Introduction

With the acceleration of global urbanization and the fast pace of modern life, the mental and physical health of urban residents has become an increasing concern. Research indicates that urban layout, green spaces, and air quality significantly influence residents’ mental health and social behavior. In the book “Digital Ethology: Human Behavior in Geospatial Context,” particularly in the chapter “How Cities Influence Social Behavior,” these factors’ profound impacts on residents’ lives are explored (Balsa-Barreiro and Menendez 2024). Well-designed urban green spaces optimize spatial use and significantly enhance residents’ happiness and sense of belonging by shaping behavior patterns and social interactions (Jacobs 1961).

Community parks, integral components of urban green spaces, offer opportunities for connecting with nature and recreation, while fostering social interactions and enhancing social cohesion (Maas et al. 2006). Additionally, studies highlight a positive link between urban green space accessibility and residents’ social behavior, underscoring the importance of community parks in enhancing interpersonal relationships and reducing stress (Takano et al. 2002).

With aging populations and rising health concerns, the design and application of rehabilitation landscapes in community parks have become a key focus in urban planning (Wang et al. 2022). Rehabilitation landscapes now focus less on aesthetics and recreation and more on enhancing residents’ mental health and social well-being through targeted design.Recent studies show that rehabilitation landscapes effectively reduce psychological stress, promote social interactions, and improve community residents’ quality of life (Putnam and Quinn 2007).

The concept of “rehabilitation landscape” originated from the concept of restorative environment proposed by Stephen Kaplan and Jane Talbot at the University of Michigan in 1983 (Zhang 2023). It is different from active rehabilitation methods and aims to help people restore their physical and mental health (Gómez et al. 2015). Rehabilitation landscapes achieve the goals of reducing patients’ physical and mental stress, enhancing their objective understanding of their own diseases, and improving public understanding of health by providing pleasant environments, conveying rehabilitation information, and attracting public attention (Plunkett et al. 2019). Community park rehabilitation landscapes mainly refer to natural or artificial environments in community parks that help residents achieve the effects of recovery or maintenance of psychological, physical, and social health through promoting physical activities, stimulating sensory experiences, and releasing beneficial substances. Such environments create positive conditions that enable residents to enjoy the beauty of nature, exercise, relieve stress, and promote overall health and rehabilitation (Peters et al. 2010; Bazrafshan et al. 2021; Chen 2020; Chen et al. 2013; Nieuwenhuijsen 2021; Liu et al. 2020). The classification of community park rehabilitation landscapes mainly includes: (1) sensory, meditation, and rehabilitation park landscapes based on the nature of the landscape; (2) observation, participation, and experiential park landscapes based on the user’s role (Zhu et al. 2017). Therefore, in future urban park construction, the rehabilitative landscape of community parks will become a new development trend, further enhancing the demand and attention of urban residents to health by providing diversified rehabilitation facilities and activities.

