Resilient Renewal of Aging Parks in High-Density Cities: Integrating Performance-Based Design and the Environmental Overlay Method in the Wuxi Case

Abstract

Climate change exacerbates challenges for old urban parks in high-density cores, intensifying urban heat islands and overcrowding hazards and causing limited extreme weather resilience. These parks face climate maladaptation, urban health risks, and reduced adaptive capacity. This study applies performance-based urban design through an “environmental analysis Overlay method,” integrating space syntax, CFD-Phoenics wind simulation, and solar analysis to translate climate adaptation, urban health, and urban resilience dimensions into measurable indicators including ventilation efficiency, crowd dispersion comfort, and flexible space capacity. Using Chengzhong Park in Wuxi as a case study, the method employs a diagnosis–optimization–validation process. Results demonstrate substantial improvements: (1) Climate: Problematic wind areas (>4 m/s or <0.5 m/s (stagnant)) decreased from 30% to 11%, while comfortable wind zones (0.5–1 m/s) increased to over 30%, achieving optimal microclimate conditions 89% of the park; (2) Health: Pedestrian circulation capacity increased by 25%, and activity areas with under 3 h of winter sunlight reduced from 26% to 19%; (3) Resilience: Spatial units consolidated from 155 to 115, with global-local integration improving from R² = 0.39 to 0.64, significantly enhancing network coherence and adaptive functionality. The findings confirm that this method provides a scientifically rigorous, replicable pathway for climate-adaptive renewal of old urban parks, supporting urban resilience agendas.

1. Introduction

In recent years, rapid urbanization and high-density development have intensified both environmental and social pressures on urban public spaces [1], posing multiple adaptive challenges for parks built in the mid-20th century. On the one hand, holiday crowding often leads to spatial fragmentation, circulation conflicts, and facility overload; in dense urban cores, limited park areas struggle to accommodate the growing demand for public use [2]. On the other hand, surrounding high-rise developments obstruct solar access and ventilation corridors, resulting in stagnant airflow and intensified urban heat islands, thereby undermining ecological buffering capacity and environmental comfort [3]. In some cases, excessive commercialization has eroded ecological functions [4], while traditional single-layout models fail to reconcile diverse needs such as leisure, sports, culture, and commerce [5]. The COVID-19 pandemic further exposed deficiencies in ventilation, spatial flexibility, and equitable access [6].

Historically, parks have served as responses to social transformation since their emergence in the 19th century: London’s Victoria Park (1845) was intended to improve the living conditions of the working class, while New York’s Central Park (1858) provided middle-class citizens with a natural retreat [7,8]. In China, urban green spaces evolved from imperial and private gardens, to public green areas in the Republican era, and later to welfare parks during the socialist construction period [9]. After the Reform and Opening-up, parks began to integrate multifunctionality, as seen in the ecological–activity synthesis of Beijing Olympic Forest Park [10], the vertical greening of Shanghai Jing’an Sculpture Park, and the traffic–landscape integration of Guangzhou Flower City Square [11]. Nevertheless, many mid-century parks in urban cores remain characterized by single-function and coarse-grained layouts, increasingly compressed by commercial expansion, and thus demand systematic renewal.

Existing studies have provided valuable insights into understanding and improving park adaptability, yet they often remain confined to a single dimension: spatial morphology studies emphasize layout and accessibility optimization [12]; environmental simulations focus on thermal comfort and ventilation [13]; sociological perspectives highlight behavioral and equity-related differences [14]. However, these dimensions are rarely integrated, rendering them insufficient to address the intertwined challenges of space, ecology, and function in aging urban parks [15]. In other words, there is still a clear research gap in developing a multidimensional assessment framework that can comprehensively coordinate spatial form, ecological resilience, and functional diversity.

To address this gap, this study adopts the concept of Performance-Based Urban Design [16], emphasizing quantifiable environmental and social performance as the basis for design and decision-making [17]. Within this framework, an “Environmental Analysis Overlay Method” is proposed, which evaluates three performance dimensions—Climatic Performance [18], Health Performance [19], and Resilience Performance [20]—through the integration of multi-source data (space syntax, mobility big data, CFD and solar simulations [21]). A case study of Wuxi is conducted to test the effectiveness of this approach in enhancing environmental performance and multifunctionality under high-density urban conditions, thereby offering methodological contributions and practical implications for the sustainable renewal of aging urban parks [15].

2. Literature Review

In recent years, mid-20th-century parks located in high-density urban cores have faced multidimensional challenges related to space, ecology, and functionality. Their planning logic and spatial form were largely shaped by the industrial-era concept of the “green lung,” yet they lack the flexibility and integration required to adapt to contemporary urban lifestyles [22,23]. Recent empirical studies in Chinese cities have documented these challenges extensively, with aging parks in Beijing and Xi’an showing significant infrastructure deterioration and functional obsolescence [24,25]. As urban design paradigms shift from formal aesthetics toward performance-oriented approaches, a growing number of digital tools based on environmental simulation and data analysis have been introduced into park renewal [26], providing new opportunities for transforming aging parks into high-performance, multifunctional spaces [27]. Case studies from Chengdu and Nanjing have demonstrated the potential of multidimensional assessment frameworks in guiding park renewal projects [28,29].

Among these methods, spatial syntax has become a widely adopted tool for quantitatively analyzing the relationship between spatial configuration and human behavior [30,31]. Extensive empirical applications in Chinese cities, including Changsha and Fuzhou, have validated its effectiveness in urban park accessibility analysis [12,32]. Through metrics such as connectivity, visibility, and spatial hierarchy, spatial syntax enables the identification of high-traffic routes and structural bottlenecks [33], thereby optimizing internal accessibility and functional organization [34]. Recent innovations in 3D spatial syntax analysis have enhanced the capability to capture complex three-dimensional spatial attributes in landscape systems [35]. When combined with POI data or human mobility big data, it can further capture dynamic crowd distributions and flow conflicts, offering quantitative evidence for spatial equity and peak-hour management [36,37]. Studies in Malmö and other European cities have demonstrated the correlation between space syntax measures and pedestrian behavior patterns [38]. In high-density urban environments, spatial syntax has also been applied to detect underserved areas and optimize the allocation of public resources [39]. Systematic reviews have highlighted both the achievements and limitations of space syntax applications in enhancing urban sociability [14].

Complementary to spatial syntax, computational fluid dynamics (CFD) has been widely employed for microclimate optimization. CFD can simulate wind velocity distribution, ventilation paths, and thermal comfort [40], and assess the impacts of terrain, vegetation, and building morphology on thermal environments [41]. Field-validated case studies in Chengdu and Hefei have demonstrated CFD’s effectiveness in evaluating outdoor thermal comfort in urban parks during different seasons [42,43]. It thus serves as a critical tool for mitigating urban heat islands and improving air circulation in dense city cores [44,45]. Recent applications in Mediterranean climates, such as studies in Annaba and Montería, have further validated CFD models for park thermal environment assessment [46,47]. When integrated with solar access analysis [48], visibility analysis [49,50], and pedestrian heat-mapping, CFD further supports comprehensive assessments of environmental comfort, usage potential, and safety [50]. Advanced multi-physics CFD simulations conducted with scStream (version 14) have enabled more sophisticated microclimate analysis in waterfront parks and urban squares [51].

Despite these methodological advances, systematic reviews reveal that most thermal comfort studies focus on individual squares or parks without establishing transferable design guidelines [52]. Nevertheless, three major gaps remain in current research: first, most studies focus on single dimensions such as spatial accessibility, microclimate simulation, or functional diversity, without achieving systematic cross-dimensional integration [53,54]; second, in the complex context of aging parks with layered histories and multifunctional demands, spatial syntax and CFD are often applied in isolation, making it difficult to reveal the coupled mechanisms between crowd organization, environmental comfort, and functional integration [55,56]; and third, few studies have combined the framework of Performance-Based Urban Design (PBUD) with spatial analysis and environmental simulation to establish an operational pathway for performance-oriented park renewal. These gaps constitute the core research void that this study seeks to address.

