat 8 OLI_TIRS and the PM_2.5 data from the China High Air Pollutant (CHAP) dataset. While these datasets are widely used, their inherent resolution limitations may impact the precision of our findings. Additionally, this study only focuses on summer daytime conditions, based on the assumption that most recreational visits occur during this period. Consequently, important diurnal and seasonal variations in temperature and air quality, as well as seasonal changes in accessibility patterns, may not have been captured. Future studies should prioritize higher-resolution data and incorporate both diurnal and seasonal cycles to provide a more comprehensive understanding of the year-round conditions. Another limitation of this study is its reliance on objective environmental and accessibility indicators. While we analyzed temperature, PM_2.5, travel time, and road networks, subjective human factors influencing recreational behavior were not included. Future research would benefit from the integration of multiple data sources, including geotagged social media data and survey responses. This mixed-methods approach would better capture visitor preferences, perceptions, and actual usage patterns, thereby enhancing the accuracy of the research.

Conclusions

National parks serve the dual function of conserving species and the integrity of natural environments, while also providing educational, scientific, cultural, and recreational opportunities that are suited to their surroundings⁶⁵. This dual role has garnered widespread global attention. While previous research has extensively discussed issues such as institutional frameworks, zoning, and ecosystem services within national parks, little attention has been paid to the buffer zones surrounding these parks and the realization of their recreational value. This study, taking the Baishanzu National Park in China as a case study, introduced a new framework that integrated environmental impacts and accessibility. Considering the park’s ecological integrity and the policy constraints it faces, this study focused on a comprehensive assessment of the park’s 30-km buffer zone, aiming to explore strategies and potential for achieving both ecological conservation and the realization of the park’s recreational value.

Our findings revealed that both LST and PM_2.5 concentrations exhibited a distinct gradient increase in the national park, its buffer zone, and the surrounding areas. Baishanzu National Park plays a significant role in cooling the surrounding environment and purifying the air. The calculations indicated that the park’s cooling area extended to 2,474.08 km², while the PM_2.5 reduction area covered 3,039.43 km². The park’s cooling intensity was 2.57 °C, and its PM_2.5 reduction intensity was 0.97 µg/m³. The environmental quality within the buffer zone was generally high, with 59.67% of the total environmental impact index exceeding 0.8. Moreover, the road network within the buffer zone is well-developed; however, the overall travel time cost for citizens remains relatively high. The accessibility index in most areas ranges from 0.78 to 0.86, accounting for 30.98% of the buffer zone’s total area. This suggested that the buffer zone benefited from the ecological spillover effect of the national park, which, while maintaining ecological integrity, also holed significant potential for sustainable tourism, eco-resorts, and health retreats.

Spatial autocorrelation analysis of environmental effects and accessibility revealed a negative correlation between the two variables, showing unique spatial clustering patterns. Synergy appeared in the high-high clusters, primarily located near the northern entrance of Longquan County’s national park. Antagonism manifested in the high-low and low-high clusters. The high-low clusters were distributed in the southeastern regions of Jingning County and Qingyuan County, while the low-high clusters concentrated in the western part of the buffer zone. Different planning strategies and policies should be implemented for each clustering pattern. Overall, this study significantly enhances our understanding of the benefits that national parks provide to both the environment and humanity, offering valuable guidance for the realization of their recreational value and sustainable development.