In recent years, there has been an increasing exploration of community park rehabilitation landscapes by scholars (Xu et al. 2019). Research on community park rehabilitation landscapes can be traced back to the sleep gardens used for therapy in ancient Greece and Rome (Sirina et al. 2017; Riechers et al. 2020; Peters et al. 2010; Peng 2020; McCormack et al. 2010; Matarrita-Cascante et al. 2010; Matsunaga 2010). In the 18th and 19th centuries, with the combination of Romanticism and medicine, Europe witnessed the revival of rehabilitation gardens (Matsuoka and Kaplan 2008; Chang et al. 2020). In the late 20th century, the reconstruction of the Yolanda Schenke Memorial Garden and the Elizabeth and Nona Evans Rehabilitation Garden designed by Deertree Walks provided important practical experiences for the practice of rehabilitation garden landscapes (Gao et al. 2021; Feng and Zhao 2020; Wu and Li 2019; Zheng et al. 2018; Zhu et al. 2018; Yan and Chen 2023; Pope et al. 2016; Zhou and Yan 2022). In recent years, about 20% to 30% of hospitals in the United States have built rehabilitation gardens with horticultural therapists for complementary treatment (Chen et al. 2021; Liu et al. 2021). It can be seen that foreign scholars’ research mainly focuses on public response to rehabilitation landscapes, the influence of environmental factors on public rehabilitation, and the impact of rehabilitation landscapes on the environment. In the domestic academic field, research on community park rehabilitation landscapes mainly focuses on aging-friendly perspectives, supply-demand perspectives, and multi-sensory experiences (Gao et al. 2021; Feng and Zhao 2020; Wu and Li 2019; Zheng et al. 2018; Zhu et al. 2018). Currently, there is limited research on the evaluation of community park rehabilitation landscapes. Scholars have proposed the Analytic Hierarchy Process (AHP) (Yan and Chen 2023) as an evaluation method, but the research methods are relatively single, and the evaluation indicators mainly focus on aesthetics, lacking scientific accuracy. It is urgent to establish a scientifically reasonable evaluation method for community park rehabilitation landscapes. The Grey Statistical Method (GST) offers a more precise method for selecting evaluation indicators, which improves the effectiveness of the evaluation model. Integrating GST with the Analytic Hierarchy Process (AHP) enables the optimization of the conventional AHP model, resulting in a more scientifically rigorous and rational evaluation framework (Pope et al. 2016; Zhou and Yan 2022; Ni et al. 2024).This study focuses on how to scientifically and accurately evaluate community park rehabilitation landscapes and proposes corresponding optimization suggestions to improve the design of community park rehabilitation landscapes. The study utilizes the GST-AHP combination evaluation model to quantify the indicator factors and landscape indices of various rehabilitation landscapes, aiming to improve residents’ daily leisure activities and psychological well-being. The Fig. 1 displays the flowchart of the current study.

Fig. 1
figure 1

Flowchart of the current study.

Full size image

Research methodology

Grey system theory

GST (Grey System Theory) is a statistical method for dealing with uncertainty in handling small amounts of data and information. It was first proposed by Professor Ju-Long Deng from Huazhong University of Science and Technology, China, at an international economic conference in 1982. This theory is a statistical approach that applies whitening functions to perform function operations and statistics. In situations where the system model is unclear and information is incomplete, GST processes evaluation data from a certain level of the system mathematically to capture trends and relationships within the system at higher levels. It effectively addresses issues where a large amount of unknown information cannot be incorporated into the model framework. Whitening functions analyze given data through whitening statistical quantities, reflecting the degree of certainty about the research object. Grey statistical methods use whitening functions to process evaluation data provided by experts, segmenting the data to calculate whitening function values for evaluating the importance of indicators as high, medium, or low. Based on these calculations, grey decision coefficients and grey decision vectors are computed to select indicators without the need for subjective assignment.

The general steps are as follows:

  1. (1)

    Distribute questionnaires to experts, who rate the importance of indicators using the Likert 7-point scale method, where 1 indicates “very unimportant” and 7 indicates “very important”, with 2–6 representing degrees of importance between the two extremes.

  2. (2)

    Construct three-level whitening functions for high, medium, and low levels of importance respectively based on the formula of grey correlation degree, and calculate the whitening function values \({f}_{k}(ab)\) corresponding to each indicator for high, medium, and low levels. Here, \(k=1,2,3\) and \({f}_{k}(ab)\) represent the whitening function values corresponding to the b-th indicator with importance level a under the high, medium, and low levels.

  3. (3)

    Calculate the grey decision coefficients for high, medium, and low levels respectively:

    $${\eta }_{k}(b)=\mathop{\sum }\limits_{a=1}^{7}L(ab)\times {f}_{k}(ab),k=1,2,3$$

    And the grey decision vector:

    $$\{{\eta }_{1}(b),{\eta }_{2}(b),{\eta }_{3}(b)\}$$
  4. (4)

    Determine the importance level of each indicator based on the values of the grey decision vector. The importance level corresponding to the maximum value indicates the importance level of the indicator. Only the indicators with an importance level of “high” are retained for further analysis, completing the indicator selection process.

