Decoding Waterfront Vitality: A Space–Experience Interaction Evaluation

Abstract

Waterfront recreational spaces, as key urban ecological resources, are distinctive in their scarcity and ecological fragility. Their sustainable revitalization requires evidence-based spatial planning and design. The analysis of the vitality of waterfront recreational spaces, which are characterized by the interaction between space and experience, essentially explores how human, water, and the city can coexist and thrive together. Based on the dual characteristics of vitality, this study presents a space–experience interactive evaluation system for waterfront recreational places that incorporates multi-source data. The vitality evaluation results can then be cross-validated with intuitive representations of vitality quantified using pedestrian flow data. Furthermore, this can be used to accurately calibrate the vitality gradient, identify and analyze the anomalous units, and provide insight into influencing factors and underlying mechanisms of vitality. The empirical investigation of the waterfront recreational area of Suzhou Jinji Lake Scenic Area (JLSA) demonstrates that this method can accurately identify spatial vitality distributions and effectively characterize the key elements of vitality zones at different levels. It can precisely decode the vitality of waterfront recreational spaces, providing fresh perspectives on understanding the space–experience interaction in waterfront recreational spaces and directing actions for enhancing vitality. In addition to serving as a supplement to existing research, it provides a flexible, scalable evaluation framework for a variety of waterfront contexts, supports the implementation of human-centered urban design, and offers theoretical and practical support for the sustainable development of waterfront areas.

1. Introduction

An urban waterfront recreational space is a vital part of contemporary urban systems and is a recreational area that is connected to the city’s water region [1]. Modern urban and landscape studies generally agree on the composition of urban waterfront recreational spaces, which are multipurpose open spaces made up of three major components: water bodies, embankment systems, and adjacent land area [2]. Being one of the most significant locations for interaction between human activity and the natural environment, recreational areas serve as a window into the city’s daily public life. However, urban waterfront recreational areas represent a unique, scarce, and fragile geographic resource. The subtle features of the leisure experience, as well as the physical surroundings, influence vitality generation.

The concept of cities as living complex systems serves as the foundation for spatial vitality research [3]. This paradigm requires holistic thinking and dynamic viewpoints in order to thoroughly examine the nonlinear linkages, systemic resilience, and co-evolutionary mechanisms among urban elements. After World War II, as critical thoughts on modernist lifestyles, automobile-dominated transportation, and urban expansion emerged, the discipline of urban planning experienced a rise in vitality studies [4]. Scholars from various perspectives have systematically elucidated the formation mechanisms of urban vitality: it is fundamentally a dynamic interaction between space and behavior, serving as both a carrier of culture inheritance and an organic integration of human activity patterns and environmental elements [5,6]. Throughout the postwar wave of urban studies, scholars thoroughly explored the generative processes of urban vitality from a range of theoretical lenses, revealing its underlying nature as a dynamic interaction between human activity and spatial functions [7]. In addition to being a composite of spatial functions and ecological carrying capacity [8], the essence of urban vitality has been revealed such as systems dynamics [9], functional responsiveness, and place elasticity. It is also a dynamic equilibrium between top-down regulation and bottom-up feedback, and it serves as a social container for diverse activities, embodying the fundamental characteristics of cities as living organisms. While Lynch K used vitality and other factors to discuss urban spatial design quality and trace urban form value standards throughout history [10], Bentley I. defined vitality as the ability to support diverse social activities and enable diverse user experiences, resulting in resonant and dynamic environments [11].

Academics began conducting systematic research on the connection between urban life and recreational places in the middle of the 20th century, motivated by postwar reconstruction and population expansion. During this time, two paradigm shifts occurred; on the one hand, physical planning gave way to humanism, as empirical research by William H. Whyte showed a strong relationship between facility layout and pedestrian accessibility and public space vitality [12]; on the other hand, ecological dimensions emerged, with Ian McHarg being the first to incorporate recreational spaces into urban ecological network analysis [13]. The methodological groundwork for the subsequent emergence of resilient city ideas and New Urbanism was established by these theoretical advances. Meanwhile, urban recreational space theory has thrived, spawning a slew of important theoretical study such as recreational geography and tourism destination life cycle theory, infusing new energy into the field of study [14,15]. The ecological and human significance of waterfront recreation areas has received unprecedented attention, and they have emerged as an important link in developing metropolitan human settlements [16]. Recent study has provided impressive results in the fields of geographic forms [17], spatial pattern evolution [18,19], spatial planning design and management [20], and geographical assessment [21]. The research scope has also grown significantly, with studies concentrating on influencing factors, constituent dimensions, and spatial perception [22].

The increasing significance of human-centered planning concepts has made sustaining urban vitality a pillar of modern urban planning. This paradigm shift has been accelerated by the new quantitative methods, such as big data analytics, which transform urban studies by delivering reliable scientific frameworks for researching and enhancing urban vitality [23,24,25]. An increasing number of researchers are embracing digital platforms to conduct data-driven studies on urban vitality that include quantitative analysis [26]. The structural equation model [27], picture semantic recognition [28], online text analysis [29], and other approaches have all been incorporated into the appropriate quantitative analysis, providing a theoretical framework and scientific instruments for renewing waterfront recreation area [30]. Multi-source data, such as mobile signaling data, social media data, traffic flow data, and POI (Point of Interest) data, greatly reduce traditional study limits and provide crucial support for assessing waterfront spatial vibrancy [31,32].

Contemporary urban waterfront revitalization projects adopt user experience as the core paradigm, reconstructing vibrant scenarios via the interconnected human-water-city symbiosis system [33]. By scientifically addressing fundamental issues like ecological sensitivity, functional complexity, and spatial dynamism in waterfront areas, the spatial vitality theory provides systematic methodological support for sustainable development models, demonstrating its exceptional practical value in this process [34,35]. Spatial vitality research can effectively address ecological vulnerability, diversity of functional requirements, dynamic spatial morphology, and other waterfront space characteristics in urban renewal projects [36,37]. It also provides scientific support for balanced development of spatial vitality [38]. By examining human–environment interactions using behavioral big data and other cutting-edge technologies, these user-centered waterfront recreational spaces have successfully produced emotionally resonant vitality scenarios [39].

While previous studies have investigated various aspects of spatial cognition, a comprehensive study loop involving spatial-perception and experience-feedback is still lacking, and they are limited by a lack of integrated frameworks. In addition, as human-centered design principles gain deeper traction, it becomes increasingly important to rigorously evaluate waterfront spaces through the lens of organic space–experience interaction. This approach may promote human-centered design in waterfront areas by identifying significant components of waterfront vitality and revealing dynamic feedback loops between spatial characteristics and subjective experience, therefore illuminating the underlying mechanics of vitality. Furthermore, to promote sustainable waterfront revitalization, the framework is applied to a practical case to verify its explanatory power and generate scalable optimization approaches.

2. Study Area

Jiangnan is a broad geographical and cultural region in China, which includes southern Jiangsu, northern Zhejiang, Shanghai, and southeastern Anhui. The broad definition of Jiangnan covers several provinces south of the Yangtze River, while the narrower definition focuses on the southern part of the Yangtze River Delta and the Taihu Lake Basin, which represents the core economic and cultural center. With more than 160 million permanent residents, this region is among China’s most densely populated and economically vibrant. Suzhou, a typical Jiangnan metropolis, with its water network spanning 25,000 km, centered on the Yangtze River, Taihu Lake, and Grand Canal. Its waterfront system is made up of “dual rings and three belts”: an outer 16 km cultural loop surrounding the old city moat, an inner loop that connects historic areas, and three belts that include the Taihu ecological, canal cultural, and Suzhou River dynamic zones. Jinji Lake is the core of Suzhou’s water network, a gem that connects ecological and culture, creating a scenic collision of history and modernity. As China’s largest waterfront leisure destination, Jinji Lake’s spatial structure and visitor dynamics make it a typical case study for investigating waterfront vitality.

