Introduction

Recreation ecosystem services contribute to both physical and mental well-being, enhancing human well-being (Wood et al., 2017). Cities, as significant ecosystems and densely populated areas, derive their Recreation Ecosystem Services primarily from public spaces such as parks, wetlands, and historical sites (Pukowiec-Kurda, 2022). For instance, urban green spaces, by providing areas for physical activity and recreation, aid in stress reduction and ultimately lower disease rates (Kondo et al., 2018). Historical sites enhance a city’s cultural services capacity, fostering cultural education and building cultural confidence. Moreover, the density, volume, spatial distribution, and accessibility of urban public spaces significantly influence human well-being. For example, high-density built environments often create a sense of spatial oppression, negatively impacting residents’ mental health (Chung et al., 2022), while undeveloped areas and disconnected road networks can contribute to the formation of impoverished communities (Luo and Jiang., 2022). Nodira Mamajonova et al. highlights the contribution of green spaces in alleviating urban challenges, enhancing ecological resilience, promoting physical and mental health, fostering social cohesion, and stimulating economic prosperity (Mamajonova et al., 2024). Extensive research demonstrates that individuals are more inclined to seek recreation and enjoyment (Recreation Ecosystem Services) in green spaces (Liang et al., 2023). Therefore, urban green spaces, as a crucial element of urban residents’ well-being, provide a valuable subject for recreation ecosystem service research, offering insights for urban planning and policy development.

In urban environments, spatial mismatches often arise between the distribution of recreation service supply areas and human demand areas due to variations in green space coverage and population distribution (Costanza, 2008). Recreation ecosystem services, however, can be transferred and delivered in the form of “flows” between supply and demand zones, effectively mitigating this spatial mismatch (Zhou et al., 2022). This has propelled the assessment of recreation ecosystem service flows (RESF) to the forefront of international academic research. Currently, the evaluation of RESF primarily encompasses two aspects: supply and demand assessment, and the assessment of flow processes.

The supply of RESF is influenced by both environmental and social factors, including climate, vegetation, water bodies, recreation facility size, infrastructure, and accessibility (Zhong et al., 2019; Zhang et al., 2021). Currently, researchers primarily employ two approaches to assess RESF supply: a potential and actual assessment based on valuation methods (Liu et al., 2021; Peng et al., 2017), public participation methods (Cabana et al., 2020; Wang et al., 2021), and model simulations (Zhao et al., 2023, Wu et al., 2022). RESF demand, on the other hand, is influenced by sociodemographic data and individual subjective factors, such as population size, population density, individual preferences, and desires (Wolff et al., 2015). Currently, demand research remains limited (Shi et al., 2020; Liu et al., 2020) and typically relies on indicators such as recreation preferences or willingness to pay (Liyan et al., 2023), population data, or socio-economic indicators combined with population data (Wu et al., 2022), and social media data (Meng et al., 2020) for demonstration.

Building upon research on RESF supply and demand, numerous scholars have investigated the flow processes that connect these two aspects (Syrbe and Walz, 2012; Wang and Hayashi, 2023; Li et al., 2022). These studies often assess the spatial characteristics of flows in terms of their boundaries, direction, speed, and quantity (Zapata-Caldas et al., 2022). For example, Liu et al. identified the boundaries of park cultural ecosystem service flows in Shenzhen City using spatial autocorrelation analysis (Zheng et al., 2023). quantified the quantity of RESF from farmland in Hangzhou City based on visitor numbers (He et al., 2019). Inácio M et al. utilized modeling and participatory mapping to identify and visualize the direction and pathways of recreation service flows along the Lithuanian coastline (Inácio et al., 2022). However, current research often focuses on only select attributes of RESF (Fagerholm et al., 2021; Thiele et al., 2020), lacking a systematic description of the entirety of the RESF. This hinders a comprehensive understanding of its characteristics.