Analytic Hierarchy Process (AHP)

The Analytic Hierarchy Process (AHP) is a significant method for multi-objective decision-making, proposed by the American operations researcher Thomas L. Saaty in the 1970s. The core idea of the AHP method is to decompose complex decision problems into several hierarchical levels, enabling decision-makers to gradually clarify their decision-making process, accurately evaluate the pros and cons of various decision alternatives, and make the final decision. The basic principle of the AHP method is to determine the relative importance between different factors at different levels through pairwise comparisons, and then construct judgment matrices based on these comparison results. Subsequently, mathematical methods are used to calculate the eigenvalues and eigenvectors of the judgment matrices, and consistency checks are performed to ensure the rationality and credibility of the comparison results. Finally, the factors are ranked based on the weight values of the eigenvectors to determine the optimal decision alternative.

Selection of evaluation indicators for community park rehabilitation landscape

Evaluation indicator screening based on GST method

The Grey Statistical Technique (GST) (Ashour et al. 2020; Wang 2019; Li 2021; Littwin and Stock 2020; Wagner and Praxmarer 2013; Abd El Aziz 2015) is a fuzzy processing method that is suitable for data with relatively limited information and uncertainty. In response to the abstractness, multicriteria nature, difficulty in determination, and difficulty in quantification of existing indicators for rehabilitation landscapes, grey statistical analysis can establish an element set through comparison, quantification, and ordinalization, thereby extracting important indicators. The use of whitening function in data processing can effectively avoid irregularities caused by different expert opinions or the influence of outliers. Through this method, general landscape evaluation indicators suitable for community park construction can be selected. The specific steps for the selection are as follows:

Initial selection of evaluation indicators

Through literature review (Liu and Xinhao 2021; Zhong et al. 2022; Peschardt and Stigsdotter 2013; Ellis 2013; Foster et al. 2011; Jayasooriya et al. 2017; Marcus and Barnes 1999; Whyte 1980) and expert consultation, a preliminary set of 30 indicators for community park rehabilitation landscapes is established, including three criterion layers: physiological rehabilitation, psychological rehabilitation, and spiritual rehabilitation.

Calculation of Grey whitening function

Questionnaire survey

A questionnaire survey was distributed to 20 experts in the field of rehabilitation landscape design through Wen Juan Xing, using the Likert 7-point scale method to score the importance of the initial selected indicator elements: 1 means not important at all, 7 means very important, and 2–6 means the importance is between the two. The raw data on the importance level of the 30 indicators for urban health parks were obtained.

Data processing

The obtained raw data of the initial importance levels of the 30 selected indicators from experts are processed using the Grey-Class Whitening Function. In Grey Statistical Technique, the whitening function is a mathematical method used to handle in-complete and uncertain information. According to the Grey Statistical Technique, the preliminary set of indicators for community park rehabilitation landscapes is divided into three categories: high, medium, and low. Based on the definition method of the Grey Correlation Degree Function, the definitions of high-level, medium-level, and low-level Grey-Class Whitening Functions are provided for \({f}_{k}(ab)\) (Lindal and Hartig 2015; Tyrvaeinen et al. 2014):

When \(k=1\), the whitening function calculation formula for “high degree of importance” is as follows:

$${f}_{1}(ab)=\left\{\begin{array}{ll}1 & {h}_{ab}\,\ge\, 7\\ \frac{{h}_{ab}-4}{7-4} & 4 \,<\, {h}_{ab} \,<\, 7.\\ 0 & {h}_{ab}\le 4\end{array}\right.$$
(1)

When \(k=2\), the whitening function calculation formula for “medium degree of importance” is as follows:

$${f}_{2}(ab)=\left\{\begin{array}{ll}0 & {h}_{ab}\,\ge\, 7\\ \frac{7-{h}_{ab}}{7-4} & 4 \,<\, {h}_{ab} \,<\, 7\\ 1 & {h}_{ab}=4\\ \frac{{h}_{ab}-1}{7-4} & 1 \,<\, {h}_{ab} \,<\, 4\\ 0 & {h}_{ab}\le 1\end{array}.\right.$$
(2)

When \(k=3\), the whitening function calculation formula for “low degree of importance” is as follows:

$${f}_{3}(ab)=\left\{\begin{array}{ll}0 & {h}_{ab}\,\ge\, 4\\ \frac{4-{h}_{ab}}{4-1} & 1 \,<\, {h}_{ab} \,<\, 4\\ 1 & {h}_{ab}\le 1\end{array}\right.$$
(3)

Where, a represents the degree of importance, \(a=1,2,\cdots ,7\), b represents the index number, \(b=1,2,\cdots ,30\), \({f}_{k}(ab)\) represents the whitening function value of the b-th index with an importance level of a, \({h}_{ab}\) represents the assigned value corresponding to the importance level of the b-th index as a. Using the above formulas (1)–(3), the whitening function values for each index can be calculated according to the three levels of importance: high, medium, and low.