3. Materials and Methods

3.1. Study Area

This study selects an urban public garden in Wuxi as the study area (Figure 1), primarily due to its complex spatial structure, ecological pressure, and its multifaceted functional role within a high-density core area [57]. Located in Liangxi District—the historical and administrative center of Wuxi—the park area typifies the adaptability challenges faced by aging parks under intense urban development [58]. The park covers approximately 9.2 hectares and features a heterogeneous spatial composition. Its eastern section preserves traditional gardens and historic buildings dating back to 1905, designated as a protected area. The western section was originally planned as a city square but has experienced disorderly commercial encroachment since 2000. The park is enclosed on three sides by mid- to high-rise commercial buildings ranging from 24 to 266 m in height, forming a highly confined urban environment. It connects to surrounding pedestrian streets and a metro station through four main entrances, highlighting the urgent need for renewal and revitalization.

The core performance issues of the park are deeply rooted in its complex developmental history. This aging urban park in Wuxi integrates different developmental phases: the eastern traditional garden area, dating from the early modern urbanization period (1900–1949), is designated as a protected heritage district, while the western section represents the modernization and ecological civilization construction era. As a park primarily designed for leisure functions, its accessibility, balance, and ventilation capacity can no longer meet the demands of today’s surging population and diversified functional requirements, urgently requiring optimization and renewal.

These historical layers have resulted in fragmented pedestrian organization with persistent congestion at entrances and exits [59]; fragmented internal spaces with broken nodes [60] (Figure 2a); deep spatial hierarchies leading to uneven accessibility and weak wayfinding [61]; dense hedges and temporary stalls causing visual obstruction, thereby reducing spatial transparency and safety [62]; and solar access problems in the southwest area caused by a 266-m super-high-rise building (Figure 2b). The temporal mismatch between the park’s original design principles and contemporary urban demands exemplifies the broader challenges facing mid-20th-century parks in rapidly densifying urban cores [63]. At the same time, the heterogeneous park boundaries, composite functions, and vertically enclosed surroundings provide a representative context for the validation and practical application of the “Environmental Analysis Overlay Method” [53].

3.2. Research Data Sources

To ensure the reproducibility and transparency of the research, this study systematically collected multi-source data required for the optimization of high-density urban parks, covering climate, spatial, and human behavioral information. These datasets support subsequent space syntax analysis, CFD simulations, and solar radiation analysis. The specific data sources are shown in

Table 1. Data Collection Form.

Data Type	Source	Key Parameters	Collection Period
Climate Data	Wuxi Liangxi District Station, China Meteorological Administration	Summer: prevailing southeast wind 4.2 m/s, temperature 30 °C; Winter: north wind 3.9 m/s, temperature 0 °C; seasonal variation in wind direction and speed	January–December 2023 (four consecutive seasons)
Pedestrian Flow Data	Baidu Heatmap + Field Surveys	Average daily flow: ~7500 visits; peak holiday flow: 18,000–22,000 visits; main concentrations at entrances, circular pathways, and street-facing plazas	October–November 2023 (15 weekdays + 15 holidays)
Spatial Data	Wuxi Planning Institute CAD Drawings + Field Surveys	Park boundaries (including preserved historical areas and urban square zones); building heights 18.5–266 m; subdivision of landscape and building spatial units within the park	September 2023
Thermal/Solar Radiation Data	SketchUp Pro + Suric Sun Plugin	(31.5757° N, 120.3019° E) Summer solstice (21 June): 07:00–18:00; Winter solstice (21 December): 09:00–15:00; shadow duration grid refined to 1 m2	September 2023 (post-modeling simulation)
Wind Environment Data	STL Models + Phoenics CFD Software	Grid resolution 0.5–1.5 m; simulation height 80 m; turbulence model: RNG k–ε; inlet wind speed 4.2 m/s; outlet zero-pressure gradient; ground rough-wall boundary	September 2023 (post-modeling simulation)

.

All datasets were unified into a standard grid system (35 cm × 40 cm), enabling multilayer overlay analysis and ensuring the integration of spatial, environmental, and behavioral data at a consistent scale. This data collection process provides a fundamental basis for the subsequent “environmental analysis overlay method” and guarantees the quantifiability, comparability, and dynamic adjustability of performance indicators.

3.3. Technical Framework of the “Environmental Analysis Overlay Method”

To optimize the spatial quality and environmental performance of aging urban parks in high-density urban cores, this study establishes a five-step technical workflow: “Data Collection → Model Simulation → Threshold Classification → Grid Overlay → Strategy Generation,” grounded in the three-dimensional performance framework of Performance-Based Urban Design (PBUD)—Climatic Performance, Health Performance, and Resilience Performance. A uniform grid size of 35 cm × 40 cm is used to ensure multidimensional spatial data alignment, with a fixed simulation boundary of X = 600 m and Y = 900 m covering the park area. The optimization focuses specifically on internal spatial layout, functional zoning, and landscape design to enhance analysis accuracy and facilitate targeted interventions.

3.3.1. Correspondence Between PBUD Three-Dimensional Performance and Technical Methods

The three-dimensional performance criteria are operationalized through specific technical approaches to quantify climatic, health, and resilience objectives, as summarized in

Table 2. The correspondence between three-dimensional performance and software technology methods.

Performance Dimension	Core Technical Tool	Quantified Objectives
Climatic Performance	CFD-Phoenics 2022 (wind simulation) [64] + SketchUp Pro 2025 (Suric Sun plugin, solar analysis) [45] + AutoCAD 2024 (area calculation)	Improve ventilation: identify and expand comfortable wind zones (0.5–2 m/s), reduce stagnant (<0.5 m/s) and strong wind (>4 m/s) areas [65]; Mitigate urban heat island: achieve ≥1.5 °C temperature reduction in summer relative to surrounding built-up areas; Ensure winter sunlight: guarantee ≥3 h daily solar exposure in key activity areas.
Health Performance	Space Syntax (DepthmapX 0.8.0; including convex space analysis, axial analysis, visibility analysis, agent-based simulation) [31] + AutoCAD 2024 (area calculation)	Optimize accessibility: maintain relative asymmetry (RA) of key nodes between 0.07–0.15; Reduce crowding; Enhance inclusivity: eliminate age-based spatial barriers.
Resilience Performance	Multi-layer Overlay (CFD wind map + solar shadow map + pedestrian heat map + space syntax accessibility map) + AutoCAD 2024 (area calculation)	Identify adaptive spaces: prolonged shade zones and safe evacuation routes; Integrate fragmented units: reduce fragmented areas by ≥50% to improve spatial and functional continuity.

.

Convex Space Analysis (Figure 3a) is carried out by dividing the park area into 155 units based on functional and enclosure structures [66]. Spatial permeability and entrance hierarchy are evaluated using total depth and relativized asymmetry (RA). Lower total depth implies higher integration and ease of movement, whereas high RA indicates excessive depth from activity nodes, reducing spatial efficiency. The calculation formula is as follows:

$$Mean Depth\left(a1\right)={\displaystyle \frac{Total Depth\left(a1\right)}{n-1}}$$

(1)

$$Relativized Asymmetry\left(a1\right)={\displaystyle \frac{Mean Depth\left(a1\right)-1}{\frac{n}{2}-1}}$$

(2)

Axial Analysis (Figure 3b), a core space syntax technique, assesses global and local integration (HH and R3) via an axial map generated from the site’s network. Integration maps use a color spectrum (red = high, blue = low) to support spatial design decisions [67].

Agent-Based Simulation (Figure 4a) models pedestrian movement along integrated paths, correlating with Baidu heatmap data [68]. Warm colors indicate high integration and crowd concentration, supporting urban centrality analysis [69].

View Analysis (Figure 4b) simulates visual integration to detect blind spots and fragmented zones—critical for ensuring safety and accessibility for children and the elderly [70]. It computes mean and standard deviation of integration across sightlines.

Wind Environment Simulation (Figure 5a) uses CFD-Phoenics with a CityEngine-generated 3D model to assess ventilation and thermal comfort. Parameters include RNG k–ε turbulence model, structured boundary conditions, and a domain of 600 m × 900 m × 80 m. Mesh refinement targets pedestrian height (Z = 1.5 m) and building edges. Results classify zones by wind speed: stagnant (0–0.5 m/s), suitable for pedestrians (0.5–1.5 m/s), and ideal for ventilation (2–4 m/s) [64]. Outputs are overlaid with other performance layers to guide planting, paving, and spatial reorganization.

Sunshine Analysis Simulation (Figure 5b): Given the presence of surrounding super-high-rise buildings, which cast extensive shadows that significantly compromise spatial usability and microclimate conditions in high-density urban parks, effective sunlight utilization becomes crucial for spatial layout optimization.

