
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
Elsevier

S2405-8440(24)12935-0
10.1016/j.heliyon.2024.e36904
e36904
Research Article
Resilience to climate change by improving air circulation efficiency and pollutant dispersion in cities: A 3D-UFO approach to urban block design
Makvandi Mehdi abc1
Yuan Philip F. philipyuan007@tongji.edu.cn
a1⁎
Ji Qunfeng b1
Li Chuancheng b1
Elsadek Mohamed a
Li Wenjing ad
Hassan Ahmad a
Li Yu a
a College of Architecture and Urban Planning, Tongji University, Shanghai, China
b College of Civil Engineering and Architecture, Wuhan University of Technology, Wuhan, China
c College of Architecture and Urban Planning, Huazhong University of Science and Technology, Wuhan, China
d Key Laboratory of Ecology and Energy-saving Study of Dense Habitat, Ministry of Education, Shanghai, China
⁎ Corresponding author. College of Architecture and Urban Planning, Tongji University, 1239 Siping Road, Shanghai 200092, China. philipyuan007@tongji.edu.cn
1 These authors contributed equally to this work.

30 8 2024
15 9 2024
30 8 2024
10 17 e3690416 4 2024
22 8 2024
23 8 2024
© 2024 The Author(s)
2024
https://creativecommons.org/licenses/by/4.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Urbanization presents significant challenges to air quality and climate resilience, necessitating pioneering urban design solutions to enhance air circulation and mitigate pollutants. This urgency intensifies in densely populated and rapidly evolving regions like Wuhan, China, where effective strategies are crucial for sustainable development. This study introduces an innovative 3D Urban Form Optimization (3D-UFO) methodology aimed at advancing urban block design configurations to improve urbanization quality. The 3D-UFO approach systematically addresses the multifaceted challenges of climate change and air quality degradation in rapidly urbanizing areas. Integrating GIS-based analysis for comprehensive Land-Use and Land-Cover Change (LULCC) evaluation with Computational Fluid Dynamics (CFD), our approach employs systematic exploration guided by established urban airflow study protocols. Robust metrics—Airspeed-Ratio (ASR) and Average-Age-of-Local-Air (ALA)—quantify the impact of diverse urban block design strategies on air-circulation efficiency and pollutant dispersion. Analysis across various urban scenarios, yielded by the proposed 3D-UFO approach, reveal significant variations in air-circulation efficiency at street and building levels (SBLs). Optimal urban air circulation achieves efficiency levels of 50–70 % when airflow aligns orthogonally across and parallel to streets. Adjusting street-level building heights, especially incorporating taller structures, boosts ventilation efficiency by 20–30 %, which is crucial for improving airflow dynamics in urban settings. Higher Height-to-Width (H/W) ratios (>5.5) yield a 218.5 % increase in ventilation in specific urban layouts. Notably, the synergy of street-aspect-ratio and building-height-ratio adjustments significantly enhance ASR and ALA, providing a quantitative foundation for sustainable urban development. This 3D-UFO methodology, fusing LULCC analysis, CFD simulations, and systematic exploration, emerge as a valuable framework for urban planners and designers. The study offers informed insights into urban sustainability challenges, demonstrating advancements in addressing environmental concerns and improving living conditions within densely populated environments.

Keywords

Urbanization
Land-use and land-cover change (LULCC)
Air circulation efficiency
CFD
3D urban form optimization (3D-UFO)
==== Body
pmc1 Introduction

Urbanization, driven by population growth and land use changes, profoundly affects socio-economic development and local climate dynamics, affecting wind patterns, air quality, and human comfort [1,2]. In China, this phenomenon exhibited a notable increase from 17.9 % in 1978 to 59.6 % in 2018, necessitating sustainability efforts to mitigate its impact on land use, urban structure, and the local temperature regimes [[3], [4], [5]].

In rapidly urbanizing cities like Wuhan, with an annual development rate of 1.12 %, challenges emerge due to prioritizing expansion over comprehensive planning [6]. This leads to densely populated regions facing air quality and heat-related challenges, exacerbated by imbalances in urbanization speed and quality, despite recent government regulations (Fig. 1e) [[7], [8], [9]]. In urban environments, building blocks' configuration and morphology are key factors for shaping thermal dynamics and mitigating challenges related to land-use and land-cover changes, including heat stress, airflow disruptions, and pollutant concentrations [[10], [11], [12]].Fig. 1 Wuhan's urban development overview: (a) Built-up areas highlighted in red; (b) Urban view from a residential peak; (c) Early urban development and (d) later urban development phases (south bird-eye-view); (e) Evolution of typical building forms under urbanization regulations; (f) Samples of urban blocks across fifteen typology categories.

Fig. 1

The objective of this study is to address these urbanization challenges by optimizing urban blocks for better urbanization quality using an innovative 3D Urban Form Optimization (3D-UFO) approach. Validated computational fluid dynamics (CFD) simulations to investigate and improve air circulation in densely built-up areas like Wuhan enrich this approach.

CFD serves as a cost-effective method for assessing urban forms' influence on natural ventilation [13,14]. Nonetheless, it faces limitations, particularly in turbulence modeling accuracy under real-world conditions [15]. Fig. S1 highlights limitations in existing research methods, from model acquisition and generation to CFD simulation. Analysts have scrutinized different turbulence models based on Reynolds-averaged Navier-Stokes (RANS) for accuracy, with each addressing specific ventilation challenges [[16], [17], [18], [19], [20], [21]].

To enhance simulation precision, researchers have systematically studied factors such as computational domain size, grid resolution, atmospheric boundary layer profiles, discretization scheme order, and convergence criteria [22]. The standard k-ε model has demonstrated superior ventilation performance in generic building scenarios, aligning with experimental observations [23]. Moreover, various modeling methods have been proposed for CFD simulations in urban environments [[24], [25], [26], [27]].

In urban ventilation studies, objectives can be categorized based on (1) the creation of 3D forms and (2) their complexities. The first category primarily uses simple forms, such as solitary cubic-shaped blocks [28,29], and urban canyon geometries [30,31] to examine building aspect ratio [32], street aspect ratio (H/W) [33], building density (BD) [34], building height variations (BHV) [35] and their impact on airflow and pollutant dispersion [36,37]. The second category investigates realistic 3D urban spaces for more sophisticated evaluations of outdoor/indoor ventilation within diverse urban contexts, including real compact heterogeneous areas [38], multi-story structures [39], canyons and pedestrian levels [40].

As evident in the literature, previous studies on urban ventilation conditions and thermal comfort in model-based urban configurations [41] have often overlooked critical factors like design ideologies, local adjustments, and socio-economic influences within urban blocks. The core objective of this study is to bridge the gap between the predominant simplistic cuboidal geometries found in urban design literature and the complex dynamics present in real-world urban environments, utilizing the 3D-UFO approach across diverse site scenarios. This holistic approach provides a comprehensive examination and evaluation of air quality concerns in rapidly evolving urban areas, considering historical development trends, regulatory frameworks, urban design parameters, idealized urban configurations, site-specific features, and ventilation performance assessment. A key study feature is incorporating block design scenarios, highlighting CFD models' pivotal role in environmental dispersion research.

Through a detailed analysis of urban development, building block diversity, and critical ventilation metrics like Average-Age-of-Local-Air (ALA) and Airspeed Ratio (ASR), our approach addresses disparities, providing a comprehensive understanding of air quality concerns and contributing to optimizing urban air circulation efficiency.

The proposed 3D-UFO approach investigates urban air quality in Wuhan's evolving urban areas through four facets: (1) the city's historical growth, urban planning, and building regulations; (2) extract the vital urban design parameters; (3) creation of idealized urban forms characterizing optimal land distribution within a generic street network; and (4) the ventilation performance of these forms using validated CFD simulations. This approach presents an effective 3D urban design solution, providing valuable insights for city stakeholders to implement.

2 3D-UFO creation for urban block evolution

Antecedent to CFD analysis, we extensively applied remote sensing and Geographic Information System (GIS) techniques to investigate the region's land-use and land-cover changes (LULCCs). This encompassed an analysis of the historical development of urban building blocks across diverse sites from 1980 to 2015. After analyzing hundreds of urban blocks, we developed practical 3D urban design frameworks considering (1) various physical characteristics (e.g., H/W, BHV, and BD), site-configuration, and block types, (2) urban design principles, (3) building block regulations, and (4) developers' profit-driven motivations.

Subsequently, we created and assessed groups of 3D urban forms using well established CFD models and validated assessment indices like mean air age and air change rate. Our innovative 3D-UFO approach enables the generation of urban block configurations compliant with planning regulations while enhancing ventilation efficiency, evaluated through the ASR and ALA indices at street and building levels (SBLs).

2.1 Wuhan's urban development

Wuhan in central China (113°41′-115°05′E, 29°58′-31°22′N) covers 8569.15 km2, featuring plains amid mountains [42]. Urban land, once less than 9 % (759.5 km2) in 1980, expanded to nearly 36 %, driven by rapid population growth from 2.52 million to 11.08 million over 40 years, causing a 405 % increase in urban areas and decreasing water bodies to 1.17 % (Fig. 1a) [42,43].

This shift impacted the wind environment, cooling demands, and land surface temperatures, intensifying the urban heat island effect. To accommodate the population surge, Wuhan's urban planning regulations evolved, leading to diverse urban block generations and building forms as shown in Fig. 1e. Wuhan's building and planning regulations underwent revisions to accommodate this population growth, leading to changes in urban blocks and building forms, as illustrated in Fig. 1e and f, respectively. During each timeframe, developers maximized built spaces within the confines of regulatory provisions and endeavored to select building forms aligned with purchaser preferences.