Calculate grey decision coefficients and grey decision vectors

Grey decision coefficients are used to measure the degree of influence of various factors on the decision results. The grey decision coefficient \({\eta }_{k}(b)\) for the b-th index in the k-th grey class is defined are shown in Table 1.

Table 1 Results of Grey statistical analysis on the importance levels of community park rehabilitation landscape indicators.
Full size table

\(L(ab)\) represents the number of experts who assign the importance value of a to the b-th index, and \({f}_{k}(ab)\) represents the whitening function value of the b-th index importance being a. Therefore, \({\eta }_{k}(b)\) is defined as follows:

$${\eta }_{k}(b)=\mathop{\sum }\limits_{a=1}^{7}L(ab)\times {f}_{k}(ab).$$
(4)

For each evaluation index, corresponding high, medium, and low three-level grey decision coefficients can be calculated using Eq. (4):

$${\eta }_{1}(b),{\eta }_{2}(b),{\eta }_{3}(b).$$

The grey decision vector is composed of these three coefficient arrays:

$$\{{\eta }_{1}(b),{\eta }_{2}(b),{\eta }_{3}(b)\}.$$

By comparing the grey decision vectors of each evaluation indicator, the evaluation indicators with high importance levels are selected, completing the selection of evaluation indicators.

Evaluation indicator screening result analysis

Analysis of physiological rehabilitation landscape evaluation indicator system

Ecological diversity is of significant importance for community park rehabilitation landscapes as it can promote ecological balance, enhance landscape quality, and contribute to the physical and mental well-being of residents. Therefore, this indicator is selected. Soil and water conservation can protect land and water resources, improve the aesthetic appeal of landscape design, and provide a sense of spiritual pleasure. Hence, this indicator is selected. Climate conditions can influence plants and landscape elements, ensuring the stability and sustainability of vegetation landscapes. Thus, this indicator is selected.

The duration of sunlight can affect people’s activity time, comfort, and overall health in the park, consequently impacting the park’s utilization and effectiveness. Therefore, this indicator is selected. Adequate noise control can reduce the perceived noise level and enhance people’s comfort, thus this indicator is selected. Water resources can enrich the landscape quality of the park, increase its beauty and attractiveness, hence this indicator is selected. Implementing effective pollution control measures can reduce the concentration of harmful substances within the park, improve air quality, and enhance the stability and diversity of the park’s ecosystem. Therefore, this indicator is selected. Physiological Rehabilitation Landscape Evaluation Index System Analysis are shown in Fig. 2.

Fig. 2
figure 2

Physiological rehabilitation landscape evaluation index system analysis chart.

Full size image

Analysis of the evaluation index system for psychological rehabilitation landscapes

Sense of security refers to the feeling of safety and level of trust that individuals have in a specific environment. It can influence whether people are willing to stay, participate, and enjoy the services and functions of the rehabilitation landscape in that environment. Therefore, this indicator is selected. Attractiveness can create a sense of pleasure and relaxation, maintaining good emotional stability in individuals. Hence, this indicator is selected. Interaction between people and the environment can stimulate engagement and experiences, thus it is selected. A smaller boundary scale can enhance the diversity of the rehabilitation landscape, while a larger boundary scale may result in a relatively homogeneous type. Therefore, this indicator is selected.

In the psychological rehabilitation criterion layer, three indicators were not selected. Readability primarily focuses on the legibility of text or information, choice opportunity mainly relates to personal choices and decisions, and indoor-outdoor correlation mainly concerns the relationship between architecture and landscape. Since the evaluation indicators of rehabilitation landscape mainly focus on the physical, ecological, social, and cultural attributes of the landscape, these indicators were not selected. Psychological Rehabilitation Landscape Evaluation Index System Analysis are shown in Fig. 3.