All simulation results—exported as raster images—are integrated with space syntax and environmental data to form a coupled spatial basis for optimizing pathways, ventilation corridors, and sunlit areas. This systematic coupling ensures that renewal strategies are theoretically grounded and empirically validated.

3.3.2. Suitability Threshold Classification

Based on PBUD objectives, simulation results are classified into “Performance Risk Zone”, “Performance Optimization Zone”, and “Transition Zone” to identify deficiencies and guide design interventions. The classification standards integrate multiple performance dimensions with adapted thresholds (

Table 3. Performance Zoning across Density, Accessibility, Wind, and Solar Exposure.

Analysis Type	Risk Zone	Optimization Zone	Transition Zone
Space Syntax Pedestrian Heat Map (Health) [74]	Red-orange area (congestion)	Green area (comfortable flow)	Yellow area (moderate flow)
Space Syntax (RA, Health) [75] Based on percentile distribution of site-specific RA data (Table A1)	(low integration): RA > 0.07	(moderate integration): 0.04 ≤ RA ≤ 0.07	(Highly integrated): RA < 0.04
Space Syntax Axis analysis [76] (Health)	Blue line segment	Red line segment	Yellow-green line segment
Space Syntax Visual field analysis [77] (Climatic)	Dark blue area	Red and yellow area	Yellow-green area
Wind Environment [78] (Climatic)	>4 m/s (strong wind risk) or <0.5 m/s (stagnant wind, heat retention)	0.5–2 m/s (comfortable for core activity areas)	2–4 m/s (green corridor adaptation; buffer strong winds)
Solar Exposure [72] (Shading; Climatic + Health)	>6 h (cold winter spaces, reduced outdoor activity)	<5 h (sufficient winter sunlight; comfortable for seasonal use)	5–6 h (seasonal adaptation via deciduous trees)

): RA values adopt a percentile-based classification approach rather than traditional space syntax theoretical thresholds, as empirical analysis revealed that site-specific RA values consistently fall below 0.1, making conventional standards (RA ≤ 0.25 for high integration) impractical for meaningful differentiation. Therefore, RA classification employs data distribution percentiles where the upper 33rd percentile represents segregated areas with relatively limited movement potential (Performance Risk Zone), the middle 33rd percentile suggests moderate accessibility (Transition Zone), and the lower 33rd percentile indicates high integration areas with optimal accessibility (Performance Optimization Zone). Wind environment thresholds are based on pedestrian comfort criteria where 0.5–2 m/s provides optimal conditions for outdoor activities, while speeds > 4 m/s or <0.5 m/s create discomfort or stagnation risks [71]. Solar exposure classification follows seasonal adaptation principles where <5 h ensures sufficient winter sunlight, 5–6 h allows deciduous tree mitigation, and >6 h indicates excessive winter exposure reducing outdoor activity [72]. Space syntax heat maps and visual field analyses use color-coded integration values where warmer colors (red/yellow-green) indicate higher accessibility and cooler colors (blue/dark blue) represent segregated or poorly connected spaces [73]. This multi-dimensional approach enables comprehensive performance evaluation while maintaining consistency with established urban design standards.

3.3.3. Implementation Details of the Five-Step Technical Workflow

This study establishes a Performance-Based Urban Design (PBUD) workflow that integrates data analysis, simulation, threshold classification, spatial overlay, and targeted optimization to guide the sustainable renewal of aging urban parks in high-density cores (Figure 6).

(1) Data Analysis: Establishing PBUD Performance Baselines

Pre-renovation site issues—including entrance congestion, fragmented landscape, age-segregated usage, and microclimate discomfort—were identified through a multi-source data collection system to support three-dimensional performance analysis:

• Spatial Data: Park topography, building layout, and functional zoning were obtained via field surveys, CAD base maps, and GIS coordinates (31.5757° N, 120.3019° E).
• Environmental Data: Wind, temperature, and sunlight parameters were derived from national meteorological records (subtropical monsoon climate in Liangxi District, Wuxi; prevailing southeast winds year round and north winds in winter; average annual wind speed of 4.2 m/s; temperature range from −7 °C to 42 °C); and on-site microclimate monitoring.
• Behavioral Data: Agent-based modeling generates virtual pedestrian flows along integrated paths, enabling validation against Baidu heatmaps. As one of the most effective space syntax techniques, crowd flow simulation reflects urban spatial centrality [69] (daily 7500–9000 visitors, weekend/holiday peak 18,000–22,000) and was combined with field photography to capture crowd aggregation and activity preferences.

(2) Model Simulation: Dimension-Specific PBUD Tools

Each PBUD dimension was simulated using its corresponding technical tools for quantitative analysis:

• Climatic Performance: CFD-Phoenics simulation was used to import STL models of the urban environment (including the park and surrounding high-rise buildings) with RNG k–ε turbulence model, fixed inlet wind speed, zero-pressure outlet, rough ground, and no-slip building walls. Simulation domain: X = 600 m, Y = 900 m, and Z = 80 m, refined at 1.5 m pedestrian height and 0.5 m near building edges. Wind vector maps were output. SketchUp Pro with Suric Sun plugin simulated summer and winter solstices (21 June, 7:00–18:00; 21 December, 9:00–15:00, hourly intervals) to generate shading duration maps.
• Health Performance: DepthmapX was used for space syntax analysis. Eighty-eight convex spaces were defined to calculate RA values and mean depth. Axial maps generated global (HH) and local (R3) integration metrics, and agent-based simulations verified pedestrian routes against Baidu heat maps. Density conversion identified accessibility bottlenecks and congestion points (e.g., entrance peak > 8 persons/m²).
• Resilience Performance: Preliminary integration of climate (wind, sunlight) and health (pedestrian flow, accessibility) layers identified potential elastic spaces (e.g., long sunlight + comfortable wind) and fragmented areas (e.g., high RA + stagnant wind).

(3) Threshold Classification: Linking to Suitability Standards

Simulation results were classified according to the thresholds in Table 1 to define PBUD performance attributes for each area:

• Wind speed: <0.5 m/s (climatic risk), 0.5–2 m/s (optimized), 2–4 m/s (transition);
• RA value: >0.07 (health risk), ≤0.04 (optimized), 0.04–0.07 (transition);
• Pedestrian density: >8 persons/m² (health risk), 1–3 persons/m² (optimized), 3–8 persons/m² (transition);
• Shading duration: >6 h (climatic + health risk), <5 h (optimized), 5–6 h (transition).

(4) Grid Overlay: Establishing a Unified Spatial Benchmark

All classified layers—including pedestrian heat maps, RA maps, CFD wind maps, and solar shading maps—were aligned to a uniform 35 cm × 40 cm grid in Photoshop 2022 (https://www.adobe.com/products/photoshop.html, accessed on 12 October 2025). The simulation boundary (X = 600 m × Y = 900 m) was strictly maintained to ensure spatial consistency and comparability across PBUD dimensions, producing multi-layer overlay maps to inform strategy generation.

(5) Strategy Generation: Targeted Optimization of Park Interior Design

Based on overlay results, three core areas were identified and targeted optimization strategies were proposed according to PBUD performance goals, using an iterative “redesign–reevaluation” mechanism:

• Performance Risk Zones (e.g., heat retention and congestion areas: <0.5 m/s stagnant wind + >6 h shading + red area (high density flow) + RA >0.07): Prioritize improving ventilation and accessibility by adding southeast-oriented wind corridors (tall trees guiding dominant winds), widening secondary paths to reduce RA ≤ 0.04, and dispersing entrance flows via 2–3 auxiliary entrances and buffer plazas.
• Performance Optimization Zones (e.g., comfortable composite areas: 0.5–2 m/s wind + <5 h shading + green area (comfortable flow) + RA ≤ 0.04): Strengthen functional diversity by arranging multi-age activity nodes (elderly exercise areas, children’s playgrounds), seating, and service facilities (water stations, rain shelters) to ensure healthy and comfortable use.
• Transition Zones (e.g., elastic buffer areas: 2–4 m/s wind + 5–6 h shading + yellow area (moderate flow) + RA 0.04–0.07): Serve as functional transition and elastic spaces, incorporating seasonal landscaping (deciduous trees) and temporary activity areas (weekend markets, science exhibitions) to enhance resilience and adaptability.

Iterative simulation verified the optimization effects, ensuring continuous improvement of PBUD three-dimensional performance. Following the identified optimization strategies, the park design was updated and re-imported into the respective simulation tools (CFD-Phoenics, SketchUp Pro, DepthmapX) to generate post-renovation performance maps, which were subsequently analyzed in AutoCAD for precise area calculation and comparative assessment of improvement percentages across all PBUD dimensions.