This transformed Wuhan into a partially high-density, high-rise urban area [44], as revealed in Fig. 1b. Current regulations permit total site coverage for buildings under 20 m and impose plot ratio limits from 8 to 10. This led to cake-like low-rise, mid-rise, and tall buildings, progressively evolving Wuhan's urban form. In Fig. 1c, the Jiangxia area in Wuhan is still in the initial transformation phase, while Fig. 1d depicts a neighborhood in an advanced development stage. Over the past decade, many sites initially undergoing these transformations have been developed as high-rise structures. Wuhan is expected to ultimately become a total high-density, high-rise urban setting under current regulations, necessitating optimization of 3D urban forms through the 3D-UFO approach proposed here.

2.2 Critical variables in realizing 3D urban Form optimization (3D-UFO)

In the quest for optimal urban development and environmental quality, certain morphological variables emerge as critical determinants. Urban road networks, site types, and building forms are pivotal elements in realizing 3D-UFO. Urban road networks hold a prominent role in urban design and overall cityscape configuration [45,46]. Extensive road network analysis in Wuhan reveals a duality in its urban fabric. The modern sector boasts a European-inspired road network featuring diverse layouts, including gridiron, loose grid, organic, radial grid, and suburban systems (Fig. 2a) [7,47,48]. Conversely, old-established areas predominantly feature an oblong road network, with street blocks exhibiting variations in width (30–50 m) and length (80–160 m), interspersed with road widths spanning 8–20 m (Fig. 2b). The presence of service roads further delineates distinctive urban forms in these areas. In the modern sector, while a comparable road network system is implemented, service routes are omitted due to the prevalence of high-rise building development.Fig. 2 (a) Comparison of common and less common road network systems, delineating distinctions in layout and connectivity; (b) Typical road network system for residential areas within the Wuhan metropolitan area, highlighting street patterns and urban design; (c) Site variations in Wuhan influenced by Development Control Regulations (DCR), showing changes in building layouts and planning approaches.; (d) Detailed process flow of a CFD simulation, outlining the steps from model setup to result analysis; (e) Schematic diagram of the proposed framework, depicting the sequential steps used in the study.

Fig. 2

Over a meticulous 35-year investigation into residential growth in Wuhan, we systematically classified sites into five groups, strictly adhering to Development Control Regulations (DCR) (Fig. 2c) (https://www.soujianzhu.cn/NormAndRules). These site classifications, comprising A, B, C, D, and E types, form the bedrock of 3D-UFO. Type-A sites are expansive areas enclosed by four roads wider than 3.5 m, boasting plot ratios ranging from 7.6 to 9.7. Type-B sites are adjacent to three roads wider than 3.5 m, while Type-C and Type-D sites each adjoin two such roads, with average plot ratios of 10.5, 8.4, and 9.4, respectively. Type-E sites are areas bordering one road wider than 3.5 m, with an average plot ratio of 8.4. According to Wuhan's regulations, plot ratios can increase by 0.1 times with the integration of specific green features into building design. Our investigation of residential growth unveiled a profound interplay between building parameters (H/W, BD, BHV) and site forms, significantly impacting air circulation efficiency and thermal environments. This intricate relationship is encapsulated within the framework of 'environmental optimization principles (EOP)' for 3D-UFO. In the densely populated urban context of Wuhan, each site category's outlook toward adjoining roads significantly impacts building characteristics' value. Developers and planners seek optimized designs aimed at enhancing environmental conditions. Fig. 2e concisely illustrates the site optimization procedure, encompassing urban area identification to 3D-UFO realization, complemented by Fig. 2d, which provides a comprehensive overview of the entire CFD process flow employed in the study's investigations.

To advance our understanding, Type-A sites are conducive to high-rise buildings, characterized by structures surrounding the core and a larger surface area. Building varieties (BV) such as cube (BV1), linear (BV2), courtyard (BV3), perimeter (BV4), slab (BV5), centralized (BV6), hybrid (BV7), tall building/tower (BV8), circular (BV9), and lozenge (BV10) are predominantly adopted within this category. In Type-B sites, the U-formed layout (BV11) is frequently use, connecting three adjacent streets and offering expansive vistas and open spaces. Type-C sites, featuring two parallel adjacent roads, often utilize an H-formed layout (BV13). Typically, Type-D sites employ L-formed structures (BV12) along the road, while Type-E sites along a single road prioritize configurations that optimize the width across the street for enhanced visibility and utilization of open spaces. Y-formed (BV14) and T-formed (BV15) configurations are also favored for Type-D and Type-E sites, respectively. These aspects, coupled with the scenarios delineated in Section 2.3, constitute the fundamental framework of the study's 3D-UFO approach, furthering our understanding of Wuhan's densely populated urban environment.

2.3 Developing design strategies for 3D-UFO and modeling

In our pursuit of effective 3D Urban Form Optimization (3D-UFO) and the development of design strategies, we meticulously crafted six distinct scenarios that align seamlessly with current regulatory frameworks governing road networks, land parcel delineations, and socio-economic considerations under urbanization process. Our paramount objective has been to formulate design strategies for 3D-UFO and modeling, which encompass the following key aspects.i. Functional zoning: To commence, we subdivided a typical urban street area into four distinct block sizes—namely, G14000 (4000 m2), G22000 (2000 m2), G31000 (1000 m2), and G4500 (500 m2). 'G' identifies the group categories. Within this framework, we executed a dual zoning strategy, comprising two distinct approaches: (1) on-site development without service paths and (2) block partitioning with the inclusion of a service path to create smaller categorized sites. For instance, Site A (G14000), denoted as G14000A, encompassing a total area of 4000 m2, underwent subdivision through the introduction of a horizontal or vertical service lines, yielding distinct Aa or Ab blocks designated as G22000Aa and G22000Ab, respectively. Consequently, this transformation gave rise to additional cases—namely G31000a, G31000b, and G4500. Following Wuhan's Building and Planning Regulations (BPR), which adhere to the Development Control Regulations (DCR), service paths were designed to facilitate both ingress and departure, aligning with Site A's layout. This meticulous effort resulted in the creation of six distinct functional zoning scenarios, as depicted in Fig. 3a.Fig. 3 (a) Analyzing block scenarios and site divisions with common building types in 3D-UFO approach and (b) Specifications of study cases, detailing relevant parameters and configurations.

Fig. 3

ii. Plot ratios: Within the framework of Wuhan's DCR, we meticulously imposed plot ratio limits for building construction and planning, with a deliberate focus on their impact on the surrounding environment. Our analytical approach drew upon Height-to-Width ratios (H/W) and Building Height Variation (BHV), yielding nuanced variations in plot ratios tailored to the specific site types. See Fig. 3b for detailed scenario descriptions. It is imperative to note that interactions between building forms and block arrangements yielded minor adjustments in plot ratios and other pertinent building parameters.

iii. Building Form in Block Configuration Scenarios: we meticulously selected building varieties (BV1 to BV15) with a primary emphasis on building forms. This precise selection process deepens our comprehension of the influence of building shape on urban design. Small G4500 blocks, encompassing D and E sites, pertain to buildings with diminutive footprints/floor areas like BV12, BV14, and BV15, while average-scale blocks (G22000Aa, G31000a, and G31000b) correlate with buildings featuring intermediate footprints/plans, exemplified by BV11 and BV13. Moreover, BV10, BV9, BV2, and BV1 may also fall into the medium category, provided they exhibit mid-density attributes (i.e. H/W = 1/1 and BD = 0.3). Spacious blocks like G14000A, as well as average blocks with mid-density and low aspect ratios like G22000Ab, are associated with larger-scale building projects, exemplified by BV8, BV7, BV6, BV5, BV4, and BV3. Furthermore, extensive site plan adjustments and building parameter modifications allow for the analysis of BV10, BV9, BV2, and BV1 within this category.

These scenarios encompass diverse building categories, encompassing low-rise (LRB; ≤3 floors or ≤10 m in height), mid-rise (MRB; 3< floors <7, 10 m < height <24 m), and high-rise buildings (HRB; >24 m in height or residential buildings exceeding 27 m) based on building form preferences. These scenarios, or models, underwent horizontal and vertical division by 20 m roads, as depicted in Fig. 4a. Notably, following Wuhan's Building and Planning Regulations (BPR) for sites smaller than 5000 m2, all tall buildings designed with five-story platforms. This approach, predominantly implemented in Wuhan, accommodates diverse spaces such as malls, supermarkets, and other amenities. A typical high-rise building in our study context comprises 33 stories, corresponding to a height of 99 m (Fig. 4b). This height aligns with the average of the 386 high-rise buildings meticulously analyzed within the Wuhan urban landscape. Overall, our approach to creating 3D-UFO design strategies is methodical and rigorous. It encompasses functional zoning, plot ratios, and diverse building forms, culminating in 15 distinct scenarios towards the study objectives in densely populated Wuhan, aligning with regulatory and socio-economic factors.Fig. 4 Urban design scenarios: (a) Models created within the 3D-UFO framework, showcasing different urban design configurations and (b) Diverse block configurations designed for the study.