Fig. 3
figure 3

Psychological rehabilitation landscape evaluation index system analysis chart.

Full size image

Analysis of the evaluation index system for psychological rehabilitation landscapes

Hierarchy can help landscape designers create more readable and usable spatial environments, thereby enhancing the effectiveness of rehabilitation landscapes. Therefore, this criterion is selected. Wholeness can create a harmonious landscape space and improve landscape connectivity, so it is selected. Symbolism incorporates cultural connotations into landscape themes through symbolic elements, maximizing the rehabilitation effect, and thus is selected. Pleasure can create a psychologically pleasing landscape, improving mental health, and therefore is selected. Aesthetic value can enhance the participation and image of the landscape, so it is selected. Cultural identity can promote the integration of the landscape with the local community, and therefore this criterion is selected.

In the mental rehabilitation criterion layer, three indicators were not selected. Permanence is a subjective feeling that is difficult to quantify and evaluate, so it is not selected. The core goal of rehabilitation landscapes is to provide an environment beneficial to physical and mental well-being, and sacredness is not a necessary condition for achieving this goal, so it is not selected. The reason for the non-selection of fluidity is that there is currently no auxiliary therapeutic method based on the sense of flow in the landscape. Mental Rehabilitation Landscape Evaluation Index System Analysis are shown in Fig. 4.

Fig. 4
figure 4

Mental rehabilitation landscape evaluation index system analysis chart.

Full size image

Building a community park rehabilitation landscape evaluation index system based on AHP

Analytic Hierarchy Process (AHP) is a multi-criteria decision-making method that combines qualitative analysis and quantitative research. It is used to solve complex problems by breaking them down into different hierarchical dimensions and making scientific judgments through quantitative thinking, providing effective solutions. It was proposed by American scholar Saaty in the 1970s and has been widely applied (Yin et al. 2022; Hansmann et al. 2007; Liao et al. 2023; Wu et al. 2019; Zhou et al. 2023). The hierarchical structure of the community park rehabilitation landscape analysis is shown in Table 2.

Table 2 Hierarchical structure of community park rehabilitation landscape analysis.
Full size table

Construction of judgment matrix

In general, the importance of different elements is different in the minds of each decision-maker. Therefore, in the AHP method, pairwise comparison of decision factors is often used to establish pairwise comparison matrices, and numerical scales are commonly used to quantify decision-makers’ judgments, and then the judgment matrix is obtained. The reference table for the scale meanings is shown in Table 3, and these values are determined based on decision-makers’ intuition and judgment in qualitative analysis.

Table 3 Scale meaning comparison table.
Full size table

Calculation of weight vector and consistency test of judgment matrix

Eigenvalue method to obtain maximum eigenvalue and eigenvector (weight vector)

(1) Calculate the eigenvalue and corresponding eigenvector for each judgment matrix:

$$\lambda =({\lambda }_{1},{\lambda }_{2},\cdots {\lambda }_{n}),{\omega }_{i}=({\omega }_{i1},{\omega }_{i2},\cdots ,{\omega }_{in}),i=1,2,\cdots n.$$

(2) Normalize the eigenvectors:

$${{\omega }_{i}}^{0}=\frac{1}{{\sum }_{j=1}^{n}{\omega }_{ij}}({\omega }_{i1},{\omega }_{i2},\cdots ,{\omega }_{in}),i=1,2,\cdots n.$$

(3) Calculate the maximum eigenvalue:

$${\lambda }_{\max }=\,\max \{{\lambda }_{1},{\lambda }_{2},\cdots {\lambda }_{n}\}.$$

Consistency test of judgment matrix

(1) Calculate the consistency index CI:

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

(2) Determine the corresponding average random consistency index RI from the table.

The values of the average random consistency index RI are shown in Table 4.

Table 4 Average random consistency index RI.
Full size table

Based on the order of the judgment matrix, the corresponding average random consistency index RI can be obtained from Table 4.