This study, building upon existing research on supply and demand, develops a framework for characterizing RESF, incorporating four key attributes: boundaries, quantity, direction, and speed. Additionally, the temporal dynamics of RESF have been relatively understudied. Existing research predominantly concentrates on specific years, such as Inácio et al.’s evaluation of recreation service supply, path, and demand along the Lithuanian coast in 2020. Johannes Hermes et al. assessed the flow quantity and revenue of RESF in Germany for the year 2021. Although a few studies have explored RESF over extended periods, a comparative analysis of flow changes across multiple time scales remains scarce (Zhao et al., 2022; Perschke et al., 2023).

In summary, while significant progress has been made in assessing the RESF, there remain limitations in the comprehensiveness of flow analysis and the depth of spatiotemporal change research. This study, therefore, focuses on the main urban area of Zhengzhou City, using 2010, 2015, and 2020 as study periods, to investigate the following: (1) A framework for RRESF, encompassing supply, demand, supply-demand relationships, quantity, direction, boundaries, and speed; (2) Spatiotemporal variations and mapping analysis of RESF. Based on these findings, the study proposes recommendations for urban management in relation to RESF.

Materials and methods

Study area

Zhengzhou City, Henan Province is located in the transitional zone between Funiu Mountain and Huanghuai Plain (112 ° 42′–114 ° 14’E, 34 ° 16′–34 ° 58’N), belongs to the warm temperate continental monsoon climate with distinct seasons and simultaneous rainfall and heat. The precipitation in Zhengzhou is slightly lower, with an average annual rainfall of about 542.15 mm. Zhengzhou city spans two major river basins, the Yellow River and the Huai River, and has abundant animal and plant resources.

In recent years, driven by policies such as the rise of the central region, the construction of the Central Plains Economic Zone, the development of the Central Plains urban agglomeration, and the construction of the national central city, the urbanization process in Zhengzhou has been rapid. As of 2020, the urbanization rate of Zhengzhou City reached 78.4%, with a permanent population of 6.4524 million in the main urban area. In the context of sustained socio-economic growth, entertainment activities have become an important component of residents’ lives. The construction of entertainment facilities and developed transportation conditions in Zhengzhou have also provided convenience for residents’ entertainment activities. This study takes the main urban area of Zhengzhou City as the research object. The research area extends from the Yellow River in the north to the Xinmi and Xinzheng borders in the south, to the Zhongmou border in the east, and to the Xingyang border in the west. It mainly includes Erqi District, Jinshui District, Zhongyuan District, Huiji District, and Guancheng District, with a total area of 1.06 × 105 hm2. The specific research scope is shown in Fig. 1.

Fig. 1: Study area.
figure 1

a Henan province in China; b Study area in Henan province; c Study area.

Full size image

Data source

The data comes from geographic spatial data cloud (https://gscloud.cn/search) Natural Resources Satellite Remote Sensing Cloud Service Platform (http://sasclouds.com/chinese/normal/). This includes three periods: 2010, 2015, and 2020. The data mainly comes from 30 m × 30 m Landsat series and 8 m × 8 m GF-1 remote sensing image data. Combining high-definition Google Maps data and on-site research from the summer and winter seasons of Zhengzhou’s main urban area in 2010, 2015, and 2020, ArcGIS software was used for visual interpretation. The final land use data obtained has a resolution of 10 meters. Land use types are divided into grassland, cultivated land, construction land, forest land, watershed, and unused land (Table 1) (Fig. 2). Road data (https://data.maptiler.com/downloads/dataset/osm/asia/china/#9.24/34.668/113.6167) downloaded from China’sroad network OpenStreetMap. The population density data is sourced from Worldpop (https://hub.worldpop.org/geodata/listing?id=29). The POI data is sourced from the Gaode MapAPI (https://lbs.amap.com/api/ios-sdk/guide/map-data/poi/). This study integrates multi-source geospatial data through a 1 km × 1 km grid to evaluate the supply and demand of entertainment services.

Table 1 Land use types.
Full size table
Fig. 2: Land use types.
figure 2

a 2010; b 2015; c 2020.