Jinji Lake Scenic Area (JLSA) is China’s largest urban lake park, with a unique blend of urban views, natural landscapes, and cultural attractions. It is not only a popular place for locals, but it is also a national 5A-level tourist destination that smoothly integrates urban living and tourism. Jinji Lake, previously a lagoon produced by a stream from Taihu Lake, is now the main water source for Suzhou Industrial Park. A Tang Dynasty geographical and historical treatise on the Wu region (centered around modern Suzhou, Jiangsu), the lake was originally known as Jinjing Lake, but it was changed to Jinji Lake during the Ming dynasty due to the legend of the golden rooster declaring the lake’s daybreak [40]. With a total water area of 7.4 km² and a total shoreline area of roughly 4.1 km², JLSA has matured into a model of urban–nature harmony over three decades of diligent development. This “City–Lake Symbiosis” plan expertly combines ancient Chinese garden aesthetics with modern cityscapes, resulting in a popular destination that drew 25 million tourists in 2025 alone, including 1.36 million during the 2025 May Day holiday. Tourists and locals make up the majority of JLSA’s visiting population. More than half of the tourists are from the Jiangnan region, and a sizable number of them are young individuals in the 25–40 age range. On the other hand, the majority of tourists who live in the area are middle-aged and older, with most being 40 years of age or older.

Based on environmental behavior theory and human-scale design principles, this study specifically defines the research scope as outdoor public areas within a 500 m buffer zone along the shoreline of JLSA (about a 5 min walk), according to effective perception distance towards water bodies in environmental behavior (500–800 m). The sample unit division was then carried out with an emphasis on water and in relation to current physical divisions (e.g., highways, bridges, and water body boundaries). The study area was divided into 70 units based on the requirement that each unit have a specific level of water visibility. It should be mentioned that, in accordance with the concept of the waterfront recreation area, its public nature and open qualities, non-open sections of residential land and current construction zones must be removed throughout this procedure [41] (Figure 1, Figure 2 and Figure 3).

3. Methods

As illustrated in Figure 4, a Vitality Evaluation framework was built based on space–experience interactions. It integrates many sources of heterogeneous data, such as geographical measures and perceptual experience, and establishes distinct quantification paths based on the attribute qualities of affecting factors. At the end of the evaluation process, it generates a regionally differentiated comprehensive vitality index V applying a decomposable vitality information hierarchy.

This study combines visitor activity data into the vitality assessment. However, to limit the influence of anomalous oscillations caused by single sources, the Vi evaluation system has been constructed on pedestrian flow heat data, providing a scientific foundation to further vitality level revision and interpretation. It uses pedestrian movement data from the scenic area to develop intuitive representations of vitality (Vi) through spatial superposition analysis. This comparison method not only ensures the framework’s integrity, but also allows for a comparative analysis with V, resulting in scientific insights for vitality-oriented waterfront architecture.

3.1. Participants: Experts Panel and Visitor Group

Before beginning the study, each participant provided their informed agreement to be included in the Questionnaire on the Vitality Evaluation of Waterfront Recreational Spaces. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Suzhou University of Science and Technology (protocol code 202504223, on 22 April 2025).

3.1.1. Experts Panel

Thirty experts from related domains were invited to participate in the factor screening and indicator-weight determination process. In order to ensure the diversity and professionalism of evaluation perspectives, the expert panel is composed of four kinds of professionals: landscape designers, urban planners, scenic area managers, and tourism management researchers. A targeted invitation and field-balancing strategy was used to ensure representativeness in terms of experience and knowledge background: seven to eight experts with intermediate or higher titles or more than five years of practical experience were chosen from each of the four fields to form a comprehensive evaluation group of thirty members.

3.1.2. Visitor Group

During the entire survey process, the random interviews involved 498 people, all randomly stratified on-site. According to statistical data of 2025, the tourist gender ratio in JLSA was 6:4 (male to female). The pilot study indicates that over 50% of JLSA users are between the ages of 19 and 35. Those between the ages of 36 and 59 come in second, and those between the ages of 5 and 11 come in third. Teenagers between the ages of 12 and 18 and those 60 and older have the lowest numbers. Out-of-town visitors made up 23.5% of the total. A total of 39.7% are from far-off parts of Suzhou city, whereas 36.8% are local residents.

The study established a visitor group of 40 volunteers who participated in discussions and subjective assessments of vitality (e.g., SEB assessment). While recruiting for the Visitor Group, the gender ratio of the volunteer group is adjusted to 6:4, and the ratio of locals to tourists is set at 1:3 to ensure that survey data are representative and accurately reflect visits while accounting for the user-composition ratio. A stratified sampling technique was adopted, with visitors stratified by age (young, middle-aged, and old).

3.2. Vitality Evaluation and Intuitive Representations of Vitality

The vitality of waterfront recreational spaces is embodied in the dynamic experiential interaction between people and space: individuals give the space vitality through social interactions (including waterfront activities) and behavioral experiences (e.g., perception of water proximity), while space provides venues for activities and guides individuals’ spatial experiences through its functional attributes (including waterfront facilities) and physical attributes (e.g., squares, roads). The two have an impact on one another: individual spatial experiences redefine the functions of the space, while space design influences human behavior. Together, they create a dynamic urban waterfront environment [42,43]. As a result, the interaction between individual experience and waterfront composition must be considered when constructing the evaluation framework of Vitality of Waterfront Recreational Space (V).

3.2.1. Vitality Evaluation Framework

The dual characteristics of spatial quality and spatial experience are linked to create the initial draft of the evaluation framework. Following this basic framework, a thorough screening and integration of relevant factors was conducted. The assessment elements were screened using word frequency and analyzed before being added into the final evaluation system. The hierarchical factor system ensures that assessments are organized while remaining compatible with the original framework’s goals. Finally, weight coefficients for indicators at each level are determined using the Delphi technique and the Analytic Hierarchy Process method. Differentiated quantitative paths are intended to handle a variety of evaluation metrics.

This approach involves collecting multidimensional data, standardizing filtering, and using hierarchical processing to create a consistent evaluation factor system. Accessibility, Spatial Coziness, Water Proximity, and Visual Attraction are the four components of spatial quality (SQ), which examine the spatial features of waterfront environments. The four factor comprise fourteen factors, including waterfront-specific environmental perception settings. Path Holistic Experience and Node User-Friendliness are two components of Spatial Experience (SE), each with two indicators focusing on embodied experience and visit quality, respectively. The names of the indicators were revised according to factor analysis, taking into account the morphological peculiarities of the waterfront area [44].

(1)

Factors and Indicators

In order to guarantee objectivity and comprehensiveness, the screening of indicators follows the principle of combining objectivity and operability, taking into account both universality and specific demands [45]. In this procedure, a high-frequency vocabulary screening method based on lexical frequency statistics was used to apply descending-order preliminary screening to the factor pool. To solve multicollinearity, orthogonal dimensionality reduction was subsequently accomplished using Principal Component Analysis (PCA) and Exploratory Factor Analysis (EFA). The K-means Clustering Technique was then used to find similarity features among components and allocate them to different categories.