Fig. 4

3 Methodologies

Long-term analysis of Land-use and Land-cover Change (LULCC), in conjunction with meteorological data, conducted to investigate modifications in urban morphologies and climatic conditions. Computational Fluid Dynamics (CFD), utilizing ANSYS FLUENT 17.1, employed to simulate and assess airflow within diverse building scenarios within the framework of the 3D Urban Form Optimization (3D-UFO) approach, with the objective of optimizing urban block-scale ventilation. The analysis adhered to established urban airflow study guidelines [[49], [50], [51], [52]], ensuring scientific rigor and replicability. This comprehensive approach forms the basis for a systematic exploration of evolving urban environments, climatic dynamics, and urban design responses over time.

3.1 K-ω SST model and simulation domain

The k-ω SST (shear-stress transport) model, a hybrid of the k-ε and k-ω models introduced by Menter [53], was employed in this study. Specifically, the k-ω model was utilized for near-wall boundary layers, while the k-ε model was reserved for wake zones located away from the wall boundary. This model has demonstrated a remarkable capacity to accurately predict airflow patterns and wind speed distribution around the buildings, as evidenced by Zhong, Jing [54]. Consequently, it was employed for investigating vortex shedding around buildings and the eddy flow through openings.

The computational analysis simulation domain adhered to established guidelines, depicted in Fig. 5a. It maintained a blockage ratio of less than 3 %, a dimension determined through wind tunnel experiments [49,55]. The dimensions of the top, lateral, and inlet boundaries for each domain model were adjusted to 5 Hmax based on the target building's height (Hmax) [50,56]. In this context, Hmax varied, with values of 13, 31, and 114 m for low-rise (LRB), mid-rise (MRB), and high-rise (HRB) buildings, respectively. Consequently, three distinct domain sizes were established: LRB = 360 × 230 × 78, MRB = 720 × 410 × 186, and HRB = 2380 × 1240 × 684 m³.Fig. 5 (a) Arrangement of computational domain including top and side views, mesh resolution, and surface mesh details; (b) Inlet boundary conditions showing profiles of mean wind speed and turbulent intensity; (c) Uniform source distribution at SBLs.

Fig. 5

A thorough analysis involved evaluating airflow from three directions: North (N, 360°), North-northeast (NNE, 22.5°), and Northeast (NE, 45°). To account for a wide range of inbound wind conditions, 32 inbound wind directions were computed, each differing by an increment of 11.25°. This approach aligned with Wuhan's recommended air ventilation guidelines (http://hz.hjhj-e.com/home/meteorologicalData/windRose). The cases were configured to accommodate an asymmetrical design based on the prevailing wind direction, determined as NNE.

3.2 Establishing boundary layers

This study assessed wind availability in Wuhan using the annual mean wind speed of 2.8 m/s recorded at Jiangxia's weather station as a reference. Adhering to the guidelines set forth by the Chinese wind code [57,58], we considered the wind velocity distribution of terrains B, F, and G, as delineated by Touma [59].

Profiles of inflow streamwise wind speed (U0 (z) = U0 (z/H)0.30) and turbulence intensity (V(z) = 0.08227 × z−0.206) were established based on power laws derived from experimental wind tunnel data [54].

Conforming to the AIJ guideline [49], we characterized the specific dissipation rate ω (z), the turbulent dissipation rate ε (z), and the vertical distributions of turbulent kinetic energy k (z) as follows:(1) kz=σu2z+σv2z+σw2z2≅σu2z=VzU0z2

(2) E(z)≅Pk(z)≅−U0′w′dU0(z)dz≅Cμ12k(z)dU0(z)dz

(3) ω(z)=E(z)k(z)Cμ

Here, σu, σv, and σw represent the root mean square (RMS) values of velocity fluctuations in the x, y, and z directions, respectively. In the k equation, the production term is denoted as Pk, with Cμ (=0.09) serving as the model constant. Applying Eq. (1) and Eq. (3) allowed us to derive inlet turbulent boundary conditions for k (z) and ω (z) in accordance with the SST k-ω turbulence model.

No static pressure was imposed for the outlet boundary conditions (=0). Notably, we incorporated improved wall/surface boundary conditions, as provided by ANSYS' ICEM, to the ground, top, and building walls within the study domain [60]. Fig. 5b demonstrates the inflowing profiles of mean wind speed and turbulence intensity.

3.3 ALA and ASR for assessing ventilation efficiency

The Average-Age-of-Local-Air (ALA) and Airspeed Ratio (ASR) were employed as prospective metrics to assess urban ventilation efficiency. ASR, represented as αASR in Eq. (4), quantifies pedestrian-level airspeed accessibility and was calculated following Wuhan's guidelines [61].(4) αASR=Wairspeedat1.5mWrInitialairspeed

W, the pedestrian-level airspeed, was determined at 1.5 m above ground in an urban environment. The initial airspeed, Wr, representing the reference wind velocity at the boundary layer's upper limit, customarily 350–650 m above the city's central area [54]. Here, we employed a Wr = 5.47 m/s at z = 500 m above the city center, as calculated by Eq. (5).(5) U0(z)=2.8(z90)0.30

The ALA, a pivotal parameter, reflects the concentration of airborne pollutants across various segments of a building. We rigorously calculated ALA by applying Etheridge's robust uniform emission theory [62], considering emissions from two distinct sources: Nitrogen Dioxide (NO2), predominant vehicular-derived pollutant at street-level, and H2O molecular diffusion at building-level, spanning the region from street-level to beneath the urban canopy. To estimate ALA precisely, we employed the following equation:(6) ⍺‾ALA=P‾(time)Pe(kg/m3×S)

Where α‾ALA represents the air turnover time, reflecting its age, P‾ signifies the pollutant concentration over time, and Pe denotes the pollutant emission volume size, set at 10−5(kg/m3×S) [63]. Utilizing Eq. (6), we calculate air turnover time within the urban canyon, including both street and building levels (SBLS). Fig. 5c illustrates the volume space of the SBLs.

3.4 Computational grid resolution and independence testing

In this study, we utilized a standardized mesh with optimized resolution to fine-tune our models, as presented in Fig. 6a, Fig. 6ba, providing an isometric view of the models with their surface meshes. The smallest surface mesh dimension, at 0.3 m, accurately captured the geometric intricacies of the building's ground, corners, and top grid. Grids located farther from the models used larger mesh sizes. Employing an extension ratio of 1.1, as proposed by Tominaga, Mochida [49], allowed us to effectively adapt grid sizes for adjacent cellular units.Fig. 6a Mesh Quality and Quantity Analysis (a) Isometric depiction of domain meshes across models, color-coded to indicate quality levels from low (<0) to high (>1). Green highlights regions exceeding the 0.3 quality threshold, indicating superior segments.

Fig. 6a

With the 3D-UFO approach, we established eighteen distinct models to examine various urban building forms and site scenarios thoroughly. Fig. 6b details the total tetrahedral grid elements for all models, with variations in cell size corresponding to the diverse geometrical features of each model. Grid independence was assessed by conducting grid convergence tests following the methodology outlined by Celik et al. [64].Fig. 6b Total number of tetrahedral grid elements for all models represented as cells.

Fig. 6b

3.5 Optimal turbulence model selection

In pursuit of optimal turbulence modeling, we employed RANS turbulence models with dual precision to address steady-state turbulent flow challenges in ANSYS FLUENT 17.1, maintaining an isothermal environment. The assessment included various RANS-based turbulence models, including Reynolds stress (RS) [65], Standard k-ω (k-omega) [66], Standard k-ε (k-epsilon) [53], Realizable k-ε (RK-epsilon) [67], RNG k-ε (RNG k-ε) [68], and Shear stress transport k-ω (SST k-omega) [69]. The analysis aimed to identify the most suitable model and determined the SST k-omega turbulence model as the optimal choice due to its superior reliability and precision. Section 4.3 delves deeper into the nuances of model validation.

4 Model validation

4.1 Empirical data analysis

Empirical wind tunnel tests conducted by Tominaga and Stathopoulos [70] and An and Fung [69], emphasizing close-surface pollutant dispersal, served for validating CFD simulations and examining block configuration outcomes. These tests were performed in a boundary layer wind tunnel measuring 1.8 × 1.8 × 13 m (width, height, and length). Boundary conditions and surface roughness were harmonized with a power-law exponent to simulate an approaching wind profile. Within this controlled environment, a passive pollutant source (tracer gas: C2H4) was emitted at the street-level's base (Fig. 7a) with a concentration of one thousand parts per million. Wind speed was measured using a split-film probe, and pollution/gas concentrations were quantified via a fast-response flame ionization detector at a 1000 Hz sampling frequency, yielding 120,000 data records in 120 s. To replicate an urban context, cuboid building blocks were arranged near the pollutant source. Fig. 7b depicts specific variations in pollutant concentration ratios along the vertical centerline. For further test data insights, review Tominaga and Stathopoulos [70], Tominaga and Stathopoulos [71], and An and Fung [69].Fig. 7 Analysis of Urban Air Quality and Computational Domain: (a) Measurement of building arrangement (left) and area of interest on building blocks (right); (b) Ratio of pollutant concentration along the vertical axis; (c) Comparative analysis of pollutant concentration at street-level; (d) Dimensions of the computational domain.