Full size image

Research procedure

Drawing upon previous research, a framework for RESF research was established, encompassing the following key elements: supply, demand, supply-demand relationships, quantity, direction, speed, and boundaries. Within this framework,

(1) Supply

The supply of Recreation Ecosystem Services (RES) includes potential supply (RPS) and opportunity supply (ROS) (Maes et al., 2012; Peña et al., 2015).

(2) Demand

Demand is related to population size (Hooftman et al., 2023), and human demand is measured using residential density.

(3) Supply and demand relationship

Evaluate the supply and demand relationship of recreation ecosystem services (RES) through supply-demand difference calculation and analysis of hot and cold spots. The supply-demand difference can identify the surplus and deficit status of RES supply and demand. Cold and hot-spot analysis is used to determine the supply and demand areas of RES, where hot spots are considered the supply areas and cold spots are considered the demand areas.

(4) Flow direction

The flow direction refers to the direction in which Recreation Ecosystem Services Flow (RESF). The beneficiary travels from region B to region A to enjoy RES, resulting in the flow of RES from region A to region B. Therefore, the direction of RESF is opposite to that of population flow. For example, Zone A is the supply zone, Zone B is the demand zone, and Zone C is the supply-demand balance zone. Not only do RESF from Zone A to Zones B and C, but there are also RES Flowing from Zones B and C to Zone A.

(5) Flow quantity

The number of recreation service beneficiaries represents flow quantity. Non-cumulativeness is a significant characteristic of recreation service flows. If area A has a supply zone, area B has a demand zone, and area C has a demand zone with demand exceeding that of area B, Recreation Ecosystem Services flowing from area A to area B, even if exceeding the demand of area B, will typically not continue to flow to area C. Instead, based on the probability of visitor choice, they will flow directly from area A to area C.

(6) Flow speed

Flow speed refers to the distance traveled by RES per unit time when they flow between two areas. Since the main body of the RESF is the beneficiary, the speed of human travel is the speed of the RESF.

(7) Flow boundary

The boundary of RESF refers to the scope of RESF. As the distance increases, the attractiveness of places providing RES gradually decreases. The main urban area of Zhengzhou has convenient transportation and easy travel for beneficiaries. The boundary of RESF can cover a wide range of the main urban areas. The specific process we studied is shown in Fig. 3, and based on the research results, we proposed suggestions for urban RESF management.

Fig. 3
figure 3

Recreational ecosystem services flow framework.

Full size image

Research methods

RES supply and demand assessment

(1) Supply Assessment

This study evaluates both RPS and ROS using six indicators: (1) naturalness, (2) infrastructure, (3) landscape diversity index, (4) water resources, (5) distance to roads, and (6) distance to scenic spots (Table 2) (Bing et al., 2021).

Table 2 Indicator system for the supply of recreational services.
Full size table

To ensure comparability between supply and demand data, normalize the results to a range of 0 to 1.

(2) Demand assessment

Residential density was utilized as a proxy for human demand. By extracting locations of dormitories, villas, and residential areas from POI data, a search radius was set, and the kernel density tool in ArcGIS was employed to calculate the density of residential points. Finally, the results were normalized to a scale of 0 to 1.

(3) Supply and demand relationship assessment

Analyze the supply and demand relationship of RES using the supply-demand difference. The supply-demand relationship is represented by the supply-demand difference of each grid (Eq. 1).

$${ESDR}={ESS}-{ESD}$$
(1)

Where, ESDR is the supply-demand difference of each grid, ESS is the RES supply of each grid, and ESD is the RES demand of each grid. When ESDR is less than 0, it means that supply is less than demand, which is a deficit. When ESDR is 0, it indicates a balance between supply and demand. When ESDR is greater than 0, it indicates that supply exceeds demand, which is a surplus.