(3) Calculate the consistency ratio CR and make a judgment:

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

When \(CR=0\), it is considered that the judgment matrix is completely consistent; when \(CR \,<\, 0.1\), it is considered that the consistency of the judgment matrix is acceptable; when \(CR \,>\, 0.1\), it is considered that the judgment matrix does not meet the consistency requirement, and it needs to be reconstructed and revised until it meets the consistency requirement.

Hierarchical indicator single sorting

Using the calculation method described above, the maximum eigenvalue, eigenvector, consistency index CI, and consistency ratio CR of each judgment matrix were calculated using Matlab software. The data for each indicator are detailed in Table 5.

Table 5 Hierarchical indicator single sorting table.
Full size table

Overall sorting of hierarchy

The weight vectors of the above two layers of indicators are only sorted within a single layer. To obtain the overall sorting of the indicator layer P in the system with respect to the target layer A, it is necessary to calculate the combined weight vector of the indicator layer P with respect to the target layer A. The formula for calculating the combined weight vector is as follows:

$${\omega }_{AP}={\omega }_{A{D}_{i}}\times {\omega }_{D{P}_{i}}.$$
(5)

By plugging in the data from Table 5 into formula (5), the combined weight vector of indicator layer P with respect to target layer A is calculated as follows:

$$\begin{array}{c}{\omega }_{AP}=(0.0419,0.2208,0.1675,0.3704,0.0714,0.0970,0.0310\\ \,0.5586,0.1226,0.0629,0.2559\\ \,0.1596,0.2504,0.1006,0.3825,0.0641,0.0428).\end{array}$$

Therefore, the weights of each factor in the comprehensive evaluation indicator system for rehabilitation landscape design effectiveness are shown in Table 6.

Table 6 Weights of each factor in the comprehensive evaluation indicator system for rehabilitation landscape design effectiveness.
Full size table

Evaluation indicator weight calculation results analysis

Through the Analytic Hierarchy Process (AHP), a comprehensive evaluation and analysis of the design effectiveness of community park rehabilitation landscapes were conducted. From the overall ranking of the weights of 17 indicators, it was concluded that the duration of sunlight (P14) had the highest weight, followed by the sense of safety (P21). These two indicators have a significant impact on the comprehensive evaluation of the design effectiveness of community park rehabilitation landscapes. The data also showed that soil and water conservation (P12) and climatic conditions (P13) were equally important in evaluating the effectiveness of rehabilitation landscape design.

With the continuous development of society and the increasing aging population, the demand for community park rehabilitation facilities has become more urgent. Therefore, it is necessary to strengthen the construction and design of rehabilitation landscapes in order to make them more human-centered and provide convenience for a larger population. The weights of other indicators, in descending order, are as follows: boundary scale (P24), pollution mitigation (P16), noise control (P15), pleasantness (P34), attractiveness (P22), ecological diversity (P11), integrity (P32), accessibility (P17), hierarchical sense (P31), human-landscape interaction (P23), symbolism (P33), aesthetic value (P35), and cultural identity (P36).

Sensitivity analysis

This study uses the Grey Statistical Method (GST) and Analytic Hierarchy Process (AHP) to assess community park rehabilitation landscapes. The aim is to clarify how various indicators affect rehabilitation outcomes. To ensure the robustness and reliability of the evaluation system, we performed a sensitivity analysis to examine how changes in indicator weights and screening criteria impact the final results.

Sensitivity analysis of indicator weights

First, we assessed how adjusting the weights of various indicators affects the comprehensive evaluation results. The analysis indicates that daylight duration (P14) and sense of safety (P21) have a significant impact, highlighting their importance in rehabilitation landscapes. Variations in these indicators’ weights lead to noticeable changes in the overall results. The sensitivity analysis also shows that soil and water conservation (P12) and climatic conditions (P13) play a crucial role in landscape design. In contrast, changes in the weights of noise control (P15), pleasure (P34), and ecological diversity (P11) have a smaller effect on the overall evaluation, suggesting a limited impact on system stability.

Sensitivity analysis of indicator screening

In the screening process, we selected 13 core indicators using the GST method. The sensitivity analysis results show that indicators excluded from the final evaluation have minimal impact on the results, confirming that the 13 selected indicators effectively reflect the characteristics of community park rehabilitation landscapes. However, indicators such as human-environment interaction (P23) and cultural identity (P36), although not part of the core system, may still influence rehabilitation outcomes in specific contexts.