3.3.4. Data Visualization and Communication Strategy

Recognizing that complex environmental data requires strategic visual communication, this study employs a hierarchical visualization approach: Level 1: Diagnostic maps use color-coded zones (red = risk) to spatially represent performance deficiencies areas. Level 2: Overlay maps employ transparency gradients to visualize multi-layer conflicts, such as areas with overlapping wind stagnation, shading, and high pedestrian density. Level 3: Validation charts provide quantitative before–after comparisons, illustrating improvements in key indicators (e.g., wind comfort, RA values, solar exposure). All visualizations adhere to consistent cartographic principles: a unified 35 cm × 40 cm grid, standardized color schemes across PBUD dimensions, and explicit threshold indicators to facilitate interdisciplinary interpretation and decision-making.

4. Process and Results

4.1. PBUD Performance Diagnosis Before Optimization

Based on simulations using DepthmapX, Phoenics, and Suric Sun, combined with pedestrian flow data, the pre-renovation site exhibits significant deficiencies across the three dimensions of Performance-Based Urban Design (PBUD).

Climatic performance: The site’s microclimate system exhibits significant imbalance, with core issues in wind environment control, uneven sunlight distribution, and pronounced urban heat island effects. Under dominant summer southeast winds (Figure 7a) (average 30°, 3.3–4.2 m/s), only 15% of the site falls within the comfortable wind speed range (0.5–2 m/s). Although central open areas have potential ventilation zones with 2.0–4.0 m/s, no continuous wind corridors are established to guide airflow. Strong wind zones (>4 m/s) along the north–south commercial streets lack vegetative buffers, and the southwest corner behind buildings forms a core stagnation zone (<0.5 m/s). Combined with dense building coverage, overall stagnation areas account for 40%. In winter, under north winds (Figure 7b) (3.9–4.2 m/s), static wind zones (<0.5 m/s) are concentrated at the site edges and are not functionally adapted for low-wind activities (e.g., outdoor uses tolerant of cold but requiring minimal airflow), exacerbating discomfort. Additionally, overlapping shadows from the 266 m high-rise and surrounding 24 m commercial buildings reduce sunlight in the western activity areas to less than 3 h at winter solstice (Figure 7c), while the eastern traditional garden areas receive over 6 h (Figure 7d), creating imbalanced light distribution and further weakening winter microclimate comfort.

These climatic deficiencies were systematically captured through the overlay analysis framework. CFD wind speed mapping combined with solar duration analysis revealed critical conflict zones where stagnant wind areas (<0.5 m/s) coincided with excessive shading (>6 h), creating compound thermal discomfort zones covering approximately 18% of the site. This overlay analysis directly informed the subsequent optimization strategy targeting ventilation corridor establishment and terrain elevation interventions.

Health performance: User experience and health safety show multiple deficiencies, primarily reflected in unbalanced accessibility, high crowding risk, and low convenience. Spatial accessibility analysis based on 155 convex spatial units (

Table A1. The pre-renovation site data analysis, which includes spatial sequence number, Mean Depth, Total Depth, and RA.

No.	Mean Depth	Total Depth	RA
1	6.8571429	1056	0.076563962
2	7.3116884	1126	0.082505733
3	8.0844154	1245	0.092606738
4	9.0779219	1398	0.10559376
5	6.318182	973	0.069518715
6	6.6168833	1019	0.073423311
7	6.0389609	930	0.065868773
8	8.1298704	1252	0.093200915
9	8.6558437	1333	0.10007639
10	8.5389614	1315	0.098548509
11	8.1233768	1251	0.093116038
12	8.0389614	1238	0.092012562
13	8.818182	1358	0.10219845
14	8.3116884	1280	0.095577627
15	8.6558437	1333	0.10007639
16	9.2987013	1432	0.10847975
17	9.3831167	1445	0.10958323
18	7.4610391	1149	0.084458023
19	7.0584416	1087	0.079195313
20	7.0584416	1087	0.079195313
21	8.4935064	1308	0.097954333
22	6.1883116	953	0.067821071
23	6.4350648	991	0.071046598
24	7.1103897	1095	0.079874374
25	7.5194807	1158	0.085221969
26	6.3311687	975	0.069688484
27	7.181818	1106	0.080808081
28	7.5129871	1157	0.085137084
29	6.8636365	1057	0.076648839
30	7.7207794	1189	0.08785332
31	7.7402596	1192	0.088107973
32	7.7532468	1194	0.088277735
33	8.4220781	1297	0.097020626
34	8	1232	0.09150327
35	7.4740262	1151	0.084627792
36	8.2532463	1271	0.094813682
37	9.4480524	1455	0.11043205
38	7.8051949	1202	0.088956796
39	6.9545455	1071	0.077837199
40	8.3506489	1286	0.096086919
41	8.590909	1323	0.09922757
42	7.1948051	1108	0.080977842
43	8.1883116	1261	0.09396486
44	7.3896103	1138	0.083524317
45	7.3311687	1129	0.082760379
46	7.6623378	1180	0.087089382
47	9.0064936	1387	0.10466005
48	7.9805193	1229	0.091248624
49	8.7142859	1342	0.10084034
50	8.2207794	1266	0.094389267
51	8.9610386	1380	0.10406587
52	8.6558437	1333	0.10007639
53	9.4610386	1457	0.11060182
54	9.590909	1477	0.11229946
55	9.7337666	1499	0.11416688
56	9.0324678	1391	0.10499957
57	6.2662339	965	0.068839654
58	7.818182	1204	0.089126557
59	8.2467537	1270	0.094728798
60	9.5584412	1472	0.11187505
61	9.954545	1533	0.11705288
62	10.292208	1585	0.12146677
63	9.5974026	1478	0.11238435
64	9.6233768	1482	0.11272388
65	9.2987013	1432	0.10847975
66	9.2337666	1422	0.10763093
67	8.7402601	1346	0.10117987
68	9.0129871	1388	0.10474493
69	9.3506489	1440	0.10915881
70	8.9025974	1371	0.10330193
71	8.3766232	1290	0.09642645
72	9.0259743	1390	0.1049147
73	9.4870129	1461	0.11094134
74	9.7402601	1500	0.11425176
75	9.1558437	1410	0.10661234
76	9.1168833	1404	0.10610305
77	9.454545	1456	0.11051694
78	5.6558442	871	0.060860708
79	6.1558442	948	0.067396656
80	6.3246756	974	0.0696036
81	8.1753244	1259	0.093795091
82	7.5194807	1158	0.085221969
83	7.2077923	1110	0.081147611
84	7.9480519	1224	0.090824209
85	8.590909	1323	0.09922757
86	8.4870129	1307	0.097869448
87	8.045455	1239	0.092097446
88	7.8376622	1207	0.089381211
89	7.6298699	1175	0.086664967
90	6.9025974	1063	0.077158138
91	6.1753244	951	0.067651302
92	6.3636365	980	0.070112891
93	5.9610391	918	0.064850181
94	7.2077923	1110	0.081147611
95	6.7857141	1045	0.075630255
96	7.6688313	1181	0.087174267
97	8.2142859	1265	0.09430439
98	7.8766232	1213	0.089890502
99	7.3831167	1137	0.08343944
100	6.6298699	1021	0.073593073
101	5.7402596	884	0.06196418
102	7.7792206	1198	0.088617265
103	7.6363635	1176	0.086749852
104	7.4740262	1151	0.084627792
105	7.1233768	1097	0.080044135
106	7.9480519	1224	0.090824209
107	8.6948051	1339	0.10058569
108	8.4870129	1307	0.097869448
109	9.2597399	1426	0.10797046
110	8.8896103	1369	0.10313216
111	7.9220781	1220	0.090484679
112	8.5649347	1319	0.09888804
113	8.1233768	1251	0.093116038
114	7.5519481	1163	0.085646383
115	7.3831167	1137	0.08343944
116	7.7077923	1187	0.087683558
117	8.4220781	1297	0.097020626
118	9.0714283	1397	0.10550887
119	8.4610386	1303	0.097529918
120	6.0649352	934	0.066208303
121	7.4740262	1151	0.084627792
122	6.7532468	1040	0.07520584
123	7.0064936	1079	0.078516252
124	7.4220781	1143	0.083948731
125	9.5649347	1473	0.11195993
126	9.3701296	1443	0.10941346
127	9.0259743	1390	0.1049147
128	8.5779219	1321	0.099057809
129	7.3506494	1132	0.083015025
130	7.8636365	1211	0.089720733
131	9.272727	1428	0.10814022
132	8.9935064	1385	0.10449028
133	8.0714283	1243	0.092436977
134	6.6948051	1031	0.074441895
135	8.4025974	1294	0.09676598
136	7.7467532	1193	0.08819285
137	7.9935064	1231	0.091418386
138	6.8571429	1056	0.076563962
139	7.2012987	1109	0.081062727
140	7.9675326	1227	0.091078855
141	6.8246756	1051	0.076139547
142	8.5129871	1311	0.098208979
143	7.9870129	1230	0.091333501
144	7.9610391	1226	0.090993971
145	8.4675322	1304	0.097614802
146	6.4935064	1000	0.071810544
147	6.7467532	1039	0.075120956
148	7.6298699	1175	0.086664967
149	6.8831167	1060	0.076903492
150	10.402597	1602	0.12290977
151	10.993506	1693	0.13063407
152	10.538961	1623	0.1246923
153	9.863636	1519	0.11586453
154	9.363636	1442	0.10932858
155	8.3961039	1293	0.096681096