Based on the evaluation of the supply-demand difference, the supply and demand areas of RES are determined through cold and hot-spot analysis, where the hot-spot is considered the supply area and the cold spot is considered the demand area. The formula for cold and hot-spot analysis is as follows (Eq. 2).

$${{Gi}}^{* }=\frac{\mathop{\sum }\nolimits_{j=1}^{n}{w}_{{ij}}{x}_{j}-\bar{x}\mathop{\sum }\nolimits_{j=1}^{n}{w}_{{ij}}}{\sqrt{\frac{\mathop{\sum }\nolimits_{j=1}^{n}{x}_{j}^{2}}{n}-{x}^{-2}}\sqrt{\frac{\left[n\mathop{\sum }\nolimits_{j=1}^{n}{w}_{{ij}}^{w}-{\left(\mathop{\sum }\nolimits_{j=1}^{n}{w}_{{ij}}\right)}^{2}\right]}{n-1}}}$$
(2)

When Gi* > 0, the study area is a hot spot. When Gi* < 0, the study area is a cold spot. The Z- and P-values in the hotspot analysis reflect where spatial clustering of high- or low-value elements occurs.

Process evaluation of recreation ecosystem service flow (RESF)

(1) Flow quantity

When an area’s ecosystem services cannot meet its own needs, it is more likely to seek compensation from neighboring areas. Meanwhile, when the region has surplus ecosystem services after its own consumption, it usually prioritizes compensating for areas closer in distance.

In addition, research has shown that high-quality entertainment service destinations are more attractive to beneficiaries (Luo et al., 2022). Therefore, we use a selection probability (Hough model) that includes multiple factors such as distance and geological quantity of entertainment service destinations to express the preferences of beneficiaries. Based on previous research, the probability of beneficiaries choosing green spaces is relatively consistent with the flow direction of entertainment services, which can be used to determine the flow direction (Xu et al., 2024) (Eq. 3).

$${P{rob}}_{{ij}}=\frac{{P}_{j}{G}_{{ij}}}{{\sum }_{j=1}^{n}{P}_{j}{G}_{{ij}}}$$
(3)

Where, is the probability Probij of a visitor from region i to region j, and is the quality of region j (expressed here in terms of green space and water area), and is the distance attenuation from region i to region j.

(2) Flow direction

The main urban area of Zhengzhou has developed transportation and convenient travel for tourists, so it is believed that tourists from all regions can reach it. The distance attenuation is represented by a Gaussian function (Eq. 4).

$${G}_{ij}=\left\{\begin{array}{ll}\frac{{e}^{-\left(\frac{1}{2}\right)\times {\left(\frac{{d}_{{ij}}}{{d}_{0}}\right)}^{2}}-{e}^{-\left(\frac{1}{2}\right)}}{1-{e}^{-\left(\frac{1}{2}\right)}}\\ 0,{if}{d}_{{ij}} > {d}_{0}\end{array}\right.,{if}{d}_{{ij}}\,\le\,{d}_{0}$$
(4)

Among them, is the dij from region i to region j, and is the d0 value from region i to region j.

Using population size to represent RES quantity, calculate as follows (Eq. 5):

$${F}_{{ji}}={R}_{i}{{Prob}}_{{ij}}$$
(5)

Among them, Fji represents the flow quantity from region j to region i, Ri represents the population of region i, and Probij represents the probability of tourists from region i visiting region j.

(3) Flow speed

We referred to previous studies (Virgl et al., 2017; Xiao et al., 2019; Yang and Cao, 2022). Considering the traffic conditions in the main urban area of Zhengzhou, three different modes of transportation have been determined: walking, cycling, and driving. The speeds for these three modes were set at 4 km/h, 15 km/h, and 45 km/h respectively.

(4) Flow boundary

According to previous research (Guo et al., 2022), we set travel times of 15 min, 30 min, and 60 min. By analyzing factors such as the distribution of RES facilities, population density, and transportation networks, the travel ranges of walking, cycling, and driving are calculated to determine the boundary range of RESF.

Results

Recreation ecosystem services (RES) supply, demand, and supply-demand relationship

RES supply assessment

From 2010 to 2020, the distribution of RPS in the main urban area of Zhengzhou City showed a gradual increase in high values (0.20–0.40) and a gradual decrease in low values (0.20–0.40), with the distribution of high-value areas tending to cluster. The distribution of ROS high values gradually increases, while the low values (0–0.20) gradually decrease. The distribution of high values becomes more concentrated in the city center, and the differences in opportunity supply between different regions in the main urban area gradually decrease.