Fig. 7

4.2 Setup of computational domains and boundaries

The computational domain, as illustrated in Fig. 7d, faithfully replicates the building block configurations as defined in the wind tunnel tests. Adhering to CFD guidelines established by Tominaga, Mochida [49], which stipulate the necessity of situating the lateral and upper boundaries at distances ≥15H from the target building (with H denoting the height of the tallest structure), shaping the domain's dimensions in H-scale along the X, Y, and Z axes. Blockage ratios for the domain models were consistently below 3.0 %, averaging approximately 2.2 %. Boundary conditions for the CFD simulations were derived from the experimental wind tunnel study, while the choice of the k-ω SST turbulence model aligned with analytical and modeling preferences.

The SST model's specific form is computed as follows:(7) ∂(ρk)∂t+∂(ρujk)∂xj=P1−β*ρωk+∂∂xj[(μ+σkμt)∂k∂xj]

(8) ∂(ρω)∂t+∂(ρujω)∂xj=γvtP1−βρω2+∂∂xj[(μ+σkμt)∂ω∂xj]+2(1−F1)ρσω2ω∂k∂xj∂ω∂xj

The variable ω is defined as:(9) ω=β*0.25k0.5β2

Where β2 = 0.07 and β* = 0.09 represent the model constants for the SST k-ω model.

The experimental setup involved ethylene (C2H4) as a tracer gas, covering 0.01H2 at a height of 0.25H with a volumetric flow rate of 5.83 × 10−6 m3/s, as specified by Tominaga and Stathopoulos [70] and An and Fung [69]. This research used wind tunnel-derived data, adopting information from an experiment of flow and pollutant concentration around buildings.

4.3 Comparative analysis and model affirmation

A comparative assessment of RANS turbulence models, namely RS, k-omega, k-epsilon, RK-epsilon, RNG k-ε, and SST k-omega, was conducted (Fig. 7c). The analysis relied on data from the cited wind tunnel experiments and large-eddy (LE) simulations [70]. The results of the RANS turbulence models exhibited strong concordance with wind tunnel and large-eddy simulation findings, particularly in relation to wind speed ratios and pollutant concentrations with respect to U0 (z). Among these models, SST k-omega displayed an average standard deviation (STDEVA) that closely approximated RK-epsilon and RNG k-ε, around 0.22. In contrast, it was lower compared to SKE, SKW, and RSM concerning the wind speed ratio (SST k-omega, RK-epsilon, RNG k-ε ≅ 0.22 < SKE, SKW, RSM). During the evaluation of the street centerline concentration of pollutants on the vertical axis, SST k-omega displayed a significantly lower standard deviation (STDEVA = 0.009) when compared to all other RANS turbulence models. Due to its accuracy and strong alignment with experimental results, the SST k-omega model was chosen for the present study.

4.4 Mesh refinement in computational models

As described in section 3.4, a 0.3 m mesh size was imposed to capture geometric intricacies. To ascertain that alterations in mesh intricacy do not influence simulation results, models were subjected to mesh refinement. Various grid types—high-resolution (19,075,286 grids, min mesh 0.1 m), medium-resolution (6,358,428 grids, min mesh 0.3 m, ideal), and low-resolution (2,725,040 grids, min mesh 0.7 m)—were evaluated for sensitivity. Uniform solver setups and computational domains were utilized for comparative analysis of airflow and pollutant dispersion at 1.5 m above ground, along planes A and B (Fig. 10a, Fig. 10b). Error analysis revealed minor distinctions between medium-resolution and high-resolution meshes, but significant divergences emerge between high-resolution and low-resolution meshes. Notably, in the A-plane (Fig. 10a, Fig. 10b), medium-resolution records NO2 concentration at 0.17, while high-resolution records NO2 = 5.12. Consequently, the validation of a medium-resolution grid (min mesh = 0.3 m) is confirmed for this study.

5 Results and discussion

5.1 Street-level airflow efficiency

Urban street-level airflow efficiency relies on two key metrics, ALA and ASR. Fig. 8a, Fig. 8b illustrates air age distributions at 1.5 m above the terrain, accompanied by the corresponding ASR values for each scenario. Fig. 9 in Subfigure A delineates average ALA characteristics observed at street-level (Fig. 9a) and within building-level (SBLs) contexts (Fig. 9c), complemented by an analysis of volume-averaged fundamental ventilation modes. Additionally, it examines the average ASR at specified points (designated in Fig. 10c), contributing to a comprehensive understanding of local atmospheric dynamics. Fig. 9 in Subfigure B further illustrates key parameters, including average momentum, dynamic pressure on buildings and the ground, average ALA at building facades, and momentum averages for all cases. Spatial ASR distributions closely mirror airflow patterns influenced by specific building arrangements. Optimal urban air circulation, achieving efficiency levels of 50–70 %, is attained when airflow aligns orthogonally across and parallel to street lanes. Our findings emphasize that increasing street width within larger sites enhances air mixing, resulting in augmented air circulation. Moreover, varying street-level building heights, particularly with taller structures, significantly enhances ventilation by 20–30 %. The street aspect ratio (H/W, building height/street width) exerts a profound influence on airflow enhancement and pollutant dispersion within urban streets, in close interaction with surface features like roofs and facades. Higher H/W ratios, for instance, H/W > 5.5, produce a remarkable 218.5 % enhancement in air access and ventilation. Street geometry, H/W ratio, and street orientation significantly influence wind control, thereby affecting wind energy harvesting from street surfaces. Therefore, the arrangement of urban patterns is pivotal.Fig. 8a Air age distributions at street-level (1.5 m above terrain) influenced by incoming dominant wind.

Fig. 8a

Fig. 8b Average volume of wind velocity/ASR at street-level (m/s).

Fig. 8b

Fig. 9 Performance Metrics and Airflow Characteristics: Subfigure A encompasses (a) Average ALA at the street-level (kg m/s), (b) Volume average of pressure (Pa), (c) Average ALA at the building-level (kg m/s), (d) Volume average of dynamic pressure (Pa), (e) Performance improvement of models, and (f) Average ASR in specified points and planes at the building-level (m/s). Subfigure B comprises (g) Average ALA at the building facade (kg m/s), (h) Average of dynamic pressure on building (Pa), (i) Average of momentum on the ground (kg m/s), (j) Average of dynamic pressure on ground (Pa), and (k) Average of momentum in specified points and planes at the building-level (m/s).

Fig. 9

On average, all models display an ASR of 1.52 during prevailing wind conditions. Specifically, when the prevailing wind aligns with the main street grid, we observe substantial wind velocity reductions in G14000A-BV6 (53.5 %), G14000A-BV4 (55.3 %), and G14000A-BV7 (56.0 %) cases. At street level, the average ALA volume was assessed across sites featuring low-rise, mid-rise, and high-rise buildings, yielding an average of 77.5. Notably, mid-rise sites recorded the lowest average at 64.8 under prevailing wind conditions. However, this value increased to 69.5, marking a 12 % rise in high-rise sites, and significantly higher at 99.8, indicating a substantial 51 % increase in low-rise sites when the incoming wind direction remained constant.

The study divulges intriguing aerodynamic insights. Specifically, it unveils a marginal ASR variance between inlet and outlet winds, denoting a value of 1.01 (G4500-BV12). In stark contrast, G14000A-BV8c and G14000A-BV4 exhibit the most marked difference, presenting a ratio of 2.8. An astounding wind velocity increase of 218.5 %, more than double the baseline, observed when the H/W ratio exceeds 5.5 and the prevailing wind aligns orthogonally with the street grid at 90°. This significant enhancement in wind velocity is primarily attributed to the incorporation of high-rise buildings with H/W > 5.5, especially those featuring simple cubic shapes and medium-density ratios, which markedly improve airflow within the urban canyon.

In this context, high-rise buildings with uncomplicated cubic configurations and medium-density proportions exhibit the highest wind velocity, intensifying the wind speed on the central street by 1.5 times. Consequently, the inclusion of high-rise structures augments canyon airflow, mitigates temperatures, and aids air pollutant dispersion within urban environments. Investigation reveals that fine-tuning the H/W ratio (shallow = H/W < 0.5, deep = H/W ≥ 2, and regular H/W = 1) unlocks a 35 % increase in avenue canyon wind velocity. At the street-level, to ameliorate concentrations of pollutants, particularly elevated in the region between the pollutant source and the leeward building facade, an H/W ratio adjustment to 1/2 is recommended.

Summarily, in urban environments like Wuhan and various other cities grappling with the transformation of street grids, the evolution of site designs and building configurations emerges as the central element shaping air dynamics in pedestrian zones. Buildings with H ≤ W, marked by the gradual widening of their structures, lead to decreased wind velocity near the building facades. This diminishment is especially pronounced in the central segments of the vertical planes adjoining the building facades (see fig. 10c for detailed plane analysis). Notably, wind speed values at the top and lateral edges remain virtually identical across all cases. In the case of buildings with H ≥ W, characterized by heightened structures, wind velocities at the edges experience acceleration.

In proximity to the building bases, an area exhibits diminished wind speeds, wherein the horseshoe vortex governs the airflow. A closer examination of the simulated patterns reveals that the height of the horseshoe vortex remains relatively constant regardless of the building's height. Maximum ASR values transpire at the top and lateral edges, where wind speeds are notably elevated. Subsequently, the elevation of building heights amplifies the maximum ASR values. In effect, a uniform ASR distribution across the central region of the facades becomes evident, representing uniform zones of low velocity and high pressure. The area characterized by the horseshoe vortex's formation displays lower ASR values, attributed to the rotational nature of the vortex, which prolongs air residence time in this locale, consequently raising the local air temperature and reducing local ASR values. Buildings with H = W also exhibit higher local ASR as their sizes increase, especially near the lateral and top edges. Examining S-plane data in Fig. 10c uncovers substantial average wind velocity reductions, particularly at 35 % (G14000A-BV8b), 35.2 % (G14000A-BV8a), 36.7 % (G14000A-BV3), and 38.5 % (G14000A-BV7).Fig. 10a Building-Level Airflow Analysis: (a) Air age distribution at the building-level influenced by dominant incoming winds;

Fig. 10a

Fig. 10b Volumetric analysis of ASR volumetric analysis at the building-level; (c) Detailed analysis at specific planes and points.