) shows a mean RA (relative asymmetry) of approximately 0.092, with a maximum–minimum difference of 0.0697 (Unit 151:0.1306 vs. Unit 78: 0.0609), indicating significant differences in spatial permeability (Figure 8a,b). The health performance diagnosis employed multi-layer spatial analysis to identify accessibility-density conflicts.

Crowding risk is highlighted by DepthmapX agent-based simulations (Figure 9a): during peak hours at the western metro entrance (weekends/holidays, 18:00–20:00), 80–100 people enter every 5 min, leading to densities of 8 persons/m². The western circular plaza disrupts natural circulation, causing backflows and intersections (Figure 9 b). Space syntax RA mapping overlaid with agent-based pedestrian flow simulation revealed that 12 spatial units with RA > 0.07 (poor accessibility) coincided with peak crowd density zones (Blue green area), creating bottleneck-isolation coupling effects. This overlay analysis quantified the spatial justice deficit, with 23% of the site experiencing simultaneous accessibility barriers and overcrowding risks, directly guiding the hierarchical pathway reorganization strategy.

Global integration (HH) versus local integration (HH-R3) has a low R² of 0.39 (Figure 10), reflecting a fragmented network: highly integrated areas are concentrated in the north, while accessibility in the south and southwest is weak, and weakly connected paths fail to divert pedestrian flows.

In terms of usability, Visibility Graph Analysis (VGA) conducted with DepthmapX reveals weak east–west visual connectivity (Figure 11a). The western circular space predominantly falls within low-visibility (deep blue) zones, where the hierarchy between primary and secondary paths is indistinct, thereby heightening congestion risks. These visual constraints particularly affect elderly and child users, who are more susceptible to disorientation, further intensifying accessibility barriers (Figure 11b). The viewshed analysis demonstrates a marked contrast between high-visibility zones (yellow and green) and blind spots (dark blue). To mitigate these shortcomings, spatial interventions are recommended, including the adjustment of hedge boundaries, removal of obstructive structures, and the establishment of visual corridors to enhance transparency, navigability, and overall wayfinding efficiency.

Resilience performance: The site demonstrates weak capacity to respond to external changes and internal demands, with core deficiencies in spatial structure, resource adaptation, and extreme weather preparedness. Structurally, the 155 spatial units (including four main functional units) are highly fragmented, and the fractured road network lacks integration. DepthmapX axial analysis confirms this weakness: global Integration (HH) shows activity concentrated in the north, while the south and southwest remain poorly connected (Figure 12a); the scatterplot comparing Integration (HH) and local Integration (HH-R3) reveals a low fit (R² = 0.29), indicating significant network fragmentation and insufficient global–local coordination (Figure 12b).

4.2. Core Optimization Strategies

Based on multi-layer overlay analysis (Figure 13), optimization strategies were developed to enhance the three-dimensional PBUD performance—climate, health, and resilience—through coordinated interventions. Figure 13 establishes the methodological bridge between diagnosis and design intervention. The framework operates through three integrated steps: (1) Grid Standardization converts heterogeneous environmental data into comparable risk/optimization zones using evidence-based thresholds—pedestrian risk zones (RA > 0.07), wind risk zones (>4 m/s or <0.5 m/s (stagnant)), and shadow duration (>6 h); (2) Multi-Layer Overlay spatially superimposes these classified layers to reveal compound problem areas where multiple deficits converge—such as zones experiencing simultaneous thermal discomfort, accessibility barriers, and spatial fragmentation; (3) Strategy Generation translates these overlay findings into coordinated spatial interventions that address multiple performance dimensions simultaneously. This systematic overlay logic ensures renewal strategies target root causes rather than isolated symptoms. Spatial navigation proved complex and highly vulnerable to disruption, with congestion unable to be relieved through alternate routes due to fragmented connectivity. Resource adaptation analysis revealed critical shortcomings: Suric Sun winter solstice simulations indicated that 26% of activity areas receive less than 3 h of sunlight, with western areas severely shaded by a 266 m supertall tower, while shadows from adjacent 24 m commercial buildings further reduced effective greenery and landscape space, creating a significant mismatch between solar resources and user needs.

The system’s extreme weather response capabilities were particularly inadequate, lacking adaptive buffering mechanisms such as shading facilities for high temperatures (causing site usage to drop to 30%) and dedicated evacuation routes for heavy rain scenarios. Fragmented connectivity and obstructed sightlines, especially east–west visual blockages, further compromised evacuation efficiency. Additionally, winter static wind zones (<0.5 m/s) and summer strong wind zones (>4 m/s) lacked appropriate transitional functional spaces, while limited spatial integration prevented rapid reconfiguration, leaving the site unable to accommodate dynamic user needs under varying climatic conditions.

In response to these identified deficiencies, comprehensive optimization strategies were developed to enhance the three-dimensional PBUD performance—climate, health, and resilience—through coordinated interventions that address spatial fragmentation, resource distribution, and adaptive capacity simultaneously.

• Climate Performance Optimization

Based on the updated wind environment simulation results (Figure 14), the optimization strategies demonstrate significant effectiveness. In the southeast wind-dominated simulation (Figure 14a), by maintaining 3-meter-wide continuous green corridors aligned with the prevailing summer southeast winds (30°), airflow is successfully channeled into the central area, with average wind speeds controlled at 0.7 m/s and maximum wind speeds reduced from previously extreme values to 2.5 m/s. The strategic use of tall deciduous trees (Cinnamomum camphora, 8–12 m height) along 3-meter-wide green corridors helps guide southeast winds while providing seasonal shade. Mixed camphor-holly tree-shrub belts in buffer zones further moderate wind speeds in previously problematic areas. In urban heat conflict zones (high solar exposure + strong winds + low accessibility), the introduction of undulating terrain and clustered vegetation effectively prevents vortex formation and enhances thermal regulation.

The north wind-dominated simulation (Figure 14b) shows more favorable wind environment improvements, with overall wind speed distribution becoming more uniform and maximum wind speeds being reduced to 3.2 m/s. The proportion of comfortable wind speed zones (0.5–1 m/s) has significantly increased to over 30%, primarily distributed in the central and southern areas of the site (shown in green and light blue zones). In the previously problematic north–south commercial street strong wind areas (>4 m/s), the installation of mixed camphor and holly tree-shrub belts successfully controls wind speeds within the 2–3 m/s range (shown in yellow-green zones), effectively mitigating interference with outdoor activities.

In winter-dominated north wind zones, edge grass mounds demonstrate clear effectiveness in buffering direct wind exposure, while low-wind areas (<0.5 m/s) are appropriately utilized for flexible activity spaces. The western topography elevation strategy effectively alleviates shading impacts from the adjacent 266-m supertall building, achieving optimized balance in sunlight distribution.

2.