In 2010, the total supply of RES in the main urban areas of Zhengzhou ranged from 0 to 0.80, with a main range of 0.20 to 0.60 (Fig. 4). The low values were mainly distributed in the west, south, and at edge of the main urban area, while the high values were scattered in different locations. In 2015, the total supply of RES in the main urban area increased compared to that in 2010 (Fig. 5). The distribution of low values was concentrated in the west and south. The regions with a total supply value between 0.40 and 0.60 increased more than in 2010, and the growth mainly took place in the northeast and southwest. In 2020, the total supply of RES in the main urban area as a whole was between 0 and 0.80, mainly ranging from 0.40 to 0.60. This represented an increase in supply capacity compared to 2015 (Fig. 6). The low values were concentrated in the northeast region. The regions with a total supply value between 0.40 and 0.60 increased more than in 2015, and the growth mainly occurred in the northwest, northeast, and south regions. The regions having a total supply value ranging from 0.40 to 0.60 have witnessed significant growth compared to 2015, with the growth predominantly taking place in the northwest, northeast, and south.

Fig. 4: Recreational ecosystem services supply in 2010.
figure 4

a RPS; b ROS; c total supply.

Full size image
Fig. 5: Recreational ecosystem services supply in 2015.
figure 5

a RPS; b ROS; c total supply.

Full size image
Fig. 6: Recreational ecosystem services supply in 2020.
figure 6

a RPS; b ROS; c total supply.

Full size image

Over the past decade, urbanization has advanced rapidly, accompanied by an augmented effort in ecological civilization construction. The green infrastructure in the main urban area has continuously increased, the area of ecological land has expanded, and the construction of green space has become more systematic. However, owing to the functions of the economic, political, and cultural center in the main urban area, the blue-green spaces in the city exhibit block-like and belt-like forms and are distributed in a scattered fashion, thereby resulting in a scattered distribution of recreational service supply areas.

RES demand assessment

The development of recreational activities is inseparable from human beings and society. The evaluation of RES demand based on residential area density is shown in Fig. 7. From 2010 to 2020, the demand for RES in urban centers increased, and the scope of demand expanded, presenting a spatial mismatch with the supply of RES. The demand for RES was concentrated in the urban center, especially in the northeast, northwest, southwest, and southeast of the main urban area. Conversely, the demand was less in the urban fringe.

Fig. 7: Recreational ecosystem services demand.
figure 7

a 2010; b 2015; c 2020.

Full size image

In 2010, the demand for RES was mainly between 0 and 0.15. The high value of demand (0.60 to 1) was relatively low, and the distribution was scattered. In 2015, the demand for RES was mainly between 0 and 0.40. The high demand value (0.60 to 1) increased, and the distribution was more concentrated than in 2010. In 2020, the demand for RES was mainly between 0 and 0.40. The high demand value (0.60 to 1) increased compared with 2010 but had only a small increase compared with 2015. The main urban area of Zhengzhou is dominated by plains, with the north side adjacent to the Yellow River and a small amount of hilly areas in the northwest and southwest. The natural environmental conditions have laid the foundation for the rapid urban expansion of the main urban area of Zhengzhou. The land use in the main urban area has become saturated, and the demand for RES is constantly increasing.

Recreation ecosystem service supply-demand relationship assessment

(1) Analysis of the supply-demand difference in RES

In the main urban areas of Zhengzhou in 2010, 2015, and 2020, there was a significant between supply and demand. The surplus areas gradually expanded, and the deficit areas were more concentrated in the urban center (Fig. 8). The surplus areas of RES were larger than the deficit areas. The supply-demand difference was greater in the surplus areas and smaller in the deficit areas. In the deficit areas, the supply-demand difference was mainly distributed between −0.60 and 0. While in the surplus areas, the supply-demand difference was mainly distributed between 0 and 0.79.

Fig. 8: Recreational services supply and demand.
figure 8

a 2010; b 2015; c 2020.