Fig. 10b

These findings illuminate the pivotal role of pressure differentials as the primary governing mechanism dictating airflow dynamics across building surfaces. This phenomenon comes to the forefront, particularly in the domain of wind pressure-driven ventilation, highlighting variances contingent upon the intricate interplay of building morphology. Optimizing synergistic relationships within site configurations, particularly the interactions between building structures and their surroundings through 3D-UFO, promotes efficient natural resource utilization and augments local wind environments. This meticulous site design enhances wind energy harvesting and reinforces sustainable urban ventilation strategies. Subsequent observations emphasize the vital importance of maintaining a minimum wind velocity threshold of 0.3 m/s to optimize aerodynamic turbulence, which promotes efficient air mixing. This optimization significantly enhances convective airflow dynamics, thereby contributing to the dispersion of airborne particulates. Our findings affirm that medium-to-high height-to-width (H/W) ratios robustly enhance air velocities and foster turbulent eddies within the urban canyon, thereby improving atmospheric quality and local thermal conditions at pedestrian levels. Heightened building density in narrow street configurations sharply curtails airflow, leading to marked reductions in flow velocity, momentum, turbulent kinetic energy, and limiting flow adjustments. This coincides with reduced in-canopy airflow in the fully developed region. In contrast, wider streets and lower building density, fueled by enhanced spatial airflow exchange, significantly bolster vertical ventilation capacity. The G4500-BV9 site, characterized by cylindrical building obstacles oriented at a 90° angle to the incoming wind, boasts the highest inflow wind velocity, coupled with the most minimal wind velocity reduction at 37.1 %. The heightened coefficient of variation underscores that irregular or rough surfaces engender amplified turbulent flow and increased flow rate variability, attributed to intensified chaotic motion of fluid molecules. This effect, despite maintaining identical mean flow rates, results in a more conspicuous reduction in wind speed at a height of 1 H and a less pronounced decrease at a height of 1.5 H from the obstruction, extending over a horizontal distance of 5 H, as compared to the impact of smooth surface turbulent flow.

Findings indicate that projects situated on expansive sites offer the most effective street-level ventilation when paired with taller structures. Among the large sites under examination, G14000A–BV4 and G14000A–BV6 exhibit superlative street-level ventilation performance across a spectrum of prevailing wind directions, demonstrating an impressive minimal average ALA at 43.4 (ASR = 1.01) and 44.4 (ASR = 0.99). Conversely, G14000A-BV8b, G14000A-BV8a, G1 4000A-BV7, and G14000A-BV8c exhibit inferior street-level ventilation efficiency, despite featuring tall buildings and identical site division scheme. Furthermore, G31000a-BV2, characterized by a smaller site and mid--rise linear buildings, presents the highest ALA yet the poorest street-level ventilation.

Utilizing the average impact index (ρv=∑βvμvNv), where βV signifies the variable mean, μV represents each variable's coefficient, and NV denotes the count of observations for each variable, G4500–BV1 emerges as the optimal model. This model excels in promoting street-level ventilation in smaller sites dominated by low-rise buildings. In the context of larger sites featuring taller structures, models G14000A-BV4 and G14000A-BV8c effectively contribute to the improvement of urban ventilation. Nevertheless, for midsize and extensive site layouts typified by G14000A–BV3, G22000Aa–BV5, G31000a–BV13, and G31000b–BV11, which are characterized by fragmented mid-rise H and U building forms, the observed ventilation levels remain relatively moderate. This observation underscores that densely packed urban plans within small and medium-sized projects experience constrained street-level ventilation in comparison to more spacious layouts featuring larger open areas.

In contrast to regular high-rise building forms like G14000A-BV8b and G14000A-BV8a, irregularly shaped high-rise structures (G14000A-BV7) integrated with mid-rise and low-rise buildings with extensive layouts improve street-level ventilation by approximately 0.5 %. Accordingly, projects with typical mid-rise buildings could potentially amplify this effect by 8.24 % (G31000b-BV11), 14 % (G31000a-BV13), 23.5 % (G22000Aa-BV5), 29.8 % (G14000A-BV3), 37.9 % (G14000A-BV6), and 39.2 % (G14000A–BV4) at street-level.

In the majority of cases, buildings characterized by higher floor area ratios exhibit negligible positive impact on the enhancement of street-level ventilation, particularly in smaller sites. Concurrently, the adoption of small site-divisional strategies featuring service lanes demonstrates a pronounced efficacy in augmenting street-level ventilation. A comparative analysis between G22000Aa-BV5 and G14000A-BV8a validates this outcome. Examination of diverse high-rise block configurations, including G14000A-BV8a, G14000A-BV8b, and G14000A-BV8c, affirms that distinct building forms represent the paramount factor influencing air dynamics at street-level. However, disparate building forms within the spectrum of low-rise, mid-rise, and high-rise structures manifest varying effects on ventilation enhancement through the strategic utilization of service paths.

5.2 Building-level airflow efficiency

Within the 3D-UFO framework, our investigation delved into building-level airflow efficiency, shedding light on airflow effectiveness from above street-level to a height of 114 m. Fig. 10 visually represented air age distributions within the average building heights above pedestrian levels, concurrently with observed ASR under prevailing wind conditions. Acknowledging ALA's pivotal role in assessing air quality near buildings for residents' welfare, we calculated its building surface mean by integrating air age values across each facet. All models exhibited an average ALA of 84.2 under prevailing wind conditions. Further analysis revealed a compatible relationship between building-level and street-level air circulation when the angle of the prevailing wind and the central street grid increased. Fig. 9 in Subfigure A featured an exhaustive analysis of critical ventilation modes, coupled with ALA and ASR at the building-level. Notably, G14000A-BV8a exhibited the highest average ALA at 115.4 (100 %).

Subsequently, G14000A-BV8b achieved 114.9 (0.43 % improvement), G14000A-BV7 reached 114.7 (0.6 % improvement), G31000a-BV2 attained 87.4 (24.2 % improvement), G22000Ab-BV2 revealed 86.9 (24.6 % improvement), G31000b-BV11 displayed 84.8 (26.5 % improvement), G31000a-BV13 exhibited 84.4 (26.8 % improvement), G22000Aa-BV5 exposed 83.5 (27.6 % improvement), and G14000A-BV3 demonstrated 82.8 (28.2 % improvement). G14000A-BV6 achieved 81.5 (29.3 % improvement), G4500-BV9 recorded 77 (33.2 % improvement), G4500-BV15 reached 76.2 (33.9 % improvement), G4500-BV14 demonstrated 76 (34.1 % improvement), G4500-BV10 attained 75.94 (34.2 % improvement), G4500-BV1 revealed 75.91 (34.22 % improvement), and G4500-BV12 exhibited 75.7 (34.4 % improvement). Results from G14000A-BV8c (ALA = 66.4, 42.4 % improvement) and G14000A-BV4 (ALA = 56.1, 51.3 % improvement) underscore enhanced urban air circulation within expansive sites (G14000 = 4000 m2). Fig. 9e elucidated these outcomes.

During our meticulous investigation, elevating the H/W (height-to-width) ratio significantly diminished canyon retention time, enhancing airflow dynamics. This revelation offers an intriguing perspective on air circulation within canyons. Additionally, implementing a centralized tall building plan can enhance local air ventilation. Compared to big sites, mid-sized sites (G22000 = 2000 m2) exhibited solely modest building-level air circulation. In smaller sites (G31000 = 1000 m2) and more compact ones (G4500 = 500 m2), the probability of attaining adequate ventilation decreased as on-site building density rose. Remarkably, G4500-BV12 employed a segmented operational strategy for diminutive sites, yet it achieved the third-lowest average ALA volume at 75.7. This reveals that enhancing ventilation in smaller sites is achievable through the implementation of well-structured symmetrical L-shaped buildings featuring central spaces strategically designed to incorporate openings aligned with predominant wind directions.

Building-level natural ventilation extends beyond its influence on urban outdoor ventilation efficiency, concurrently exerting an impact on indoor air quality. Fig. 9g in Subfigure B provided a quantitative assessment of the indoor natural ventilation potential of each building, characterized by ALA measurements at the facade. These measurements indicate the accessibility of fresh outdoor air for indoor natural ventilation, enhancing overall air quality. Among the building configurations assessed, G14000A-BV8c achieved the highest average ALA at the building facade, an impressive 152.5 (100 %), attributed to its distinctive cuboid morphology and voluminous depth.