Stability and reliability of results

Overall, the sensitivity analysis shows that the evaluation system is stable. Core indicators like daylight duration and sense of safety consistently influence the comprehensive evaluation results across different weight settings, whereas changes to other indicators have a lesser effect on the evaluation system.

Strategies for optimizing the quality of community park rehabilitation landscapes

Establishing integrated functional spaces with clear audience focus

The design of community park rehabilitation landscapes needs to consider the diverse landscape requirements of different age groups, creating integrated and interconnected spaces. For example, when catering to youth and children, it is necessary to consider the diverse and safe landscape colors for children while also catering to the trendiness and novelty preferences of young adults (Semeraro et al. 2021). When addressing the needs of elderly and middle-aged individuals, the design should focus on the therapeutic functions and privacy considerations for the elderly, as well as social needs and stress relief functions for middle-aged individuals (Geng et al. 2022; Du et al. 2017). Creating composite spatial landscapes that cater to different activity demands can promote the healing process for the audience. Combining dynamic and static activity spaces can meet people’s daily activities and socialization needs while fulfilling their psychological desire for a sense of belonging (Kemperman and Timmermans, 2014; Hunter et al. 2015).

Creating a comprehensive sensory experience in a natural environment

In the design of community park rehabilitation landscapes, integrating the five senses (vision, hearing, taste, smell, and touch) with rehabilitation design and promoting interaction between individuals and various elements is essential for achieving effective rehabilitation outcomes (Shuvo et al. 2020; Shen et al. 2021). By satisfying basic aesthetics and functionality, utilizing both soft and hard landscape elements to stimulate the five senses, and facilitating interaction between people and the environment, the design can contribute to the holistic well-being of individuals. The hierarchical arrangement of green vegetation and the richness of water features can create natural and visually appealing landscape spaces (Zhai et al. 2020). The coexistence of green spaces and blue spaces can better promote the healing process for users.

Emphasizing human-centric design details for rehabilitation landscapes

Human-centric design details can create a comfortable environment and provide a pleasant psychological experience for people. They can also increase the utilization rate of community parks and enhance the potential for rehabilitation. For instance, using appropriate colors and contrasts, providing accessible pathways, ensuring landscape facilities meet safety standards, creating moderate contrasts with landscape elements, and incorporating multifunctional features can enhance the attractiveness and comfort of rehabilitation landscapes. This approach caters to the individual needs of rehabilitation participants, providing them with a comfortable, safe, and positive environment (Yu et al. 2021; Ernstson, 2013; Labib et al. 2020).

Conclusion and discussion

The evaluation of community park rehabilitation landscape quality involves a comprehensive analysis process that considers multiple attributes and objectives. During the evaluation, it is necessary to consider the influence of both subjective and objective factors in landscape perception (Store et al. 2015; Ye and Qiu 2021). The GST-AHP combination model enables the evaluation of non-quantitative and quantitative factors. This model allows for the assessment of individual community park rehabilitation landscape quality as well as the comparison of multiple community park rehabilitation landscapes (Van den Bosch et al. 2016; Opdam 2020). By evaluating the results and weights of various indicators, targeted improvements can be made, providing important references for this type of landscape design and truly enhancing the level of landscape design.

Through the application of the GST-AHP combination method, this study explored the construction of a community park rehabilitation landscape quality evaluation model and the selection of indicator factors: (1) Establishing a scientific and rational evaluation system for community park rehabilitation landscapes is the foundation for maintaining the vitality and development of community parks (Wei et al. 2022). (2) By adopting the GST-AHP combination model, a quality evaluation system for community park rehabilitation landscapes was constructed to analyze the factors influencing the rehabilitation landscape in community parks. This further enhances the rehabilitative functionality of community parks and provides a theoretical reference for rehabilitation landscape design. (3) Further research can be conducted through survey questionnaires to explore the scientificity and accuracy of indicator factors at different criteria levels, making the evaluation system for rehabilitation landscapes more standardized, scientific, rational, and accurate.