Full size image

(2) Analysis of the supply-demand regions of RES

From 2010 to 2020, the distribution of RES hot spots in the main urban area of Zhengzhou expanded from east to west, and some hot spots in the northeast gradually disappeared. In addition to the city center, the distribution of cold spots also appears at the edge of the main urban area, and the cold spots in the west gradually disappear. In order to present the flow of RES more intuitively, the main urban area of Zhengzhou was divided based on the analysis of hot and cold spots. The hot spots in 2010, 2015, and 2020 were divided into g1-g4 zones, the cold spots were divided into d1-d6 zones, and the insignificant zones were divided into n1-n5 zones (Fig. 9).

Fig. 9: Hot and cold spot zone.
figure 9

a 2010; b 2015; c 2020.

Full size image

Evaluation of the flow process of recreation ecosystem service flow (RESF)

RESF quantity and direction assessment

(1) Probability of beneficiary arrival

From 2010 to 2020, in the main urban area of Zhengzhou city, RES beneficiaries had the highest probability of reaching the hot-spot area with the largest RES, followed by the non- significant area with the largest size.

In 2010, the arrival probability of the beneficiaries of RES was between 0 and 5% (Fig. 10a). For the beneficiaries of RES, the G4 area had the highest arrival probability (the probability of all areas reaching G4 was more than 3%). The arrival probability of the beneficiaries of RES in the d5 and n3 areas was less than 0.03%. The arrival probability of beneficiaries in the D6 area was 0. This was because the D6 area was located at the edge of the main urban area. It was small in size, and its land use type did not contain ecological land.

Fig. 10: The arrival probability of beneficiaries by areas.
figure 10

a 2010; b 2015; c 2020.

Full size image

In 2015, the arrival probability of the beneficiaries of RES was between 0 and 7.30% (Fig. 10b). The arrival probability of recreation service beneficiaries in the g3 area was the highest (the probability of all areas reaching g3 was more than 4.90%). The arrival probability of the beneficiaries of the RES in g1, g2 and d1 areas was less than 0.02%.

The probability of the arrival of the beneficiaries of RES in 2020 was between 0 and 8.10% (Fig. 10c). The arrival probability of recreation service beneficiaries in the g1 area was the highest (the probability of all areas reaching g1 was more than 5%). The probability of the beneficiaries of RES in the n1 area was less than 0.02%. The arrival probability of beneficiaries in the n3 area was 0. This was because the n3 area was located at the edge of the main urban area, the area was small, and the land use type did not contain ecological land.

(2) RESF quantity and direction

The RESF is bidirectional, and the flow of RES in each supply and demand area is divided into two directions: inflow and outflow. Red represents inflow, and blue represents outflow. From 2010 to 2020, the number of areas with a two-way flow of RES decreased, and the areas with the largest inflow were usually the cold spot area and hot-spot area, while the areas with the largest outflow were usually the hot-spot areas (Fig. 11).

Fig. 11: Quantity and direction of recreational ecosystem services flow.
figure 11

Inflow (a–c); outflow (d–f).

Full size image

In 2010, except for d5-g2, g3-d5, g1-d5, and g2-g1, the flow of RES was a two-way flow between two areas (Fig. 11a, d). The area with more inflow was d3. The area with more outflow was g4. The flow of d3 and n1 areas, d3 and g4 areas, d3 and n2 areas, and d3 and d4 areas had great changes with the change of direction. The inflow (n1-d3) quantity was 61,684 people, while the outflow (d3-n1) quantity was 7318 people. The inflow (g4-d3) quantity was 110,842 people, while the outflow (d3-g4) quantity was 47 people. The inflow (n2-d3) quantity was 12,018 people, while the outflow (d3-n2) quantity was 18 people. The inflow (d4-d3) quantity was 7832 people, while the outflow (d3-d4) quantity was 357 people.