Subsequently, G31000a-BV2 at 84.2, G4500-BV9 at 74, G22000Aa-BV5 at 69.1, G31000b-BV11 at 64.8, G14000A-BV7 at 63.2, G14000A-BV6 at 63.06, G4500-BV10 at 63.03, G4500-BV15 at 62.75, G14000A-BV3 at 62.02, G4500-BV1 at 61.3, G22000Ab-BV2 at 59.5, G31000a-BV13 at 59.4, G4500-BV12 at 58.54, G14000A-BV8a at 58.47, G14000A-BV8b at 57.1, and G4500-BV14 at 55.8 demonstrated substantial air circulation improvements of 44.7 %, 51.4 %, 54.6 %, 57.5 %, 58.5 %, 58.64 %, 58.66 %, 58.8 %, 59.3 %, 59.8 %, 60.9 %, 61.04 %, 61.63 %, 61.65 %, 62.5 %, and 63.4 %, respectively.

Overall, G14000A-BV4 exhibited the highest potential in ameliorating indoor air quality, characterized by a minimal 50.2 ALA at the building facade within an expansive 4000 m2 site, and optimal ventilation efficiency, resulting in a substantial 67 % enhancement. The investigation, grounded in 3D-UFO analysis, unequivocally revealed a robust correlation between natural ventilation capacities within interior spaces and the intricate interplay of building forms and configurations, notably accentuated in elevated constructions and high-rise block schemes within larger projects. Emphasizing the indispensable role of integrating urban building geometry and directional wind patterns under the H/W ratio for optimal design efficiency, our elucidation markedly advanced the understanding of ventilation potential across diverse site classifications. The study conclusively demonstrated the superiority of meticulous design in small-scale sites with Y-formed structures (G4500-BV14) for enhancing indoor natural ventilation. Subsequent observations identified this superiority in larger sites with T-shaped configurations (G14000A-BV8b) and regular high-rise building plans (G14000A-BV8a).

Following, G4500-BV12 highlighted systematically configured L-shaped construction plans in small-scale projects, revealing a 61.63 % enhancement in ventilation efficiency. This underscores the attainability of optimal airflow within enclosed spaces through a symmetrical design, integrating central courtyards and strategically positioned openings along both vertical and horizontal axes. In G31000a-BV13, a symmetric U-shaped and H-shaped integration, and in G22000Ab-BV2, elongated, depth-oriented geometric configurations led to significant improvements of 61.04 % and 60.9 % in indoor natural ventilation. Across a spectrum of site scales, encompassing large to small with different building plans, G14000A-BV8c (high-rise), G31000a-BV2 (mid-rise), and G4500-BV9 (low-rise) exhibited reduced potential for internal natural ventilation. Observations elucidated that a heightened H/W value exerted a detrimental influence on indoor natural ventilation within these scenarios, wherein turbulent airflow patterns induced, impeding the efficient interchange of indoor and outdoor air and consequently diminishing the overall efficacy of ventilation.

5.3 Discussion

This study employed the innovative 3D-UFO methodology to scrutinize air circulation and pollutant dispersion within densely built-up urban environments. The results realistically elucidate multifaceted outcomes and portray the challenges associated with urban pollution, offering quantitative insights into potential strategies for improving air quality and fostering healthier living conditions. Our approach, grounded in a comprehensive analysis of LULCC, coupled with CFD and systematic exploration, guided by established principles of urban airflow study protocols, facilitated robust simulation and assessment of airflow in diverse site scenarios and block configurations.

In large sites of approximately 4000 m2, distinguished by expansive building plans, a discernible inclination manifested toward optimizing ventilation at the building-level. Nevertheless, these sites often demonstrated limited capacity to improve indoor air circulation and generally did not assure high air circulation performance at the street-level. This observation aligned with prior findings by Zhang et al. [35]. Among large plan projects featuring high-rise buildings, G14000A-BV8c exhibited a grade A street-level ventilation performance of 10.7 %, particularly effective in a deep avenue canyon (H/W ≥ 2), ranking first in building-level ventilation (ALA = 66.4). However, it does not prioritize indoor ventilation. Contrarily, G14000A-BV4 demonstrated superior building-level ventilation, ranking highest in air circulation and achieving elevated indoor ventilation.

G14000A-BV6 exhibited optimal building-level air circulation performance, securing the highest rank in this aspect, and attained the second position in street-level ventilation. Notably, a limitation manifested in its indoor ventilation, characterized by suboptimal conditions. This observation aligned with insights from Antoszewski's study [34], wherein various contributing factors, including the strategic positioning of elevated central buildings with a relatively modest frontal area, played a pivotal role in shaping air circulation dynamics within urban settings. Therefore, implementing a strategy for centrally located elevated buildings with a relatively small frontal area and preventing deep spaces with enclosed plans, particularly under 3D-UFO, can promote airflow within the urban context. These findings resonate with those by Nugroho et al. [72], which explored how geometric parameters (such as form and dimensions) and volumetric characteristics influence the thermal microclimate around high-rise structures, thereby enhancing our understanding of how spatial configurations impact airflow patterns and environmental quality in dense urban settings.

Our 3D-UFO study revealed that additional service paths enhanced street-level ventilation in small and medium-sized sites (approximately 500 m2, 1000 m2, and 2000 m2). Conversely, such strategies may lead to limited ventilation at building-level in larger site projects. For instance, G2 2000Aa-BV5 ranks fourth in street-level and mid-rank building-level ventilation in mid-sized projects. Small and medium-sized sites may employ distinct strategies, such as small floor plans and elevated structures along block axes, to optimize building-level air circulation and indoor ventilation, unlike service path approaches. The most reliable examples for sites with low-rise buildings were G4500-BV14 and G4500-BV12, each characterized by superior indoor air circulation and satisfactory building-level ventilation, albeit with comparatively less dependable street-level ventilation. Within this classification, G4500-BV1 manifested the most delicate air circulation at street-level, whereas G4500-BV10, concurrently attaining the second-highest rating in street-level air circulation, demonstrated the most optimal ventilation at the building-level.

In sites with mid-rise buildings, G31000a-BV2 ranked first in poor ventilation at street-level, building-level, and indoor air quality compared to mid-sized sites (1000–2000 m2). Following a discernible airflow pattern, G31000a-BV2 ranked in the lower tertile, demonstrating analogous poor ventilation at street-level, building-level, and indoor air quality.

Ventilation degradation, eliciting adverse effects, exhibited amelioration at both street and building levels (SBLs) within larger sites (G22000). Notably, mid-rise structures, exemplified by G22000Aa-BV5, secured third and fourth rankings in this enhancement paradigm. G22000Ab-BV2 exhibited penultimate suboptimal ventilation at SBLs, attributed to its profound avenue canyon (H/W = 4 ≥ 2). Jamei and Rajagopalan [73] expounded this intricacy, emphasizing the connection between air temperature within deep canyons and the magnitude of the H/W ratio. This denotes that an elevated H/W ratio engenders heightened air temperatures, attributable to constrained air circulation and the influence of surface temperatures.

Canyon investigations elucidated the intricate relationship between enhanced ventilation dynamics and spatio-temporal aspects of 3D-UFO planning. Increased ventilation efficiency notably curtailed retention times, in contrast to scenarios with decreased efficiency. Elevated H/W ratios correlated with prolonged retention, especially on the wind-facing sides of tall buildings, in contrast to shorter retention on leeward sides in canyons. These results aligned with prior research [74], highlighting the adverse impact of increased building spacing on pedestrian ventilation. Maintaining an H/W ratio of 1.0 proved crucial in preserving a significant amount of air age along main streets, a result further substantiated by our numerical outcomes. This study also extend the work of Du et al. [75], who observed that building height variability significantly influences urban ventilation patterns, supporting our findings that indicate the strategic placement of high-rise structures can enhance air circulation.

The collision of airflows caused a three-dimensional winding stream on the leeward side of structures and an increasing channeling stream on the main street. Consequently, adjusting the aspect ratio of the three-dimensional street canyon influenced the pattern and distribution of airflow throughout the urban canopy. By holding the H/W ratio at 1.0 or less (H/W ≤ 1.0), increasing the street aspect ratio decreased ASR on the main street while elevating it within the street canyon. According to Chen et al. [76], canyons with lower H/W ratios showed increased net longwave radiation emission and enhanced convective cooling.

Maintaining H/W ≥ 1.0, an increase in the street aspect ratio elevated ASR on the main street and reduced it in the street canyon. The significance of geometry remained evident in wind performance, exemplified by mitigating heat dissipation through an increase in H/W ratio to 5.5 [77]. As the street aspect ratio increased, the surface layer's roughness in the urban environment augmented, intensifying turbulent flow and enhancing channeling flow within the main street. Notably, the tallest buildings on the leeward side experienced subdued winds. Maintaining H/W ≤ 1.0 resulted in a progressive growth of ASR value in the outermost street canyon along the flow direction. This pattern became unobservable when the H/W ratio exceeded 1.0, possibly because the street corner vortex improves vertical flux and enhances airflow in the street canyon simultaneously in building configurations with low H/W ratios. Echoing Huang et al. [78], this study emphasizes the critical importance of integrated design strategies and urban form optimization for enhancing air quality.

Through 3D-UFO can profoundly alter incoming wind patterns within urban blocks, inducing discernible shifts in turbulence characteristics and momentum flux by strategically manipulating the urban skyline. Thus, variations in urban skylines, characterized by low-rise, mid-rise, and high-rise buildings in classified sites, can either reduce or enhance urban ventilation capacity. The geometry of building arrays can be a primary factor influencing pollutant dispersion in urban blocks, demonstrating a strong correlation between density and configuration, consistent with insights from Antoszewski et al. [34].