Health Performance Optimization

The updated crowd flow simulation demonstrates significant improvements in spatial accessibility and circulation efficiency (Figure 15). The 1–2 h pedestrian flow path analysis (Figure 15a) reveals enhanced movement patterns, where high-density flow areas (red zones) now form coherent corridors connecting key activity nodes, whereas secondary circulation routes (blue paths) provide effective distribution throughout the site, eliminating previous bottleneck conditions. The updated field of view analysis (Figure 15b) shows substantial improvements in visual connectivity, with expanded high-visibility zones (green/yellow areas) replacing previous fragmented sightlines. Strategic removal of obstructive elements and hedge boundary adjustments have successfully addressed VGA blind spots, particularly enhancing east–west visual permeability across the central zones. Spatial depth analysis (Figure 15c,d) confirms the effectiveness of space unit consolidation from 155 to 115 functional areas. The improved integration is evident through the expanded medium-depth zones (green areas) that now provide better spatial continuity, while high-integration nodes (red zones) are strategically positioned at key circulation intersections to maximize accessibility.

3.

Resilience Performance Optimization

Functional Adaptation: Flexible activity areas (e.g., skate park) are placed in winter low-wind zones (<0.5 m/s); movable shading structures are installed in high-sunlight zones (<5 h shadow) to adapt to extreme heat, raising usage rates to 75%. Western shaded areas (<3 h sunlight) are mitigated with elevated terrain and vegetative buffers (Figure 16).

Spatial Elasticity (Figure 17): A central 6 m raised green area combined with terraced performance zones and sunken plazas forms a flexible system. Networked paths connect peripheral green spaces, reducing fragmentation. Primary paths also serve as emergency evacuation routes during heavy rainfall, enhancing extreme weather resilience. Scale Coordination: Ecological corridors and soft grass mounds at high-rise boundaries create a transitional buffer, reducing building dominance. The combination of elevated and sunken terrain generates varied spatial volumes, strengthening adaptability and disturbance resistance. These strategies collectively enhance connectivity, continuity, and compositeness, achieving integrated PBUD performance improvements in high-density urban park renewal.

4.3. Quantitative Validation of PBUD Performance After Optimization

The effects of the site renewal were quantitatively verified using the same simulation tools applied before optimization—DepthmapX for space syntax, Phoenics for wind environment, and Suric Sun for sunlight analysis. Results demonstrate that the three-dimensional PBUD performance—climate, health, and resilience—achieved the preset targets, with key indicators showing substantial improvements as follows:

Climate Performance: Substantial Improvement in Microclimate Comfort and Regulation. Climate optimization focused on “ventilation enhancement, wind stagnation elimination, and heat island mitigation.” As demonstrated in the CAD analysis, problematic wind areas (static zones and excessive wind speeds > 4 m/s) were significantly reduced through strategic site modifications. Figure 18a shows that, before the update, areas affected by surrounding buildings had occupied about 30% of the site, while Figure 18b reveals that after applying terrain-undulation and airflow-introduction measures, windless and strong-wind zones were significantly reduced, occupying only 11% of the site, representing a 19-percentage-point improvement.

This improvement is attributed to the 3 m wide continuous green corridors guiding southeast winds, along with mixed tree–shrub belts (camphor + holly) in strong wind areas (>4 m/s), stabilizing wind speeds along main paths (0.5–1 m/s) and central activity zones (1–2 m/s), fully meeting human comfort requirements. The terrain undulations and strategic airflow introduction effectively resolved the original problematic areas, with wind shadow zones eliminated via terrain elevation and wind corridor connectivity. The combined effect of green corridors and vegetative buffers effectively reduced temperature differences with surrounding built areas, significantly enhancing thermal regulation and achieving optimal microclimate conditions across 89% of the park area. The design shifted from scattered tree placement to functional zoning: wind corridor vegetation (deciduous trees with 8–10 m spacing) along southeast wind paths, dense multi-layered buffer vegetation (trees + shrubs + groundcover) along commercial streets, and clustered planting (>60% canopy coverage) in the southwest thermal island zones. This systematic vegetation strategy directly supported the measured microclimate improvements.

Health Performance: Enhanced Accessibility, Safety, and Sunlight Exposure. The health performance evaluation focuses on three core indicators: “pedestrian flow management, spatial accessibility, and sunlight adaptation.” Analysis of red zones in pedestrian flow comparison maps during peak traffic hours reveals that before the renovation, the site suffered from poor east–west connectivity, with pedestrian congestion areas covering 3038 square meters (15.6% of the total site area) (Figure 19a). After the renovation, under high pedestrian flow patterns, the red and yellow-green high-density pedestrian areas expanded to 3798 square meters (19.5% of the site area) (Figure 19b). This represents a 25% increase in the effective area for accommodating pedestrian gathering and circulation, significantly enhancing the site’s pedestrian capacity and accessibility.

Reorganized circulation (main paths 5–8 m; secondary paths 3 m) with direct connections from the west entrance to the central area eliminated flow conflicts, confirmed by DepthmapX agent-based simulation. Space syntax metrics improved: RA (relative asymmetry) mean increased from 0.0921 (Table A1) to 0.0904 (

Table A2. The post-renovation park area data analysis, which includes spatial sequence number, average depth, total depth, and RA.

No.	Mean Depth	Total Depth	RA
1	7.2105265	822	0.10992082
2	5.7456141	655	0.083993167
3	6.1842103	705	0.091755942
4	5.4649124	623	0.079024993
5	5.2894735	603	0.075919889
6	5.1140351	583	0.072814777
7	5.7192984	652	0.083527401
8	4.5175438	515	0.062257413
9	5.1491227	587	0.073435798
10	5.0087719	571	0.070951715
11	5.7719297	658	0.084458932
12	5.6140351	640	0.081664339
13	6.7719297	772	0.10215805
14	6.4649124	737	0.096724108
15	6.4473686	735	0.096413597
16	6.2982454	718	0.093774259
17	5.4649124	623	0.079024993
18	5.1315789	585	0.073125288
19	4.4473686	507	0.061015371
20	4.8947368	558	0.068933398
21	5.5263157	630	0.080111787
22	5.8859649	671	0.086477257
23	6.8421054	780	0.1034001
24	5.1666665	589	0.073746316
25	6.4561405	736	0.096568853
26	5.2543859	599	0.075298868
27	6.5175438	743	0.097655647
28	6.6842103	762	0.1006055
29	6.1315789	699	0.090824403
30	5.0701756	578	0.072038502
31	7.0175438	800	0.1065052
32	5.7807016	659	0.084614187
33	6.1491227	701	0.091134913
34	5.6842103	648	0.08290638
35	5.3421054	609	0.07685142
36	5.6315789	642	0.081974849
37	7.1842103	819	0.10945506
38	7.5087719	856	0.11519951
39	4.9210525	561	0.069399163
40	4.7982454	547	0.067225583
41	5.4736843	624	0.079180248
42	4.877193	556	0.068622887
43	4.8245616	550	0.067691356
44	5.9122806	674	0.086943023
45	6.4035087	730	0.095637321
46	6.6140351	754	0.099363454
47	6.6315789	756	0.099673964
48	6.4385967	734	0.096258342
49	7.1491227	815	0.10883404
50	7.1842103	819	0.10945506
51	6.4298244	733	0.096103087
52	7.2192984	823	0.11007608
53	7.3157897	834	0.11178388
54	6.3859649	728	0.095326811
55	5.8245616	664	0.085390471
56	5.8245616	664	0.085390471
57	7.5789475	864	0.11644155
58	4.8508773	553	0.068157122
59	4.8596492	554	0.068312377
60	5.3508773	610	0.077006675
61	5.4736843	624	0.079180248
62	5.6315789	642	0.081974849
63	5.1578946	588	0.073591061
64	5.4561405	622	0.078869738
65	5.6315789	642	0.081974849
66	5.2280703	596	0.074833103
67	5.8947368	672	0.086632513
68	6.0526314	690	0.089427106
69	6.4473686	735	0.096413597
70	7.1842103	819	0.10945506
71	6.3157897	720	0.094084769
72	5.4912281	626	0.079490766
73	6.8421054	780	0.1034001
74	7.2894735	831	0.11131812
75	5.4912281	626	0.079490766
76	5.7894735	660	0.084769443
77	5.2807016	602	0.075764634
78	6.5964913	752	0.099052943
79	7.1403508	814	0.10867877
80	5.2543859	599	0.075298868
81	6.622807	755	0.099518709
82	6.0789475	693	0.089892872
83	6.9385967	791	0.1051079
84	6.3684211	726	0.095016301
85	6.4385967	734	0.096258342
86	5.9649124	680	0.087874554
87	6.2543859	713	0.092997983
88	5.9210525	675	0.087098278
89	6.7894735	774	0.10246856
90	6.4298244	733	0.096103087
91	7.1403508	814	0.10867877
92	7.1491227	815	0.10883404
93	7.5175438	857	0.11535476
94	6.7894735	774	0.10246856
95	7.5175438	857	0.11535476
96	7.8421054	894	0.12109921
97	5.9824562	682	0.088185064
98	7.7192984	880	0.11892563
99	7.9561405	907	0.12311753
100	7.5877194	865	0.1165968
101	6.9298244	790	0.10495265
102	6.2982454	718	0.093774259
103	5.9035087	673	0.086787768
104	6.2456141	712	0.092842728
105	8.1578951	930	0.12668841
106	8.7368422	996	0.13693526
107	8.4210529	960	0.13134606
108	7.6315789	870	0.11737308
109	7.4649124	851	0.11442322
110	6.7105265	765	0.10107126
111	6.9035087	787	0.10448688
112	6.1754384	704	0.091600686
113	6.7543859	770	0.10184754
114	7.0438595	803	0.10697097
115	7.2894735	831	0.11131812