To start, comprehensive searches were conducted in CNKI (China National Knowledge Infrastructure) and WOS (Web of Science) high-quality indexed databases using keywords “Vitality,” “Recreation Space,” and “Waterfront Recreation Space.” The core databases of CNKI (China National Knowledge Infrastructure) and Web of Science, were chosen to provide complete coverage of Chinese academic publications while also providing access to high-quality international resources. CNKI, China’s largest and most comprehensive academic platform, offers substantial coverage for research on Chinese contexts and localized phenomena. Scopus provides broad discipline coverage, while Web of Science (WOS) has more stringent inclusion requirements. Therefore, in alignment with the research focus on “high-quality indexed literature,” Web of Science was prioritized to ensure greater academic rigor and representativeness of international publications.

The study used not only exact match retrieval but also composite search phrases that included synonyms, near-synonyms, and related terms with Boolean logic operators (such as AND, OR, and NOT). A preliminary selection of 3368 relevant articles was produced by the search (from 2016 to 2025). A total of 3082 articles were kept for factor screening following thorough examination and quality assessment. A total of 458 representative factors were identified by statistical analysis of frequent factors. Then, 66 effective components were filtered from 458 previously chosen elements using factor analysis. Word frequency ranking in descending order generated 30 high-frequency components after low-frequency items were eliminated. After initial factors were extracted using the principal component analysis approach and the factor structure was clarified using Varimax orthogonal rotation, six common components were then retrieved as factors (2nd Level). The cumulative variance contribution rate of 82% demonstrated that the six factors could sufficiently explain the data from the original 30 variables. Ultimately, the space–experience indication was employed to match and identify factors at all levels. Following rotation, common characteristics from the factor loading matrix were summarized and used to rename the factors. A three-level vitality evaluation system was established, as indicated in

Table 1. Vitality evaluation framework and weights.

Category		Factor		Indicator
Item	Weight	Item	Weight	Item	Weight
Spatial Quality (SQ)	0.543	Accessibility F1	0.204	Attainability F1-1	0.587
				Traffic Convenience F1-2	0.413
		Spatial Coziness F2	0.363	Water Clarity Index F2-1	0.201
				Water Exposure Index F2-2	0.133
				Waterfront Cleanliness Index F2-3	0.183
				Sky Openness Index F2-4	0.088
				Green View Index F2-5	0.122
				Amenity Density Index F2-6	0.113
				Recreational Facility Density F2-7	0.161
		Water Proximity F3	0.210	Waterfront Facility Coverage Index F3-1	0.529
				Waterfront Safety Index F3-2	0.471
		Visual Attraction F4	0.223	Waterfront Aesthetic Index F4-1	0.429
				Recreation Space Aesthetic Index F4-2	0.364
				Recreation Space Color Comfort Index F4-3	0.207
Spatial Experience (SE)	0.457	Path Holistic Experience F5	0.459	Continuity F5-1	0.515
				Trail Richness F5-2	0.485
		Node User-Friendliness F6	0.541	Activity Support F6-1	0.530
				Facility Friendliness F6-2	0.470

.

(2)

Weights

Thirty experts from the Experts Panel established evaluation criterion weights through a three-round Delphi method. Experts gave feedback and scored criteria (1–5 points) in each round. For statistical analysis, including mean score computation, consistency checks (Kendall’s W = 0.82, p < 0.01), and handling outliers, SPSS (IBM SPSS Statistics 26.0.0) was utilized. Reliable consensus was ensured by using normalized mean scores to determine final weights.

3.2.2. Intuitive Representations of Vitality

Since indicators like pedestrian movement and aggregation density are frequently used for assessing spatial vitality, the results of its evaluation can be cross-validated using Intuitive Representations of Vitality (Vi) [46]. The average pedestrian flow heat value for each spatial unit is used to statistically estimate the Intuitive Representations of Vitality (Vi). The average pedestrian flow heat value for every unit is calculated during several time periods (Hw for working days, Hr for rest days), offering a consistent metric for contrasting patterns of spatial vitality. This data-driven method makes it possible to evaluate human activity and the effectiveness of urban-space utilization objectively.

$$Hw={\displaystyle \frac{{\displaystyle {\sum }_{\mathrm{i}=6}^{23}H{w}_i}}{18}}$$

(1)

$$Hr={\displaystyle \frac{{\displaystyle {\sum }_{\mathrm{i}=6}^{23}H{r}_i}}{18}}$$

(2)

In Formulas (1) and (2), Hw_i is the heat value in the sample unit for the i-th hour of work days, Hr_i the heat value of the unit for the i-th hour of rest days (i = 6, 7, …, 23). After the data from each unit has been imported into the ArcGIS platform and divided into three levels, the spatial connection and visualization processing can be finished. This approach is based on time-segmented cumulative data of pedestrian flow heat values, and dynamic heat-map is used to show how vitality varies over time. The static comprehensive vitality map visually represents the total intensity of pedestrian flow heat in each unit from a spatial perspective, confirming the accuracy of the spatial gradient distribution of the aforementioned evaluation (V) and providing dual data support for spatiotemporal vitality research.

3.3. Comparative Analysis

This comparative analysis employs a cross-validation technique to ensure the scientific rigor of the previously presented evaluation system and promote further interpretation. Vitality Evaluation (V) and Intuitive Representations of Vitality (Vi) gradient maps are spatially contrasted, and the consistency and variations in spatial distribution between the two datasets are discussed. It should be pointed out that gradient segmentation uses the natural discontinuity approach (Jenks Optimization) to cluster standardized thermal values and then combines significant differences (such as data differences between adjacent units that surpass a threshold) to establish breakpoints.

Vitality levels can be thoroughly calibrated across units throughout this process by comparing the gradients of Vitality Evaluation (V) and Intuitive Representations of Vitality (Vi). A correlation between two sets of outcomes can be constructed with data standards and visualization tools. It compares the quantitative differences in vitality gradient classifications using a combination of statistical analysis and spatial overlay mapping. The 70 units were categorized as High-Vitality Units, Medium-Vitality Units, and Low-Vitality Units. Additionally, as this is an Enhanced Vitality Grading procedure, anomalous units can be recognized, discussed, and re-matched to their vitality gradients.

Following calibration, the relevant data groups of units within each vitality gradient should be examined and calculated, along with metrics like variance (S²), range (R), standard deviation (σ), and Interquartile Range (IQR). Finally, using field surveys and visitor feedback, an in-depth examination of the vitality unit groupings throughout the three gradients was conducted, revealing their vitality traits and underlying causes.

$$S^2={\displaystyle \frac{{{\displaystyle {\sum }_{\mathrm{i}=1}^n(x_i-\overline{x})}}^2}{n-1}}$$

(3)

$$R=\max (x_i)-\min (x_i)$$

(4)

$$\sigma =\sqrt{{\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n(x_i-\mu )}}$$

(5)

$$\mu ={\displaystyle \frac{1}{n}}{\displaystyle {\sum }_{\mathrm{i}=1}^n{x}_i}$$

(6)

$$IQR=Q_3-Q_1$$

(7)

In Formula (3) through (7), n represents the total number of data, and x_i represents the i-th data value.

Q₁ (first quartile): the value located at the 25% position.

Q₃ (third quartile): the value located at the 75% position.

3.4. Data and Quantification

The data-gathering procedure is carefully prepared using composite sampling techniques to guarantee the scientific validity and reliability of study results. It guarantees sample representativeness and enhances the data’s capacity to reflect intricate real-world scenarios by attaining thorough coverage in the time dimension, dynamic balance in the spatial dimension, and astute processing of outlier data, laying the foundation for the validity of research findings.