To summarize, our findings significantly advance the field of urban planning and environmental science by providing robust, quantitative insights into the dynamics of urban airflow and pollutant dispersion. By systematically analyzing various urban block configurations using the 3D-UFO approach, this study offers valuable guidelines for designing urban environments that prioritize air quality and climate resilience. The evidence supporting the efficacy of specific building arrangements and configurations in enhancing air circulation and reducing pollutant concentrations can inform future urban planning and policy decisions, contributing to healthier and more sustainable cities.

Despite the comprehensive nature of this study, certain limitations must be acknowledged. The simulations were based on idealized urban block configurations and may not fully capture the complexities of real-world urban environments. Additionally, the focus on specific building typologies and configurations may limit the generalizability of the findings to other urban contexts. Future research should aim to incorporate more diverse urban forms and account for additional environmental variables, such as varying climatic conditions and pollutant sources. By addressing these limitations, future studies can build on our findings and develop more refined approaches to urban airflow and pollutant dispersion analysis, ultimately contributing to more effective urban design strategies.

6 Conclusion

This research has advanced the field of urban environmental studies by introducing the 3D Urban Form Optimization (3D-UFO) methodology, which seamlessly integrates GIS-based Land-Use and Land-Cover Change (LULCC) analysis with Computational Fluid Dynamics (CFD) simulations. Our comprehensive findings emphasize the pivotal role of urban block design in modulating air-circulation efficiency and pollutant dispersion within densely constructed environments. The study key findings are as follows.• Significant Variability in Air-Circulation Efficiency: Significant variability in air-circulation efficiency was observed at both street and building levels across the evaluated urban configurations using the 3D-UFO approach.

• Enhanced Airspeed-Ratio (ASR) and Average-Age-of-Local-Air (ALA): Adjustments in street-aspect-ratio and building-height-ratio significantly enhanced ASR and ALA, yielding marked improvements in urban air quality metrics.

• Optimization in Large Sites: Large sites (approximately 4000 m2) demonstrated a focus on optimizing ventilation at the building level, though they often exhibited limited street-level air circulation performance.

• Performance of High-Rise Buildings: In larger urban settings, high-rise buildings with Height-to-Width (H/W) ratios >5.5 show a 218.5 % increase in ventilation efficiency. This enhancement, particularly notable with cubic forms and medium-density layouts, significantly enhances wind speeds, emphasizing the critical role of site characteristics in optimizing urban airflow dynamics.

• Impact of Service Paths: The introduction of additional service paths in small and medium-sized sites (approximately 500–2000 m2) enhanced street-level ventilation but frequently compromised building-level air circulation.

• Effectiveness in Low-Rise Building Sites: Low-rise building sites with smaller floor plans and elevated structures along block axes achieved superior indoor air circulation and building-level ventilation.

• Challenges in Mid-Rise Building Sites: Sites with mid-rise buildings and deep canyons (H/W ≥ 2) exhibited poor ventilation at both street and building levels. This corroborates prior research on the negative impact of high H/W ratios on air circulation, highlighting that the beneficial effects of higher H/W ratios on ventilation are context-dependent and vary based on urban configuration and building forms.

• Maintaining Street Aspect Ratio: Maintaining a street aspect ratio with H/W ≤ 1.0 significantly improved air circulation within the street canyon, enhancing convective cooling and reducing air retention times.

Despite these advancements, the study has limitations. The reliance on static configurations of predefined urban layouts potentially overlooks dynamic urban planning scenarios that might influence airflow dynamics differently. Furthermore, necessary simplifications in CFD simulations for computational feasibility could have constrained the accuracy of airflow predictions under specific environmental conditions. Future research should explore the socio-economic determinants shaping urban planning decisions and their implications for implementing 3D-UFO strategies. Enhanced methodologies could incorporate real-time data for dynamic modeling of urban airflow and pollutant dispersion, thereby improving predictive accuracy and robustness. Advanced computational techniques should be explored to simulate complex urban environments with greater fidelity and detail. Overall, the 3D-UFO methodology, beyond its specific applications in the current case study, holds significant promise for broader implementation in urban contexts globally. By enhancing our understanding of how urban form influences environmental quality and public health, this research contributes substantively to the evolving field of sustainable urban planning and design.

Funding

This article is supported by the Shanghai Science and Technology Committee (Grant No. 21DZ1204500 ), 10.13039/501100012166 National Key R&D Program of China (2023YFC3806900 ), 10.13039/501100012166 National Key R&D Program of China (2022YFE0141400 ), 10.13039/501100001809 National Natural Science Foundation of China (Grant No. U1913603 ), the Shanghai Municipal Science and Technology Major Project (2021SHZDZX0100 ), the Fundamental Research Funds for the Central Universities, the Postdoctoral Fellowship Program of CPSF (GZC20241226 ), and Special Funding for Postdoctoral Researchers from the 10.13039/501100002858 China Postdoctoral Science Foundation (2024T170669 ).

Data availability

All of the data are available in the manuscripts.

CRediT authorship contribution statement

Mehdi Makvandi: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Philip F. Yuan: Supervision, Funding acquisition. Qunfeng Ji: Formal analysis. Chuancheng Li: Resources. Mohamed Elsadek: Methodology. Wenjing Li: Visualization, Formal analysis. Ahmad Hassan: Software. Yu Li: Visualization, Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A Supplementary data