) (−1.8%), achieving the accessibility target; consolidated 40 space units showed a more balanced RA distribution, with reduced disparity between maximum and minimum values, enhancing spatial permeability and wayfinding. Global (HH) and local integration (HH-R3) correlation R² jumped from 0.39 to 0.64 (+64.1%), exceeding the target (≥0.6), effectively repairing the previously fragmented network and improving pedestrian flow efficiency.

Sunlight optimization significantly reduced the proportion of activity areas receiving less than 3 h of sun on the winter solstice from 26% to 19%. As illustrated in Figure 20, the spatial distribution of shadows was consolidated. The mitigation of shading, aided by the terrain elevation against the 266 m supertall building, was pivotal. This intervention enabled 81% of activity areas to receive 5–8 h of sunlight, ensuring adequate thermal comfort in winter. The comparison shows that the updated layout expands the activity area northward into zones with longer sunshine hours, creating a more contiguous and sun-rich environment.

Resilience Performance: Enhanced Spatial Integration and Extreme Weather Adaptability. Resilience performance aimed to “eliminate fragmentation and improve elasticity.” Space units were consolidated from 155 to 115 (−25.81%), well below the ≤40 unit target, eliminating redundant transitional spaces. The four main functional units were networked, resolving landscape and functional fragmentation. The comprehensive site analysis demonstrates the systematic approach to achieving spatial integration (Figure 21). The master plan (Figure 21a) illustrates the strategic arrangement of nine key functional zones, from the West Entrance Landscape (01) to the Quiet Rest Space (09), creating a cohesive network of interconnected spaces. The site analysis diagrams (Figure 21b) reveal the multi-layered design strategy through vegetation analysis, functional zoning, and streamline optimization, where main circulation routes (solid red lines) and secondary pathways (dotted red lines) form an integrated movement network.

The design process (Figure 21c) addresses scale-related fragmentation issues through terrain undulation strategies combined with strategic greenery placement, establishing appropriate landscape enclosure relationships. The spatial analysis (Figure 21d) showcases how the undulating terrain responds to prevailing wind patterns while the central square serves as a highly accessible hub. This integrated approach transforms the previously fragmented landscape into a resilient, adaptable system capable of accommodating diverse activities while maintaining spatial coherence during extreme weather conditions.

In summary, the optimized site achieved measurable improvements across all three PBUD performance dimensions, with all key indicators meeting or exceeding preset targets. These findings confirm the methodological rigor and practical effectiveness of the proposed framework, offering a quantifiable and verifiable reference for the sustainable renewal of aging parks in high-density urban cores.

5. Discussion

5.1. The Value of the Environmental Analysis Overlay Method Under the PBUD Framework

This study’s primary contribution lies in operationalizing PBUD theory through a novel analytical integration that addresses three fundamental gaps in existing park renewal methodologies. First, the Performance-Technology Synergy establishes a systematic correspondence between abstract PBUD dimensions and quantitative spatial analysis tools. Unlike previous studies that apply space syntax, CFD, or solar analysis in isolation, this method creates a unified analytical framework where spatial accessibility (DepthmapX), microclimate regulation (CFD-Phoenics), and environmental comfort (solar analysis) are coordinated through consistent grid-based overlay techniques. This integration enables the identification of performance conflicts—such as ventilation corridors that compromise pedestrian accessibility—that remain invisible to single-dimension approaches. Second, the Quantitative Operational Pathway introduces a replicable closed-loop methodology that transforms qualitative design intuitions into measurable performance criteria. The threshold classification system (Table 3) provides explicit quantitative targets (e.g., comfortable wind speed 0.5–2 m/s, optimal RA ≤ 0.04, resilience units ≤ 40), while the 35 cm × 40 cm grid overlay enables precise spatial mapping of performance conflicts and opportunities. This systematic approach moves beyond case-specific solutions to establish a transferable analytical protocol applicable across diverse urban park contexts. Third, the method addresses the Climate-Health-Resilience nexus through integrated performance assessment rather than sequential optimization. The overlay analysis reveals compound vulnerabilities—such as areas experiencing simultaneous thermal discomfort, accessibility barriers, and spatial fragmentation—that compromise park functionality under both routine and extreme conditions. This holistic approach enables targeted interventions that achieve multiple performance improvements simultaneously, as demonstrated by the Wuxi case study’s coordinated gains across all three PBUD dimensions.

1. Performance–Technology Synergy: Translating PBUD Theory into Spatial Practice

The method tightly couples the three-dimensional PBUD goals (climate, health, resilience) with spatial syntax (DepthmapX), CFD wind simulation (Phoenics), and sunlight analysis (Suric Sun), overcoming the limitations of single-technology approaches that tend to prioritize simulation over integrated design [79]. For example, by overlaying CFD wind fields with space syntax-based pedestrian heatmaps, conflicts between ventilation corridors and high-traffic areas were avoided, while simultaneously achieving “spatial justice” (restoring poorly connected zones), “ecological continuity” (microclimate regulation), and “functional compositeness” (flexible multi-use spaces). After optimization, the number of spatial units decreased from 155 to 115, RA mean decreased by 1.8%, confirming the alignment of PBUD performance with the core propositions of performance-oriented planning theory [16] and demonstrating the translation of abstract theory into perceptible spatial improvements.

2. Quantitative Operational Pathway: A Closed-Loop Tool of Threshold–Overlay–Feedback

The method introduces a unified 35 cm × 40 cm grid across multiple data layers (pedestrian flow, wind speed, sunlight, RA, visibility) to implement a closed-loop workflow: performance threshold classification → spatial overlay analysis → strategy generation → effectiveness validation. Thresholds explicitly define quantitative targets, e.g., comfortable wind speed 0.5–2 m/s, healthy RA 0.12–0.15, resilience units ≤ 40; overlay analysis visualizes “heat island conflict zones,” “microclimate coordination zones,” and “priority intervention zones”; feedback validation uses the same simulation tools (e.g., DepthmapX, Phoenics) to confirm results, such as R² for HH-R3 vs. HH increasing from 0.39 to 0.64. This establishes a replicable, data-driven operational tool for rapidly guiding urban park renewal in aging high-density cores.

3. Addressing Core “Climatic–Health–Resilience” Needs

At the climatic dimension, CFD and sunlight overlays enable bioclimatic micro-applications, such as summer southeast wind guidance and winter sunlight provision. For health performance, spatial syntax and agent-based flow simulation ensure accessibility, crowd safety, and fair use. Resilience performance is supported through multi-layer overlays that inform extreme-weather adaptation measures, such as movable shading for high temperatures and emergency evacuation corridors during heavy rain. In the Wuxi case, indicators including reduced heat island intensity, lower peak crowd density, and higher usability under extreme heat demonstrate the method’s alignment with thematic objectives, offering a practical methodological reference for similar urban park renewal studies.