• Periodic stratified sampling: To improve the temporal representativeness of the data, prevent seasonal or unintentional biases, and guarantee the completeness and continuity of data coverage over the whole year’s time series, set a defined sample interval on a daily or weekly basis.
• Time-segmented balanced distribution: For sampling, split each day into three time segments: morning, afternoon, and evening. This improves the data’s spatial balance, removes the influence of diurnal variations, and captures the dynamic changes over several time periods.
• Multi-batch dynamic adjustment: Multiple rounds of sampling are performed within the framework of fixed cycles and time periods. Cross-validation between batches improves data stability, and sampling tactics are constantly modified based on real collection conditions, increasing data adaptability.
• Abnormal data removal and retesting: To address potential data interference caused by urban construction in some units, manual detection and removal of abnormal data in the unit and nearby site units is performed. After work is completed, retesting is performed to guarantee the integrity of the restored data and that the study conclusions are unaffected by local disturbances.
• A multi-source data integration approach is used during the evaluation process to guarantee the effectiveness and precision of impact factor quantification: instrumental measurements obtain high-precision environmental data, field surveys accurately capture micro-level characteristics, online multi-source data (with repeated sampling to supplement dynamic information) is integrated, and urban management data (including administrative records) is systematically integrated (Table 2). Each sample unit is spatially associated with quantitative data that has been imported into the ArcGIS 10.8 program. The Natural Breaks technique (Jenks Optimization) was used to classify data categories based on intra-class similarity and inter-class difference, leading to more accurate spatial clustering patterns.

  Table 2. The data collection path of the indicators.

  Indicator	Quantitative Pathway	Data Collection and Integration (Field Survey/Software/App)
  F1-1	Depthmap: integration and choice.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F1-2	Kernel density analysis of public transport stations.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-1	Point measurement by Secchi disk experiment method along the shoreline.	Point measurement on-site
  F2-2	Water Exposure Index = (Visible Water Area/Total Site Area) × 100%	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-3	Scoring is conducted in accordance with the Standards for Urban Water Area Cleaning Operations and Quality [47] and the Suzhou Urban Environmental Sanitation Quality Standards [48].	Data from Jinji Lake Scenic Area Management Center
  F2-4	Image semantic analysis Sky Openness Index = (Visible Sky Area/Total View Area) × 100%	GPU-CUDA-enabled Semantic Segmentation App. v1.0, NVIDIA, Santa Clara, CA, USA
  F2-5	Image semantic analysis	GPU-CUDA-enabled Semantic Segmentation App. v1.0
  F2-6	ArcGIS kernel density analysis	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-7	It is measured by the density of seating facilities: ΡC9 = total number of facilities/unit area	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F3-1	Evaluate according to the types and quantities of hydrophilic facilities.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F3-2	Evaluate based on the quantity and quality of safety facilities.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F4-1	The Semantic Differential Evaluation (SBE) method is used to calculate the visual aesthetic perception degree of urban skylines.	Data from Jinji Lake Scenic Area Management Center IBM SPSS Statistics 26.0.0, IBM Corporation, Armonk, NY, USA
  F4-2	The Semantic Differential Evaluation (SBE) method is used to calculate the aesthetic perception degree of nodes on recreational paths.	IBM SPSS Statistics 26.0.0, IBM Corporation, Armonk, NY, USA
  F4-3	Color Impact calculates the color comfort index: VCD = ∑ (2.787 + 0.081 Vni + 0.037 Cni − 0.00075 Hni) × Rni/3	Datacolor App, Datacolor Tools SV, Lawrenceville, NJ, USA
  F5-1	It is measured by the path connection index: γF5-1 = Σmi/N	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F5-2	It is measured by the density of path nodes: ρF5-2 = nodes/total path length × 100	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F6-1	Statistics on revisit intention, residence frequency	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F6-2	Calculation of facility diversity, Kernel density analysis in ArcGIS	Data from Jinji Lake Scenic Area Management Center ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA

  Table 2. The data collection path of the indicators.

  Indicator	Quantitative Pathway	Data Collection and Integration (Field Survey/Software/App)
  F1-1	Depthmap: integration and choice.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F1-2	Kernel density analysis of public transport stations.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-1	Point measurement by Secchi disk experiment method along the shoreline.	Point measurement on-site
  F2-2	Water Exposure Index = (Visible Water Area/Total Site Area) × 100%	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-3	Scoring is conducted in accordance with the Standards for Urban Water Area Cleaning Operations and Quality [47] and the Suzhou Urban Environmental Sanitation Quality Standards [48].	Data from Jinji Lake Scenic Area Management Center
  F2-4	Image semantic analysis Sky Openness Index = (Visible Sky Area/Total View Area) × 100%	GPU-CUDA-enabled Semantic Segmentation App. v1.0, NVIDIA, Santa Clara, CA, USA
  F2-5	Image semantic analysis	GPU-CUDA-enabled Semantic Segmentation App. v1.0
  F2-6	ArcGIS kernel density analysis	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F2-7	It is measured by the density of seating facilities: ΡC9 = total number of facilities/unit area	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F3-1	Evaluate according to the types and quantities of hydrophilic facilities.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F3-2	Evaluate based on the quantity and quality of safety facilities.	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F4-1	The Semantic Differential Evaluation (SBE) method is used to calculate the visual aesthetic perception degree of urban skylines.	Data from Jinji Lake Scenic Area Management Center IBM SPSS Statistics 26.0.0, IBM Corporation, Armonk, NY, USA
  F4-2	The Semantic Differential Evaluation (SBE) method is used to calculate the aesthetic perception degree of nodes on recreational paths.	IBM SPSS Statistics 26.0.0, IBM Corporation, Armonk, NY, USA
  F4-3	Color Impact calculates the color comfort index: VCD = ∑ (2.787 + 0.081 Vni + 0.037 Cni − 0.00075 Hni) × Rni/3	Datacolor App, Datacolor Tools SV, Lawrenceville, NJ, USA
  F5-1	It is measured by the path connection index: γF5-1 = Σmi/N	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F5-2	It is measured by the density of path nodes: ρF5-2 = nodes/total path length × 100	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F6-1	Statistics on revisit intention, residence frequency	ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA
  F6-2	Calculation of facility diversity, Kernel density analysis in ArcGIS	Data from Jinji Lake Scenic Area Management Center ArcGIS Pro 3.4.3, Esri, Redlands, CA, USA

The data-gathering period lasted from March 2024 until September 2025. Given the impact of updated construction projects in the study area (Jinji Lake Waterfront Greenway, Jinji Inn Multi-Service Station, Suzhou Museum of Contemporary Art, etc.) on data collection, the study conducted a three-month data supplement collection period (September 2025–December 2025).

4. Results

4.1. Vitality Evaluation

4.1.1. Overall Vitality

The overall outcome of the vitality evaluation shows considerable geographical heterogeneity, demonstrated in Figure 5. Vitality scores ranged from 2.005 to 4.329 (on a 5-point scale) according to a statistical analysis of 70 evaluation units. The graph shows a gradient distribution pattern with higher values along the northwest and west shorelines than the northeast shoreline. High-V units (V ≥ 4.00) make up 8.6% (6 units) of the total and are scattered in a dotted pattern around the commercial core (Gate of the Orient), interspersed with medium-V units. Units with high scores (3.50 ≤ V < 4.00) make up 18.6% (13 units) and are evenly spread over the eastern cultural leisure belt, with medium-V units scattered locally. In contrast, the northern ecological wetland zone and southeast traffic interface score largely below 3.00, indicating a contiguous low-V distribution pattern.

