Impact of Atmospheric Stability on Urban Bioaerosol Dispersion and Infection Risk: Insights from Coupled WRF–CFD Modeling

Abstract

The rapid pace of global urbanization has exacerbated the urban wind-heat environment, posing a severe threat to public health and sustainable urban development. This study explores the aerodynamic transport characteristics of bioaerosols in a local urban area of Beijing following an accidental bioaerosol release. By coupling the Weather Research and Forecasting (WRF) model with a Computational Fluid Dynamics (CFD) model, the research accounts for the temporality of urban airflow and atmospheric stability. A dose–response model was employed to assess the exposure risks to Beijing Institute of Technology personnel. The findings reveal substantial differences in flow fields and bioaerosol dispersion under varying atmospheric stability: the infection area ratio was 42.19% under unstable conditions and 37.5% under stable conditions. Infection risk was highest near the release source, decreasing with distance. Under the three stability conditions, the probability of infection is highest near the release source and decreases with increasing distance. Contaminants propagate more rapidly under unstable conditions, while stable conditions have a higher concentration of high-risk areas. Gender-based analysis indicated a higher infection probability for males due to elevated inhalation rates. This study elucidates the critical role of atmospheric stability in bioaerosol dispersion and provides a robust scientific foundation for biosafety planning, including early warning, mitigation, and emergency evacuation strategies.

1. Introduction

Amid the accelerating pace of urbanization, cities—major centers of human activity—face escalating biosafety challenges with potentially severe and far-reaching consequences. Historical incidents include the 1984 methyl isocyanate leak in Bhopal, India [1]; the 2015 Middle East Respiratory Syndrome (MERS) outbreak in South Korea [2]; and the 2019 brucellosis outbreak at a biopharmaceutical facility in Lanzhou, China [3]. The high population density and intricate infrastructure of urban areas amplify the complexity and difficulty of mitigating such incidents. In this context, a systematic investigation into the transport mechanisms of bioaerosols and the associated infection risks within urban environments is urgently needed. Such research is critical not only for safeguarding public health but also for enhancing the resilience and sustainable development of cities. By addressing these challenges, this work contributes to the scientific foundation necessary for effective biosafety management, emergency response, and urban sustainability on a global scale.

Airflow plays a pivotal role in determining the aerodynamic transport characteristics of biological aerosols [4]. To accurately capture the complex wind dynamics within urban environments, researchers typically rely on two primary modeling approaches: computational fluid dynamics (CFD) simulations and meteorological models, such as the Weather Research and Forecasting (WRF) model, for predicting flow fields [5]. For example, Liu et al. [6] developed a full-scale urban model incorporating detailed building geometries and employed Reynolds-averaged Navier–Stokes (RANS) simulations to investigate community-level wind patterns. Their results demonstrated that full-scale models provide greater accuracy in predicting wind speeds compared to micro-scale approaches. Similarly, Ramponi et al. [7] utilized the RANS method to explore the influence of wind direction and urban street layouts on ventilation efficiency in surrounding areas. Huang et al. [8] further enhanced spatial resolution by nesting a WRF model layer with two CFD models, achieving precise downscaling of wind flow assessments at the pedestrian level. These advancements underscore the importance of integrating meteorological and CFD modeling techniques to improve the characterization of urban airflow and its implications for bioaerosol transport.

Furthermore, Kang et al. [9] demonstrated the potential of integrating a CFD model with a tree drag parameterization scheme to assess the impact of urban vegetation on pedestrian wind comfort. Their findings revealed that strategically placed trees can significantly enhance comfort under specific wind conditions. While CFD simulations provide high-resolution predictions of airflow distributions around buildings, they are computationally demanding [10]. In contrast, meteorological models excel in large-scale airflow predictions, effectively capturing the spatial distribution of atmospheric flow and wind speed, albeit with lower spatial resolution [11]. To leverage the strengths of both approaches, coupling WRF and CFD models has emerged as a promising strategy for achieving a more precise representation of urban flow fields. He et al. [5] investigates the impact of sea–land breeze (SLB) on wind speed, pollutant dispersion, and flow patterns in coastal cities using a WRF–CFD coupled model. It finds that SLB collisions cause vortex structures and pollutant accumulation in streets, with the lowest pollutant concentrations occurring during the evening peak traffic period due to higher wind speeds. Miao et al. [12] used a coupled WRF-OpenFOAM model to examine airflow and pollutant dispersion in Beijing, validating the CFD model with wind-tunnel data. The results showed that local circulations, influenced by buildings, caused complex pollutant dispersion patterns, with higher concentrations found in specific areas due to weaker vertical flow or vortex formations, highlighting the model’s potential for studying urban environments. However, despite these advancements, the influence of atmospheric stability on urban wind environments remains underexplored.