5.2. A Framework for PBUD Renewal Pathway for High-Density Urban Aging Parks

Based on PBUD performance improvements observed in the Wuxi case reflecting a practical translation of urban design theory, three scalable urban renewal strategies emerge [80]. First, climate–space synergy is achieved through a “ventilation corridor + solar adaptation” logic, preserving 3–5 m wide green wind corridors along the prevailing summer southeast wind to channel airflow (0.5–1.5 m/s) while planting green belts in stronger wind zones (2–4 m/s), and elevating terrain (1.2–2.5 m) to mitigate shading from the 266 m supertall building, ensuring winter solstice sunlight of 4–6 h and reducing <3 h areas from 26% to 19%, with heat conflict zones converted to ecological green belts. Second, health–accessibility is optimized via a three-tiered path hierarchy: primary paths (5–8 m) in high-integration areas connect the west metro entrance to central activity zones, secondary paths (3 m) link peripheral weakly connected areas, and tertiary paths (1–2 m) integrate with green belts and embedded seating to enhance social accessibility, dispersing crowds and supporting intergenerational interaction. Third, resilience–elasticity is promoted through spatial nesting and temporal programming, placing flexible activity zones (skate parks, extreme sports) in winter stagnation areas, movable shading in sunny zones to increase extreme heat usability, and configuring a nested structure of one main and multiple subspaces to support high- and low-intensity functions, temporal activity shifts, and age-friendly intergenerational use, fully realizing the goals of ventilation accessibility, temporal flexibility, and life-cycle continuity.

Underpinning these strategies, vegetation design serves as critical performance infrastructure [81]. The transformation from scattered single-layer planting to a functionally differentiated three-tier system directly enabled measured improvements (Figure 22). Pre-renovation monotonous tree arrangements lacked spatial hierarchy and microclimate regulation capacity (Figure 22a). Post-renovation design established: (1) wind corridor vegetation with 8–10 m tree spacing to guide airflow while permitting solar penetration; (2) multi-layered buffer zones (trees + shrubs + groundcover) moderating wind speeds along commercial streets; (3) thermal mitigation clusters with >60% canopy coverage in southwest zones. This systematic vegetation hierarchy created spatial depth and visual openness (Figure 22b). The emphasis on deciduous species (70% of tree palette) enables seasonal adaptability—summer shade combined with winter solar access—demonstrating that systematic planting constitutes measurable performance intervention in high-density park renewal.

5.3. Research Limitations and Future Directions

The existing urban park renewal performance evaluation framework still has room for development. This is particularly evident in the challenge of capturing the full spontaneity of human behavior. While our study employs agent-based modeling and heatmap validation to approximate pedestrian behavior [82], we acknowledge that human spontaneity and social appropriation of space—as exemplified by the Pruitt-Igoe complex—cannot be fully captured by quantitative models alone. Future research should integrate ethnographic observations and participatory design methods [83] to better capture the nuanced and evolving nature of lived urban experience, complementing the computational analysis presented here. Thus, the PBUD framework serves as a performance baseline rather than a deterministic prescription, allowing for adaptive and human-centered urban park design. Finally, it is important to note that this study presents a simulation-based conceptual design optimization. Its effectiveness was validated through an internal closed-loop ‘diagnosis-optimization-validation’ technical workflow but has not yet been empirically tested within a physically constructed park project. This remains a key limitation of the current study and highlights an important direction for future research.

First, in terms of climatic adaptation, most studies focus on subtropical monsoon climates (e.g., Wuxi), while validation and optimization are needed for diverse climatic conditions, such as temperate monsoon (Beijing), plateau monsoon (Kunming), and tropical (Guangzhou). Regional adjustments of comfort wind speed, sunlight duration, and shading/drainage strategies can help establish a climate-adaptive PBUD performance threshold guide.

Second, health performance quantification can be enhanced by integrating physiological data (e.g., heart rate, body temperature) collected via wearable devices with subjective comfort surveys, constructing a model linking objective indicators with human health responses. Third, the dynamic simulation capability of technical models should be strengthened by incorporating AI and machine learning, using real-time meteorological and pedestrian flow data to enable adaptive management under extreme weather and peak crowd conditions. Recent advances in machine learning applications for urban planning challenges [84] demonstrate significant potential for complex problem-solving based on experimental data in urban applications, including configurations of the built environment and architectural design.

The pathway from planning support to plan-making through AI integration [85] offers promising directions for enhancing the dynamic capabilities of performance-based urban design frameworks. Finally, PBUD core indicators (e.g., green accessibility, coverage of ventilation corridors) can be translated into urban renewal policy tools, forming a “theory–method–policy” closed loop to provide actionable and verifiable guidance for the sustainable renewal of old parks in high-density urban cores.

Despite these acknowledged limitations, the Wuxi case study provides empirical evidence supporting the potential of the Environmental Analysis Overlay Method for aging park renewal. The quantitative performance improvements are consolidated in

Table 4. Quantitative comparison of pre- and post-renovation performance across climate, health, and resilience dimensions.

Performance Dimension	Key Indicator	Pre-Renovation	Post-Renovation	Change
Climate	Comfortable wind zones (0.5–1 m/s)	15%	>30%	+100%
	Problematic wind areas	30%	11%	−63%
Health	RA (Relative Asymmetry)	0.0921	0.0904	−1.8%
	Integration correlation (R2)	0.39	0.64	+64%
	Effective pedestrian capacity	3038 m2 (15.6%)	3798 m2 (19.5%)	+25%
Resilience	Number of spatial units	155	115	−26%
	Activity areas with <3 h winter sunlight	26%	19%	−27%

to summarize the observed outcomes. While these results are based on a single case and should be interpreted within the methodological constraints discussed above, the measurable gains across climate, health, and resilience dimensions suggest that the approach offers a promising framework for practitioners and researchers working in comparable high-density contexts. Further validation across diverse climatic regions and park typologies would strengthen confidence in the method’s broader applicability.

6. Conclusions

This study addressed the multifaceted challenges of renewing aging parks in high-density urban cores, where issues of climate maladaptation, public health risks, and resilience deficiencies are acutely concentrated. Using Chengzhong Park in Wuxi as a case study, this research implemented and validated a performance-based urban design (PBUD) framework through a novel “Environmental Analysis Overlay Method.” This method successfully integrated multi-source data and simulation tools—including space syntax (DepthmapX), computational fluid dynamics (CFD-Phoenics), and solar radiation analysis (Suric Sun)—within a unified spatial grid to diagnose, optimize, and quantitatively validate park renewal strategies.

The core achievement of this study lies in translating the three-dimensional PBUD objectives—Climatic, Health, and Resilience Performance—into a replicable, closed-loop technical workflow of diagnosis, optimization, and validation. The post-renewal simulation results confirm significant improvements across all targeted dimensions:

• Climatic Performance: The proportion of comfortable wind zones (0.5–1 m/s) increased dramatically from 15% to over 30%, while stagnant wind areas were reduced from 30% to 11% (a reduction of 29 percentage points). Problematic wind areas (static zones and excessive wind speeds > 4 m/s) were reduced from 30% to 11% of the park.
• Health Performance: Spatial accessibility was greatly improved, with the mean Relative Asymmetry (RA) decreasing from 0.0921 to 0.0904 (−1.8%) and the correlation (R²) between global and local integration rising from 0.39 to 0.64 (+64.1%). The effective area for pedestrian gathering and circulation increased by 25%. Winter sunlight exposure for activity zones was significantly increased, reducing areas receiving less than 3 h of sun on the winter solstice from 26% to 19%, while 81% of activity areas now receive 5–8 h of sunlight.
• Resilience Performance: Spatial fragmentation was addressed by consolidating the number of spatial units from 155 to 115 (−25.81%), significantly exceeding the target reduction. The usability rate during extreme heat was dramatically improved, and the redesigned pathway network now doubles as an efficient emergency evacuation route.

The findings demonstrate that the “Environmental Analysis Overlay Method” provides an effective, quantifiable pathway for the sustainable renewal of aging urban parks. It moves beyond singular, isolated analyses to offer a holistic tool that synchronizes environmental performance with socio-spatial functionality. By establishing clear performance thresholds and a grid-based overlay system, the method enables designers and planners to identify conflict zones, prioritize interventions, and generate targeted strategies that are both scientifically grounded and practically actionable.

Notably, this study has certain limitations. First, the validation is based on a single case study; future research should apply this method across diverse geographic and climatic contexts to test its broader applicability and refine performance thresholds. Second, the CFD simulations, while rigorous, would benefit from validation with on-site sensor data to enhance accuracy. Third, the assessment of health and social performance primarily relied on quantitative spatial and flow data; integrating qualitative user feedback and physiological data (e.g., from wearable devices) would provide a more comprehensive understanding of perceived comfort and usability.

Future work should focus on expanding the case base, incorporating real-time data streams and AI-driven adaptive management for dynamic performance optimization, and translating the PBUD indicators into standardized policy tools [86]. This will further bridge the gap between theoretical research, design practice, and urban governance, ultimately supporting the development of more climate-adaptive, healthy, and resilient cities.