As demonstrated in Figure 6, the total score for Spatial Quality (SQ) is higher than that for Spatial Experience (SE). Only 27.1% (19 units) had a higher SE score than SQ score. In the second-level evaluation factors, the score gradient of these influencing factors is Water Proximity (F3) > Visual Attraction (F4) >> Spatial Coziness (F2) > Accessibility (F1), with Accessibility (F1) receiving the lowest overall score; there is no significant difference in the scores for Path Holistic Experience (F5) and Node User-Friendliness (F6).

Figure 7 depicts the score gradient distribution of the third-level indicators for each unit. The Waterfront Cleanliness Index (F2-3), Waterfront Aesthetic Index (F4-1), and Activity Support (F6-1) have reasonably high values in the vast majority of units according to data statistics.

However, the scores for Traffic Convenience (F1-2), Amenity Density Index (F2-6), Recreational Facility Density (F2-7), Recreation Space Color Comfort Index (F4-3), and Facility Friendliness (F6-2) are relatively low, indicating that each unit should prioritize the optimization of recreational experience based on improving spatial quality.

4.1.2. Spatial Quality (SQ)

The quality of waterfront space is the “invisible engine” that powers the experience. It directly impacts whether consumers can immerse themselves in it by combining the physical environment, functional design, and narrative aspects. The four primary aspects of Spatial Quality (SQ) are Accessibility, Spatial Coziness, Water Proximity, and Visual Attraction. The gradient of the overall values of these factor is the following: Water Proximity (average F3 = 3.82) > Visual Attraction (average F4 = 3.82) >> Spatial Coziness (average F2 = 3.00) > Accessibility (average F1 = 2.82). The scores and spatial distribution may offer new insights into the waterfront vitality and the relationship between spatial aspects.

(1)

Accessibility (F1)

The average value of Accessibility (F1) is 2.824, ranking the last among the six indicators when compared to the average scores of other indicators. Accessibility (F1) is evaluated through Attainability (F1-1) and Traffic Convenience (F1-2), which can demonstrate the degree of integration between waterfront space and urban functions, as well as the convenience of users’ actual experience. Compared to Traffic Convenience (F1-2), which has an average value of 1.99, Attainability (F1-1) has an average value of 3.41. The gradients and distributions of each unit are shown in Figure 8.

Units with high Accessibility (F1-1) frequently indicate that access to waterfront spaces is more straightforward and relatively quick, attracting more visitors and considerably contributing to the location’s vitality. Furthermore, units with limited accessibility are often disregarded for activities in waterfront settings. Using accessibility data to identify “blind spots” in waterfront areas (such as inadequate pedestrian accessibility in certain areas) can guide the larger-scale optimization of walkways or transportation infrastructure. The results show that 43.9% of the units have high Attainability (F1-1 ≥ 4), with an emphasis on recreational pathway turning points such squares, bridges, entrances and exits to scenic spots, etc. (units 157, 02–04, 34–37, etc.). A total of 18.2% of the routes in units with poor Attainability (F1-1 < 2) are waterfront trails that are isolated from city traffic. User satisfaction and duration of stay are directly impacted by Traffic Convenience (F1-2). However, only three units (units 54, 62, and 63) have exceptional traffic convenience (F1-2 > 4), and 27.14% of the units have low assessments (F1-2 < 2). This is largely consistent with the traffic conditions observed during the field survey: public transportation stops and public bicycle rental points are concentrated in the two major business districts (units 57, 08–14) of Harmony Plaza (a premier shopping, dining, and entertainment destination along east bank, with the famed 500 m-long LED canopy) and Suzhou Center. However, the walking distance from each traffic node to the waterfront space remains 700–1200 m.

(2)

Spatial Coziness (F2)

The average value of Spatial Coziness (F2) is 3.002, ranking fourth among the six indicators when compared to the average scores of other indicators. Spatial Coziness (F2) is made up of seven indicators with considerable variations in their average values and a noticeable gradient in the average values of each indicator: Waterfront Cleanliness Index (4.36) >> Sky Openness Index (3.24) >> Water Clarity Index (2.91) ≈ Green View Index (2.90) > Water Exposure Index (2.69) > Recreational Facility Density (2.06) > Amenity Density Index (2.00). The gradients and distributions of each unit are shown in Figure 9.

The Water Clarity Index (F2-1), Water Exposure Index (F2-2), Waterfront Cleanliness Index (F2-3), Sky Openness Index (F2-4), and Green View Index (F2-5) all have high scores for the majority of units. The Amenity Density Index study, however, reveals a considerable variance in the score of each unit (F2-6). The Amenity Density Index values for over half of the units are low (F2-6 ≤ 2). Ligong Causeway (units 30–36) on the south bank and Harmony Plaza and Times Square (units 08–12) on the northwest side are the highest density places for public facilities. The overall Recreational Facility Density (F2-7) is poor. A polarization between inactivity (units 31–36) and inadequate supply (units 02–16) is seen in the irrational allocation of many rest areas.

(3)

Water Proximity (F3)

The average value of Water Proximity (F3) is 3.816, ranking first among the six indicators when compared to the average scores of other indicators. The Waterfront Facility Coverage Index (F3-1) and Waterfront Safety Index (F3-2) are used to represent the Water Proximity (F3). The average score of the later (3.84) is rated higher than the latter (3.79). The gradients and distributions of each unit are shown in Figure 10.

Of the 70 units, 17.14% have insufficient facility coverage (F3-1 < 3), while 43 enjoy an outstanding Waterfront Facility Coverage Index (F3-1 ≥ 4). Low-scoring units were concentrated on the north bank of JLSA, whereas over two-thirds of units received good safety ratings (F3-2 ≥ 4). According to the common understanding of space–experience interaction in waterfront areas, tourists’ awareness of the water bodies is diminished when certain locations lack waterfront amenities, which limits their engagement in waterfront activities. More specifically, as some units have high spatial closure and blocked visibility on the nearby walking path, making it difficult to see water bodies (units 05–07), or that the absence of safety maintenance facilities on either side of the trail makes it challenging to preserve the water experience (units 11–15).

(4)

Visual Attraction (F4)

The average value of Visual Attraction (F4) is 3.746, ranking second among the six indicators when compared to the average scores of other indicators. The Visual Attraction is represented by Waterfront Aesthetic Index (F4-1), Recreation Space Aesthetic Index (F4-2), and Recreation Space Color Comfort Index (F4-3). The former has a relatively high overall evaluation (4.19), whereas the latter two elements have significant variation in evaluations between units. The gradients and distributions of each unit are shown in Figure 11.

A total of 80.3% of the units are continually distributed (units 02–23, 45–53) and have good Waterfront Aesthetic Index assessments (F4-1 ≥ 4), providing a comprehensive experience for dynamic touring. High-score units for Recreation Space Aesthetic Index (F4-2 ≥ 4) have a discontinuous distribution. The lack of harmony between man-made and natural landscapes is directly linked to 17.14% of units with low ratings (F4-2 < 3) (units 4,6,7, 28–29). The northeast and south sides of JLSA are adjacent to several units with low Recreation Space Color Comfort Index scores. This is a consequence of the fact that natural areas dominate the northeast shore space (units 60–64). However, certain artificial landscapes on the south bank presently have restricted recreational space, owing their proximity to enclosed management areas (units 29–33).

4.1.3. Spatial Experience (SE)

In essence, the vitality of waterfront areas is a result of spatial experience-driven behavioral agglomeration. When the waterfront is combined with a variety of amenities (commercial, recreational, and cultural) to create a positive experience, it will trigger various behaviors, creating a perception-feedback cycle that will continuously increase the area’s attractiveness. Conversely, experience rupture can cause the behavioral link to break, resulting in a decline of vitality. There are four indications in this process that have an impact on the Path Holistic Experience (F5) and Node User-Friendliness (F6).