In 2015, except for d2-g1, d4-n2, d1-n2, d1-d5, and d2-n2, the flow of RES was a two-way flow between two areas (Fig. 11b, e). The areas with more inflow were d3 and g3, and the area with more outflow was g3. The flow of d3 and g3 areas, d3 and n1 areas, and d3 and n2 areas had great changes with the change of direction. The inflow (g3-d3) quantity was 218,346 people, while the outflow (d3-g3) quantity was 12,611 people. The inflow (n1-d3) quantity was 120,356 people, while the outflow (d3-n1) quantity was 12077 people. The inflow (n2-d3) quantity was 2444 people, while the outflow (d3-n2) quantity was 13 people.

In 2020, except for n3-g1, n3-d2, n3-n2, n3-n5, d3-d1, n1-d1, n1-n2, the flow of RES was a two-way flow between two areas (Fig. 11c, f). The areas with large inflow were d2 and g1. The area with a large outflow was g1. The flow in the d2 and n4 areas, g1 and n2 areas, d2 and n2 areas, and d2 and d3 areas had great changes with the change of direction. The inflow quantity from n4 to d2 was 24,440 people, while the outflow quantity from d2 to n4 was 616 people. The inflow (n2-g1) quantity was 7159 people, while the outflow (g1-n2) was 293 people. The inflow (n2-d2) quantity was 21,277 people, while the outflow (d2-n2) quantity was 1 person. The inflow (d3-d2) quantity was 3852 people, while the outflow (d2-d3) quantity was 122 people.

RESF supply and boundary

As can be observed from Fig. 12a–c in 2010, the boundaries of the areas accessible by walking at a speed of 4 km/h were predominantly located in the center and north of the main urban area, along with areas such as d1, n2, n3, d4, and d6 in the east and south. The flow speed of RES after a 60 min walk was still zero. The boundary of the area is reachable by cycling at a speed of 15 km/h and nearly covers the entire main urban area, with only the edge areas of g1, d1, and d5 having zero flow velocity. The boundary of the area reachable at a speed of 45 km/h nearly covers the entire main urban area. In 2010, the urban construction of Zhengzhou was relatively incomplete. Green spaces are mainly distributed in the city center and northern areas, which can be reached within a 30 min walk.

Fig. 12: Speed and boundary of recreational ecosystem services flow.
figure 12

a 2010 walking, b 2010 cycling, c 2010 driving, d 2015 walking, e 2015 cycling, f 2015 driving, g 2020 walking, h 2020 cycling, j 2020 driving.

Full size image

As can be observed from Fig. 12d–f in 2015, the boundaries of areas accessible at a walking speed of 4 km/h expanded compared to 2010, with additional locations not only at the city’s periphery but also in the city center. Particularly in the eastern and southern parts of the main urban area, there are more areas that can be reached within 60 min on foot. The boundary of the area is reachable by cycling at a speed of 1 km/h nearly covers the entire main urban area, with only a few areas on the periphery of the main urban area, such as n2 and d6, having zero flow velocity. Compared to 2010, the number of areas reachable within 30 min by cycling has increased. The boundary of the area reachable at a speed of 45 km/h nearly covers the entire main urban area. Compared to 2010, the number of areas reachable within 15 min by car has increased.

As can be observed from Fig. 12g, h, j in 2020, the boundaries of the areas accessible by walking at a speed of 4 km/h have expanded compared to 2015, with more locations situated in the eastern part of the main urban area. The area reachable within 15 and 30 min on foot has increased significantly. The boundary of the area is reachable by cycling at a speed of 15 km/h nearly covers the entire main urban area. Compared to 2015, there are more areas reachable within 15 min by cycling. The boundary of the area reachable at a speed of 45 km/h essentially covers the entire main urban area.