The following is the Supplementary data to this article.Multimedia component 1

Multimedia component 1
==== Refs
References

1 Sun Y. Zhao S. Spatiotemporal dynamics of urban expansion in 13 cities across the Jing-Jin-Ji Urban Agglomeration from 1978 to 2015 Ecol. Indicat. 87 2018 302 313
2 Qin H. Influence of site and tower types on urban natural ventilation performance in high-rise high-density urban environment Build. Environ. 179 2020 106960
3 Zhang P. Remote sensing modeling of urban density dynamics across 36 major cities in China: fresh insights from hierarchical urbanized space Landsc. Urban Plann. 203 2020 103896
4 Makvandi M. Urban heat mitigation towards climate change adaptation: an eco-sustainable design strategy to improve environmental performance under rapid urbanization Atmosphere 14 2023 10.3390/atmos14040638
5 He X. High-resolution dataset of urban canopy parameters for Beijing and its application to the integrated WRF/Urban modelling system J. Clean. Prod. 208 2019 373 383
6 Makvandi M. Advancing Urban Resilience Amid Rapid Urbanization: An Integrated Interdisciplinary Approach for Tomorrow’s Climate-Adaptive Smart Cities—A Case Study of Wuhan, China Smart Cities 7 2024 2110 2130 10.3390/smartcities7040084
7 Hong L. Tao W. Urban design practice towards planning management in China—urban design guidelines in wuhan city as examples Pan Q. Cao J. Recent Developments in Chinese Urban Planning: Selected Papers from the 8th International Association for China Planning Conference, Guangzhou, China, June 21 - 22, 2014 2015 Springer International Publishing Cham 349 367
8 Wu Y. Chen H. The diffusion of traffic pollutants in different residential blocks based on spatial morphological clustering Build. Environ. 228 2023 109860
9 Zhu J. Tian L. Editorial: inclusive urbanization in the 21st century China Cities 60 2017 415 419
10 Shareef S. The impact of urban morphology and building's height diversity on energy consumption at urban scale. The case study of Dubai Build. Environ. 194 2021 107675
11 Shareef S. Altan H. Urban block configuration and the impact on energy consumption: a case study of sinuous morphology Renew. Sustain. Energy Rev. 163 2022 112507
12 Liu D. Clarke K.C. Chen N. Integrating spatial nonstationarity into SLEUTH for urban growth modeling: a case study in the Wuhan metropolitan area Comput. Environ. Urban Syst. 84 2020 101545
13 Shen P. Wang Z. How neighborhood form influences building energy use in winter design condition: case study of Chicago using CFD coupled simulation J. Clean. Prod. 261 2020 121094
14 Calautit J.K. Development of a natural ventilation windcatcher with passive heat recovery wheel for mild-cold climates: CFD and experimental analysis Renew. Energy 160 2020 465 482
15 Mirzaei P.A. CFD modeling of micro and urban climates: problems to be solved in the new decade Sustain. Cities Soc. 69 2021 102839
16 Khalilzadeh A. Ge H. Ng H.D. Effect of turbulence modeling schemes on wind-driven rain deposition on a mid-rise building: CFD modeling and validation J. Wind Eng. Ind. Aerod. 184 2019 362 377
17 Gimenez J.M. Bre F. Optimization of RANS turbulence models using genetic algorithms to improve the prediction of wind pressure coefficients on low-rise buildings J. Wind Eng. Ind. Aerod. 193 2019 103978
18 Wu C.-H. Kopp G.A. A quasi-steady model to account for the effects of upstream turbulence characteristics on pressure fluctuations on a low-rise building J. Wind Eng. Ind. Aerod. 179 2018 338 357
19 Liu S. CFD simulations of wind distribution in an urban community with a full-scale geometrical model Build. Environ. 117 2017 11 23
20 Shirzadi M. Naghashzadegan M. Mirzaei P.A. Improving the CFD modelling of cross-ventilation in highly-packed urban areas Sustain. Cities Soc. 37 2018 451 465
21 Barbano F. A numerical study of the impact of vegetation on mean and turbulence fields in a European-city neighbourhood Build. Environ. 186 2020 107293
22 Kosutova K. Cross-ventilation in a generic isolated building equipped with louvers: wind-tunnel experiments and CFD simulations Build. Environ. 154 2019 263 280
23 Tominaga Y. Blocken B. Wind tunnel analysis of flow and dispersion in cross-ventilated isolated buildings: impact of opening positions J. Wind Eng. Ind. Aerod. 155 2016 74 88
24 Shirzadi M. Tominaga Y. Mirzaei P.A. Experimental and steady-RANS CFD modelling of cross-ventilation in moderately-dense urban areas Sustain. Cities Soc. 52 2020 101849
25 Brown A.L. An urban dispersion inspired scenario for CFD model validation Fire Saf. J. 2020 103130
26 Perén J.I. CFD analysis of cross-ventilation of a generic isolated building with asymmetric opening positions: impact of roof angle and opening location Build. Environ. 85 2015 263 276
27 Jia H. Kikumoto H. Construction of urban turbulent flow database with wavelet-based compression: a study with large-eddy simulation of flow and dispersion in block-arrayed building group model J. Wind Eng. Ind. Aerod. 2020 104433
28 Kozmar H. Surface pressure on a cubic building exerted by conical vortices J. Fluid Struct. 92 2020 102801
29 Kubilay A. Numerical simulations of wind-driven rain on an array of low-rise cubic buildings and validation by field measurements Build. Environ. 81 2014 283 295
30 Liu T. Lee W.L. Evaluating the influence of transom window designs on natural ventilation in high-rise residential buildings in Hong Kong Sustain. Cities Soc. 62 2020 102406
31 Yim S.H.L. Air ventilation impacts of the “wall effect” resulting from the alignment of high-rise buildings Atmos. Environ. 43 32 2009 4982 4994
32 Zhang X. Evaluation of pedestrian wind comfort near ‘lift-up’ buildings with different aspect ratios and central core modifications Build. Environ. 124 2017 245 257 32287974
33 Aboelata A. Vegetation in different street orientations of aspect ratio (H/W 1:1) to mitigate UHI and reduce buildings' energy in arid climate Build. Environ. 172 2020 106712
34 Antoszewski P. Krzyżaniak M. Świerk D. The future of climate-resilient and climate-neutral city in the temperate climate zone Int. J. Environ. Res. Publ. Health 19 2022 10.3390/ijerph19074365
35 Zhang Y. Numerical studies of passive and reactive pollutant dispersion in high-density urban models with various building densities and height variations Build. Environ. 177 2020 106916
36 Foroutan H. Numerical analysis of pollutant dispersion around elongated buildings: an embedded large eddy simulation approach Atmos. Environ. 187 2018 117 130
37 Tamura Y. Xu X. Yang Q. Characteristics of pedestrian-level Mean wind speed around square buildings: effects of height, width, size and approaching flow profile J. Wind Eng. Ind. Aerod. 192 2019 74 87
38 Antoniou N. CFD and wind-tunnel analysis of outdoor ventilation in a real compact heterogeneous urban area: evaluation using “air delay” Build. Environ. 126 2017 355 372
39 Gamero-Salinas J. The influence of building form variables on the environmental performance of semi-outdoor spaces. A study in mid-rise and high-rise buildings of Singapore Energy Build. 230 2021 110544
40 Kaseb Z. A framework for pedestrian-level wind conditions improvement in urban areas: CFD simulation and optimization Build. Environ. 184 2020 107191
41 Wang W. Large-eddy simulations of ventilation for thermal comfort — a parametric study of generic urban configurations with perpendicular approaching winds Urban Clim. 20 2017 202 227
42 Zhang Y. Urban expansion simulation towards low-carbon development: a case study of Wuhan, China Sustain. Cities Soc. 63 2020 102455
43 Makvandi M. The Interactive Impact of Building Diversity on the Thermal Balance and Micro-climate Change under the Influence of Rapid Urbanization vol. 11 2019 1662
44 Yuan M. Examining the associations between urban built environment and noise pollution in high-density high-rise urban areas: a case study in Wuhan, China Sustain. Cities Soc. 50 2019 101678
45 Shi Z. Street grids for efficient district cooling systems in high-density cities Sustain. Cities Soc. 60 2020 102224
46 Xiao Y. Street network accessibility-based methodology for appraisal of land use master plans: an empirical case study of Wuhan, China Land Use Pol. 69 2017 193 203
47 Yue H. Zhu X. Exploring the relationship between urban vitality and street centrality based on social network review data in wuhan, China Sustainability 11 16 2019
48 Bray D. Designing to govern: space and power in two wuhan communities Built. Environ. 34 4 2008 392 407
49 Tominaga Y. AIJ guidelines for practical applications of CFD to pedestrian wind environment around buildings J. Wind Eng. Ind. Aerod. 96 10 2008 1749 1761
50 Mochida A. Comparison of various k-ε models and DSM applied to flow around a high-rise building-report on AIJ cooperative project for CFD prediction of wind environment Wind Struct. 5 2_3_4 2002 227 244
51 Franke J. The COST 732 Best Practice Guideline for CFD simulation of flows in the urban environment: a summary Int. J. Environ. Pollut. 44 1–4 2011 419 427
52 Yoshie R. CFD prediction of wind environment around a high-rise building located in an urban area Proceedings of the Fourth International Symposium on Computational Wind Engineering 2006 CWE2006 Yokohama, Japan
53 Menter F. Two-equation eddy-viscosity transport turbulence model for engineering applications. AIAA journal AIAA J. 32 8 1994 1598 1605
54 Zhong H.-Y. Wind-driven pumping flow ventilation of highrise buildings: effects of upstream building arrangements and opening area ratios Sci. Total Environ. 722 2020 137924
55 Tominaga Y. Computational fluid dynamics simulation of snowdrift around buildings: past achievements and future perspectives Cold Reg. Sci. Technol. 150 2018 2 14
56 Shirasawa T. Development of CFD method for predicting wind environment around a high-rise building part 2: the cross comparison of CFD results using various k-models for the flowfield around a building model with 4:4:1 shape AIJ Journal of Technology and Design 18 2003 169 174
57 China J.s.b. Load Code for the Design of Building Structures, GB50009-2001 : National Standard of the People's Republic of China 2002 China Architecture & Building Press Beijing
58 Gb Load Code for the Design of Building Structures 2012
59 Touma J.S. Dependence of the Wind Profile Power Law on Stability for Various Locations Journal of the Air Pollution Control Association 27 9 1977 863 866 10.1080/00022470.1977.10470503
60 Zhiyi Z. Potential of cross-ventilation channels in an ideal typical apartment building predicted by CFD and multi-zone airflow model J. Build. Eng. 44 2021 103408
61 Jie Y. The ventilation efficiency of urban built intensity and ventilation path identification: a case study of wuhan Int. J. Environ. Res. Publ. Health 18 2021 11684
62 Etheridge D. NATURAL VENTILATION OF BUILDINGS 2012
63 Hang J. Sandberg M. Li Y. Age of air and air exchange efficiency in idealized city models Build. Environ. 44 8 2009 1714 1723
64 Celik Procedure for estimation and reporting of uncertainty due to discretization in CFD applications J. Fluid Eng. 130 7 2008
65 Nakajima K. Ooka R. Kikumoto H. Evaluation of k-ε Reynolds stress modeling in an idealized urban canyon using LES J. Wind Eng. Ind. Aerod. 175 2018 213 228
66 Saleh A. Lakkis I. Moukalled F. A modified k-ω turbulence model for improved predictions of neutral atmospheric boundary layer flows Build. Environ. 223 2022 109495
67 Brozovsky J. Simonsen A. Gaitani N. Validation of a CFD model for the evaluation of urban microclimate at high latitudes: a case study in Trondheim, Norway Build. Environ. 205 2021 108175
68 Ameen A. Numerical investigation of the flow behavior of an isothermal corner impinging jet for building ventilation Build. Environ. 223 2022 109486
69 An K. Fung J.C.H. An improved SST k−ω model for pollutant dispersion simulations within an isothermal boundary layer J. Wind Eng. Ind. Aerod. 179 2018 369 384
70 Tominaga Y. Stathopoulos T. CFD Modeling of Pollution Dispersion in Building Array: evaluation of turbulent scalar flux modeling in RANS model using LES results J. Wind Eng. Ind. Aerod. 104–106 2012 484 491
71 Tominaga Y. Stathopoulos T. CFD simulations of near-field pollutant dispersion with different plume buoyancies Build. Environ. 131 2018 128 139
72 Nugroho N.Y. Triyadi S. Wonorahardjo S. Effect of high-rise buildings on the surrounding thermal environment Build. Environ. 207 2022 108393
73 Jamei E. Rajagopalan P. Effect of street design on pedestrian thermal comfort Architect. Sci. Rev. 62 2 2019 92 111
74 Zhang Y. Du X. Shi Y. Effects of street canyon design on pedestrian thermal comfort in the hot-humid area of China Int. J. Biometeorol. 61 8 2017 1421 1432 28197732
75 Du Y.X. Mak C.M. Tang B.S. Effects of building height and porosity on pedestrian level wind comfort in a high-density urban built environment Build. Simulat. 11 6 2018 1215 1228
76 Chen G. Effects of urban geometry on thermal environment in 2D street canyons: a scaled experimental study Build. Environ. 198 2021 107916
77 Martinelli L. Matzarakis A. Influence of height/width proportions on the thermal comfort of courtyard typology for Italian climate zones Sustain. Cities Soc. 29 2017 97 106
78 Wang J. Low-carbon-oriented commercial district urban form optimization and impact assessment analysis Build. Environ. 254 2024 111377