(1)

Path Holistic Experience (F5)

When compared to the average scores of the other factors, Path Holistic Experience (F5) ranks fifth out of the six indications with an average value of 2.877. Path Holistic Experience (F5) is evaluated by assessing Continuity (F5-1) and Trail Richness (F5-2). The average score of Continuity (3.07) outperforms the average score of Trail Richness (2.67). The gradients and distributions of each unit are shown in Figure 12.

While over half of the units obtained good continuity ratings (F5-1 ≥ 3), low continuity units were found along the east and south shorelines. This is due to the fact that the east shore sections adjacent to residential property have fewer branch roads (Units 19–29), while the business blocks on the south shore have insufficient links with nearby units (units 33–44, 65–69). The majority units along the north and south banks are of High Trail Richness units (F5-2 ≥ 4) (units 14–18, 40–50, 62–65). The northeast and southwest banks’ pathways (units 04–13, 33–37) are mainly linear, with simple shapes and functions. Field investigations indicate these paths, due to their monotonous and constant design, provide little stimulation to visitors on their route. Given the absence of important activity nodes that connected them, these walkways generally serve a pass-through function, rarely encouraging various activities or prompting loitering on the waterfront.

(2)

Node User-Friendliness (F6)

Node User-Friendliness (F6) has an average score of 3.043, placing it third out of the six indicators. Activity Support (F6-1) and Facility Friendliness (F6-2) represent this factor (Figure 13). The former’s overall evaluation is outstanding (4.06), but the latter’s average score (1.09) is far from ideal.

The spatial distribution of high-scoring Activity Support (F6-1) areas is more uniform, and the evaluation findings show that the state is generally outstanding. The western shore region has a significantly higher overall Activity Support (F6-1) score than the north and southeast coastline regions. Units 43–60, for example, provide wide beachfront paths and wide open green spaces, allowing for unlimited activities. The north shore (units 61–64, 01–10) and southeast shore (units 20–30) have less space and amenities, and automotive and pedestrian paths are frequently congested. The overall ranking for Facility Friendliness (F6-2) is low. The commercial zones Gate of the Orient and Times Square have the highest concentration of recreational and fitness facilities, with units 51–58 and 15–19 accounting for over half of the total number of facilities in the study area. The remaining units (02–06, 17–25, 28–31) are dominated by open green spaces and trails, with minimal recreational facilities, making them unsuitable for various public activities.

4.2. Intuitive Representations of Vitality

4.2.1. Gradient Overview

The gradient of Intuitive Representations of Vitality (Vi) depicts the pedestrian flow heat distribution over a given time period [49]. As illustrated in Figure 14, 70 units are classified into three levels based on Vi value: high (7.11 ≤ Vi), medium (4.77 ≤ Vi < 7.11), and low (0.00 ≤ Vi < 4.77). However, several units (units 08–11) had significant heat-value spikes and vitality spillover effects due to their close proximity to major commercial areas and venues of cultural activity. High-Vi units exhibit continuous vitality by excelling in both recreational space foundations and experience quality. The quantity of medium-Vi units fluctuates just slightly between weekdays and rest days. The variation in these unit numbers corresponds to the amount of high-Vi units, indicating potential synergies: they have good spatial design but might benefit from more experience features. Areas with low-Vi continuously perform worse in terms of both recreational value and spatial quality. Interestingly, high-Vi units exhibit a bell-curve trend within a single day, whereas low-Vi units follow a U-shaped distribution pattern. This dynamic shift reflects the temporal rhythms of urban public space efficiency and is very consistent with public leisure activity patterns.

4.2.2. Dynamic Characteristics

The overall gradient structure of Vi varied significantly in terms of time between weekdays and rest days (Figure 15). While the two peaks on rest days are extremely close, Vi changes on weekdays and rest days exhibit clear double-peak characteristics. The following are the criteria for units for each Vi gradient. There are 43 low-Vi units (about 62%), 10 medium-Vi units (14%), and 17 high-Vi units (24%) on weekdays. Between 11:00 and 21:00, respectively, the quantity of High-Vitality units increases. Low-Vi units always predominate on weekdays, with the lowest percentage of medium-Vi units. Although it fluctuates widely, the overall Vi level is modest. On rest days, there are 31 low-Vi units (about 44%), 16 medium-Vi units (23%), and 23 high-Vi units (33%). The number of high-Vi units increases between 15:00 and 19:00, respectively. Rest days, on the other hand, have significantly different vitality characteristics: the number of high-Vi units increases by about 35% when compared to weekdays, the fraction of low-Vi units decreases roughly 28%, and the overall Vi level improves significantly.

The thermal gradients of 70 units reveal theme-based clustering patterns, with leisure-consumption zones rising on weekends and commercial hubs dominating during the week. Suzhou Culture and Arts Centre—Suzhou Contemporary Art Museum district (units 08–15) exhibits high levels of pedestrian flow heat throughout the weekdays, while on weekends it concentrates around the Gate of the Orient and Times Square (units 53–54).

4.3. Comparative Analysis

4.3.1. Comparison and Enhanced Vitality Grading

Different levels of granularity offer an original foundation for comparative analysis. Vitality of Waterfront Recreational Space (V) (intrinsic vitality) employs a 5-level gradient classification for high-resolution internal mechanism analysis. This fine-grained technique accurately catches small changes in vitality levels. Intuitive Representations of Vitality (Vi) (extrinsic vitality) adopts a three-level gradient classification, recognizing that external manifestations have distinct clustering properties and simplified gradients might show substantial distinctions. This differential design offers a straightforward representation of external performance while maintaining the accuracy of internal mechanism analysis. A correlation between two sets of results can be established with the aid of data standards and visualization tools. The quantitative differences in vitality gradient classifications are compared in this work using a mix of statistical analysis and geographical overlay mapping. Among the three gradient datasets, Spatial Quality (SQ) shows a significant difference compared to Spatial Experience (SE). The High-Vitality units group has a more significant score set in Spatial Quality (SQ_H = 3.82) than in Spatial Experience (SE_H = 3.79). In detail, Water Proximity (F3) exhibits a substantial disparity across the three gradient units, followed by Visual Attraction (F4) also standing out prominently. Path Holistic Experience (F5) and Node User-Friendliness (F6) also demonstrate certain variations. The disparities in Accessibility (F1) and Spatial Coziness (F2) are relatively less pronounced.

• V Gradient (1–5 levels) shows a flattened decay, with Level 3 accounting for 42%, which indicates multi-tiered radiation in spatial vitality (Figure 15).
• Vi Gradient (1–3 levels) displays a power-law distribution, with the top 20% of data points accounting for 65% of vitality values (Figure 16).

Figure 16. Comparison of V and Vi.

Figure 16. Comparison of V and Vi.

Comparing the evaluation findings between V and Vi reveals a striking consistency in the vitality gradient classification: 18 out of 19 units (Approximately 90%) in 1st–2nd V gradients totally overlap with 20 high-Vi units. This 90% overlap rate not only verifies the dependability of the vitality measurement approach used in this study for area identification, but it also shows substantial convergence between the two evaluation methods when analyzing recreational vitality levels in waterfront environments. Further investigation revealed that 82.8% units (24 of 29 units) of low-Vi gradient are entirely overlain by fourth-fifth V gradients (39 units), with only 6 units (units 4, 19, 56, 66, 68, 69) revealing gradient classification differences. The significant level of alignment between V and Vi in crucial vitality gradients lends more evidence to the validity of vitality evaluation.