Discussion

Framework for RESF

The framework’s construction facilitates a systematic analysis of the entire RESF process. Previous studies often focus on individual attributes of RESF, such as supply and demand (Dang and Li, 2023；Xu and Wang, 2023), quantity (Vallecillo et al., 2019), and direction (Hermes et al., 2024). This study systematically analyzes the complete set of attributes of RESF, encompassing supply, demand, quantity, direction, boundaries, and speed. By employing this framework across multiple spatial and temporal scales, a comprehensive understanding of the overall RESF status in a region can be achieved. For example, regarding supply and demand, the areas with surplus RES supply in Zhengzhou City’s main urban area have gradually increased between 2010, 2015, and 2020, while deficit areas have become more concentrated in the city center. This spatial mismatch in the supply and demand of RES in the urban core cannot meet the needs of the population. Hotspot analysis reveals that the distribution of RES hot spots in Zhengzhou City’s main urban area has expanded from east to west between 2010 and 2020, with hot spots in the northeast gradually disappearing, reflecting the city’s evolving spatial orientation of RES. In terms of boundaries and speed, the distribution of green spaces in Zhengzhou City in 2010 was primarily concentrated in the urban center and north, with these areas being accessible within a 30 min walking radius. In 2015, the area accessible within a 30 min cycling radius expanded. By 2020, the boundary of the area reachable by car at 45 km/h encompassed almost the entire main urban area.

RESF research and urban construction

The analysis based on the framework for RESF not only provides a deeper understanding of the dynamic changes in service flows within Zhengzhou City’s main urban area but also offers decision support for optimizing urban layout and enhancing urban construction.

Regarding supply and demand, our findings align with studies conducted by Zhenhua Bing (Bing et al., 2021) and Yue Deng (Deng et al., 2020), reflecting the current state of recreation service supply and demand in the city. These findings provide targeted recommendations for urban construction from a supply and demand perspective. However, previous studies have not addressed the impact of flow direction and speed on urban development. This research expands upon this aspect, as detailed below. In terms of flow quantity and direction, the areas with the highest inflow in 2010, 2015, and 2020 were, respectively, D3, D3, and D2, all located in the city center (Fig. 11). Given the limited availability of developable green spaces in the city center, these areas require an increase in recreation service outflow. Specifically, it is recommended to establish green spaces around areas D3, D3, and D2 to enhance recreation service outflow.

In terms of speed and boundaries (Fig. 12), different travel modes can impact the accessibility of RES based on their respective velocities and reachable boundaries. Based on accessibility analysis, specific optimization measures can be identified. For example, although walking speed is 4 km/h, the accessible RES areas are concentrated and limited, primarily distributed in the city center and parts of the northern and southern areas. It is recommended to optimize transportation routes in conjunction with RES paths, enhancing pedestrian accessibility in the eastern and western parts of the city.

Conclusions

The framework facilitates a systematic description of the complete characteristics of RESF

This framework enables a comprehensive understanding of the full spectrum of RESF in Zhengzhou City. From 2010 to 2020, the areas with a surplus in RES supply in the city’s main urban area have gradually increased, while deficit areas have become more concentrated in the city center, highlighting a spatial mismatch between supply and demand. Bidirectional flow areas of recreation ecosystem services in Zhengzhou City have decreased. Areas with the highest inflow are D3, D3, and G3, as well as D2 and G1; areas with the highest outflow are G4, G3, and G1. Accessibility of RES varies significantly across different flow velocities.

RESF evolves with urbanization

Urbanization has brought about changes in population density and distribution, as well as in the quantity and distribution of green spaces, leading to shifts in both the supply and demand of RESF. Between 2010 and 2020, the total supply of RES in Zhengzhou City’s main urban area increased from 0.2–0.6 to 0.4–0.6, while demand primarily increased from 0–0.15 to 0–0.4. Areas with a surplus in supply outweighed areas with a deficit, and the difference between supply and demand in surplus areas gradually widened (0–0.79). The probability of RES beneficiaries reaching their destination has gradually increased. Bidirectional flow areas have decreased in number, and flow paths have expanded in the eastern and southern parts of the main urban area. The walking radius of accessible areas has also gradually expanded.

RESF assessment results provide decision support for urban planning

The assessment of RESF plays a crucial role in urban planning. By evaluating the supply and demand, boundaries, quantity, direction, and speed of RES flows, a comprehensive understanding of RES supply and demand dynamics can be achieved. This information provides a multi-dimensional basis for urban planning. For example, supply and demand analysis can inform the quantity of green space allocation, while flow direction and quantity analyses offer additional data for green space distribution. The analysis of flow speed and boundaries can ensure that green space distribution meets the needs of diverse user groups.