To compare the results of both methods (three-level and five-level gradient divisions), align the data and re-examine the breakpoints for each gradient. It is important to recognize and discuss the effects of outliers, including the obvious agglomeration influence of large commercial complexes and cultural amenities on regional vitality [50]. The vitality of the 70 units can be categorized into three gradients based on the data comparison and geographic coverage overlay (Figure 17).

(1)

High-Vitality Unit Group

• High-Vitality units satisfying both conditions: 3.50 ≤ V < 4.50 and Vi ≥ 7.11
• The High-Vitality unit group consists of 18 units, with a mean vitality value V_H = 3.83125, a variance of approximately 0.174, a standard deviation of approximately 0.417, a range of 1.11, and the highest degree of dispersion among the three groups.

(2)

Medium-Vitality Unit Group

• Medium-Vitality units including: 4.77 ≤ Vi < 7.11 and units not covered by High-Vitality units or Low-Vitality units
• The Medium-Vitality unit group consists of 29 units, with a mean vitality value of VM = 3.17625, a variance of approximately 0.163, a standard deviation of approximately 0.404, and a range of 1.12, indicating a high degree of dispersion.

(3)

Low-Vitality Unit Group

• Low-Vitality units satisfying both conditions: 2.00 ≤ V < 3.00 and 0.00 ≤ Vi < 4.77
• The Low-Vitality unit group consists of 23 units, with a mean vitality value VL = 2.66625, a variance of approximately 0.134, a standard deviation of approximately 0.366, and a range of 0.92, indicating a low degree of dispersion.

4.3.2. Decoding Waterfront Vitality: High-Vitality Units

The data fluctuations of High-Vitality unit group are higher than those of the other two groups. The range, standard deviation, and interquartile range (IQR) of the Vi data set are all more than an order of magnitude larger than those of the V data set, indicating that the variability in Vi data set is significantly higher. Vi data set has a wide range due to exceptional values, while there are no obvious outliers in the V data set.

• V data set: The range is approximately 0.818, the standard deviation is about 0.269, and the interquartile range (IQR) is roughly 0.513. The data set exhibits low dispersion and is concentrated in distribution.
• Vi data set: Range approximately 33.722, standard deviation approximately 9.382, interquartile range (IQR) approximately 10.563, indicating high dispersion and significant fluctuations.
• While the median of F1 is comparatively low in the High-Vitality unit group, the medians of F3, F4, and F6 are considerably high. While F2 has a relatively moderate dispersion degree, F3 and F4 have rather big dispersion degrees. There are a few outliers in F1 that are comparatively far from the box, more outliers in F2 that are dispersed on the side of lower values, and a definite number of outliers in F3 through F6, but their distribution is more dispersed (Figure 18).

Figure 18. Data distribution of High-Vitality unit group.

Figure 18. Data distribution of High-Vitality unit group.

• The waterfront vitality of this group is driven by a combination of “Spatial form optimization,” “Social and humanistic interaction” and “Ecological base support.” It not only meets practical requirements, but also increases user loyalty through humanistic care and environmental integration.
• High-Vitality waterfront units are easily accessible and recognized, providing efficient public transportation and a large range of supplementary amenities, according to visitor feedback. Each unit is dominated by three spatial categories: leisure walkways, urban squares, and recreational green spaces. They have notable waterfront features that provide numerous chances for close-water activities as well as a dense urban interaction from a distance. The public is free to participate in a variety of recreational activities since there are plenty of areas that can accommodate people of all ages and a wide range of activities. Additionally, it promotes positive interactions between individuals and environments by offering visitors adequate thematic activities and items. Notably, commercial and cultural activity in commercial complexes and cultural hubs causes agglomerated High-Vitality states to spread outward. The presence of some outliers is associated with substantial commercial cultural activity.

4.3.3. Decoding Waterfront Vitality: Medium-Vitality Units

The data fluctuations are moderate when compared to the other two groups. The range, standard deviation, and interquartile range (IQR) show that the variability in Vi data set is significantly higher than in V data.

• V data set: Range approximately 1.508, standard deviation approximately 0.371, interquartile range (IQR) approximately 0.371, the degree of dispersion is low, and the data set distribution is concentrated.
• Vi data set: The range is approximately 9.167, the standard deviation is about 1.996, and the interquartile range (IQR) is roughly 2.027, with a relatively high degree of dispersion and significant fluctuations.
• F3 and F4 feature comparatively high means and medians, along with a significant amount of data dispersion and a few outliers. F2 has outliers but also a shorter box, which suggests a lesser data dispersion. F1, F5, and F6 contain comparatively symmetric data distributions and close medians and means (Figure 19).

Figure 19. Data distribution of Medium-Vitality unit group.

Figure 19. Data distribution of Medium-Vitality unit group.

• The waterfront vitality of this group is driven by a combination of “Social and humanistic interaction” and “Natural-Artificial transition harmony”. The vitality of these units is positively influenced by surrounding High-Vitality units and holds great potential. With appropriate guidance, it can also radiate vitality to some extent toward lower-vitality areas.
• Medium-Vitality units are frequently found near High-Vitality nodes and are significantly impacted by their radiation/spillover effect. Their vitality fluctuates between rest and work days, indicating a lack of long-term tenacity. The spatial categories of Medium-Vitality units are more complex than those of High-Vitality units, and they often comprise at least two of the three types: leisure walks, urban squares, and recreational green areas. They also feature more composite spatial attributes than Low-Vitality units. An on-site study found that Medium-Vitality units have dynamic user activity patterns and shorter dwell periods. The combination of manufactured surroundings and natural water bodies provides sufficient Spatial Coziness (F2) and Water Proximity (F3), but these units struggle to handle large-scale, extremely vibrant events due to their often limited waterfront areas and limited space capacity.

4.3.4. Decoding Waterfront Vitality: Low-Vitality Units

The degree of data fluctuation of the Low-Vitality unit group is minimal when compared to the other two groups. The variability in Vi data set is substantially greater than in V data set, as evidenced by the range, standard deviation, and interquartile range (IQR). But compared to the other two groups, the fluctuations are relatively small.

• V data set: Range approximately 0.948, standard deviation approximately 0.245, interquartile range (IQR) approximately 0.315, the degree of dispersion is low, and the data distribution is concentrated.
• Vi data set: The range is approximately 2.833, the standard deviation is about 0.750, and the interquartile range (IQR) is roughly 0.917, with a relatively low degree of dispersion and relatively low fluctuations.
• Overall, the median-lines and means of F3 and F4 are relatively high, the degree of data dispersion is significant, and there are some outliers. The F2 box is short, with little data dispersion, but there are some outliers. The median lines and means of F1, F5, and F6 are reasonably close, and the data distribution is fairly symmetric (Figure 20).

Figure 20. Data distribution of Low-Vitality unit group.

Figure 20. Data distribution of Low-Vitality unit group.

• The vitality is primarily driven by “ecological experiences” and the spillover effect of surrounding vitality. Some units are typically designated as leisure green spaces or eco-friendly leisure trails; however, half of the units have long been ignored in terms of management, resulting in concerns like rubbish pollution, vegetation overgrowth.
• Low-Vitality units are unable to sustain considerable activity clusters during weekdays and on weekends, and most remain dormant for extended periods of time since they are distant from commercial or cultural areas and have monotonous structures. According to visitor feedback, the lack of entertainment venues and services leaves the experience essentially blank, while paths and nodal spaces are disjointed and stiff, preventing waterfront activities from taking place. Although it serves a useful purpose as a landscape backdrop for the surrounding dynamic areas, the lack of spatial vitality has greatly decreased its overall attractiveness.

5. Discussions

5.1. Reinterpreting Waterfront Vitality Through Spatial-Experience

Traditional waterfront vitality evaluation studies frequently focuses on one-dimensional or qualitative techniques, which lack systematic quantitative backing through multi-method integration [51]. Many scholars have acknowledged this constraint and launched creative attempts in recent years, causing the profession to gradually shift from “subjective perception” to “objective measurement” [52]. This study is an innovative endeavor under this changing backdrop. It proposes a multi-method integrated evaluation framework, along with comparison analysis, to accurately identify and systematically analyze weak vitality links, propelling waterfront vitality research to a novel stage of “comprehensive evaluation + precise identification.”

By developing a comparative analysis framework of Vitality Evaluation (V) and Intuitive Representations of Vitality (Vi), this study not only effectively avoids the subjective bias presented in conventional evaluation based on manual observation, but it also systematically compensates for static evaluation’s limitations in revealing spatial vitality. This technology-driven evaluation method offers a more scientific quantitative way for investigating the viability of waterfront areas. It incorporates the concept of system synergy, emphasizing that the creation of waterfront space vitality is the result of the coupling and interplay of numerous variables. According to research findings, waterfront vitality exhibits the following effects and mechanism:

• Nonlinear Leap Effect: Low-Vitality locations achieve cross-regional vitality transmission through landscape support functions, and the waterfront activities display a dynamic pattern of continuity and leaps.
• Spillover and Decay Effect: Periodic “cultural-commercial activities” provide spillover vitality effects, lead to a vitality radiation zone spanning two–four geographical units, with strength decreasing with distance.
• Dynamic Coupling Mechanism: Through long-term interactions, the dynamic linkage of natural and cultural landscapes produces persistent vitality stimulating effects and strengthens place identity. The research of waterfront areas, including the Seine River in Paris and the Qinhuai River in Nanjing, has extensively validated this impact [53].

It is noteworthy that this study demonstrates that “periodic cultural–commercial activity” may lead to a notable vitality spillover phenomena, resulting in a leapfrog in vitality levels over two to four nearby units. This effect has a fundamental stimulating and radiating impact on the vitality of waterfront spaces, which is in complete agreement with the main contention of previous research on the superior activation efficacy of Cultural Consumption Pioneer Zones (CCPZs) over commercial facilities [54]. Therefore, the external radiation effect of High-Vitality units should be utilized. The spatiotemporal extent and intensity of vitality spillover can be determined by gathering and combining multidimensional data, which enables the strategic placement of inexpensive, non-commercial public service facilities and activity venues in neighboring units inside the “spillover zone.” Little Island Park in New York (art installations, open-air plays, etc.) and Shanghai’s Suzhou River (Dragon Boat Riverside Scroll pop-up market) both rely on recurrent cultural events and markets to generate active waterfront environments, leveraging the Spillover and Decay Effect. In addition to supporting a variety of engagement activities in the region, it can effectively minimize spatial exclusion caused by economic constraints, resulting in a more comprehensive improvement of waterfront experiences.

5.2. Limitations and Future Research

Taking the waterfront area of JLSA as an example, this study provides an empirical evidence and theoretical exploration for analyzing the waterfront space vitality based on the space–experience interaction. However, as a multidimensional composite concept, urban vitality research naturally incorporates interdisciplinary perspectives from sociology (human behavioral interaction), geography (spatial form distribution), economics (commercial benefit-driven), and urban planning (design intervention strategies) [55]. This study has limitations due to the dual influence of inherent barriers between disciplines and dynamic urban construction, most notably fragmentation of ecological, social, and economic data, as well as deviation in data collection produced by rapid urban regeneration. While the analysis framework takes into account the fundamental composition of waterfront areas, socioeconomic factors are still incomplete.

Future research may establish a multidimensional analysis framework that incorporates social equity and economic factors, as well as introduce cutting-edge technologies like digital twins to strengthen dynamic simulation and optimization of spatial experience and improve the closed-loop technical system: Evaluation, Diagnosis, and Intervention Simulation Integration. Although the current data collection covers the entire year cycle, in the future, data dimensions can be expanded through refined grouping (such as by season or extreme weather type), a more robust vitality database can be constructed, and cross regional, cross type, and all season tracking and comparative research can be conducted.

The space–experience evaluation methodology developed for this study divides the macro concept vitality into operational variable aspects with unambiguous weight allocations. The findings provide a clear aim for shifting research from descriptive evaluation to diagnostic design, as well as a solid empirical foundation for scientifically developing optimization strategies. Based on the interactive relationship between spatial elements and recreational experience, it is possible to scientifically establish a dynamic response mechanism between user demands and spatial quality optimization [56,57]. In subsequent procedures, the target quantification module or Smart Map can be used based on the assessment system. It uses a three-dimensional “service guidance control” network to drive targeted modifications while simultaneously gathering and integrating visitor feedback with quantitative data. By superimposing this feedback on geographical vitality gradients, the method allows for precise diagnosis of inactive zones and guides the construction of hierarchical water-friendly areas, enhancing the sense of place and boosting long-term environmental sustainability.

The study offers a comprehensive and expandable analytical approach. By conducting empirical study on China’s largest urban waterfront park JLSA, which has complicated compositional elements, the framework’s high flexibility and inclusivity in complicated real-world contexts have been fully validated. Moreover, given that the framework is expandable and can be adjusted during establishment based on evaluation objectives and geographical characteristics, it can be applied to a variety of water bodies, including rivers, streams, and even coastal waterfronts.

6. Conclusions

Based on the co-construction of spatial features and human experiences of vitality, the study develops an evaluation method for waterfront recreational space from a space–experience interaction perspective, as well as a compare analysis. By identifying the factors that influence vitality in the two dimensions of Spatial Quality (SQ) and Spatial Experience (SE), a three-layered evaluation framework and assessment method for systematic study has been built. This approach reexamines the intrinsic relationship between physical space quality and experiential perception. Building on previous research on the vitality of waterfront areas, it underlines how the public experience perspective brings humanistic care into the quantitative examination of space.

This study develops a scalable framework for vitality evaluation, providing a critical scientific foundation for the precise and dynamic place-making of waterfront areas. During the evaluation system’s formative phase, a combination of word frequency screening and factor analysis was used to screen and filter the high-frequency publications and vocabulary data crawled by the machine, improving the quality and reliability of factor collection and avoiding misleading conclusions due to data quality issues. This approach fundamentally reinforces the basic concept of sustainable development continually, in addition to providing a precise tool for the scientific management of waterfront areas. It effectively preserves the waterfront area’s resilience while continually promoting regional vitality and social–cultural identity through dynamic assessment and the guidance of precise decision-making, offering a long-term model for human–water coexistence and sustainable urban development.

This study uses multimodal data integration to demonstrate the significant connection between “spatial-data” and “experience-perception” in waterfront recreational places. It presents a two-dimensional comparison of Vitality Evaluation and Intuitive Representations of Vitality, proving the validity of the proposed evaluation methods and allowing overlay analysis to drive spatial design for vitality. This approach not only increases assessment efficiency, but it also improves the scientific validity of study results by combining objective data with subjective experience. By comparing and decoding the spatial disparities of influencing factors, it proposes targeted optimization strategies for the precise management of sensitive and vulnerable waterfront spaces. This process helps to move waterfront recreation space renewal upward from the traditional design and toward an evaluation/diagnosis-driven design paradigm.