Atmospheric stability refers to the tendency of the atmosphere to resist or promote vertical air motion. It is determined by the temperature profile of the atmosphere, specifically the rate at which temperature decreases with altitude, known as the lapse rate. When the atmosphere is stable, air parcels that are displaced vertically will return to their original position, while in an unstable atmosphere, displaced air parcels will continue to rise or sink [13]. Guo et al. [14] used CFD simulations to study the impact of atmospheric stability on flow and pollutant dispersion in urban street canyons, revealing that stability significantly affects vortex intensity, plume dilution, and pollutant concentration [14]. Wang et al. [15] addressed the limitations of the neutral atmospheric boundary assumption in air ventilation assessment (AVA) by comparing field measurements, wind-tunnel tests, and large-eddy simulations (LES), demonstrating that unstable conditions improve ventilation performance, while the neutral assumption underestimates pedestrian-level velocities and weakens the correlation between street orientations and ventilation. Incorporating atmospheric stability into coupled WRF–CFD frameworks could refine our understanding of urban airflow dynamics and significantly enhance the accuracy of bioaerosol transport modeling in complex urban settings.

Current research predominantly focuses on the dispersion characteristics of bioaerosols in enclosed or semi-enclosed environments. For instance, Cao et al. [16] explored aerosol diffusion within hospital wards under three distinct ventilation modes, emphasizing the pivotal role of ventilation design in mitigating indoor cross-infection risks. Similarly, Esther et al. [17] conducted a detailed evaluation of indoor air quality, pollutant levels, and the relationship between airborne respiratory viruses and biological samples in university classrooms in Spain. Their findings revealed significant variations in ventilation efficiency and pollutant concentrations, highlighting the critical influence of airflow management on indoor air quality. Passo et al. [18] investigated PM_2.5 and bioaerosol concentrations in subway metro stations, utilizing the Multipath Particle Dosimetry Model to assess their deposition in human respiratory tracts and the associated health risks. These studies collectively underscore the importance of understanding aerosol behavior in confined environments to inform effective ventilation strategies and minimize health risks associated with bioaerosol exposure. Incorporating time-dependent meteorological factors, particularly shifts in atmospheric stability, is crucial for achieving more accurate predictions of bioaerosol transport and associated health risks.

Health risk assessments are integral to guiding emergency evacuation strategies and implementing preventive measures, especially in high-occupancy environments such as hospitals, laboratories, and schools [19,20]. However, while substantial research has addressed bioaerosol dispersion in enclosed spaces and broader urban contexts [21,22], comparatively little attention has been directed toward school environments. Schools, characterized by high population densities, unique spatial layouts, and close interpersonal interactions, present conditions conducive to the rapid spread of pathogens during epidemics or biosafety incidents [23,24,25]. Addressing this gap requires targeted research to refine risk assessments and enhance biosafety preparedness in educational settings, thereby bolstering resilience against sudden biosafety threats.

Atmospheric stability has been widely recognized as a critical factor influencing airflow dynamics, which in turn shape the spatiotemporal distribution of bioaerosols and associated infection risks [26,27]. Additionally, previous studies have highlighted the importance of individual respiratory rates and spatial positioning in determining infection probability [28,29]. Based on this, an innovative framework was proposed in this study to investigate airflow patterns and bioaerosol transport on a university campus in Beijing under varying atmospheric stability conditions. By integrating the WRF model with a CFD model, the study captures the detailed aerodynamic transport characteristics of bioaerosols following an accidental release. A dose–response model (DPM) is further applied to quantify infection risks among exposed individuals. These findings deepen our understanding of bioaerosol transmission dynamics in urban settings and provide actionable insights for optimizing biosafety strategies. This research not only enhances public health preparedness but also supports the sustainable and resilient development of urban environments in the face of emerging biosafety challenges.

2. Method

2.1. Study Area

This study site is located at the Beijing Institute of Technology in the Zhongguancun area of Beijing, covering 108 hectares with a plot ratio of 3.01 (Figure 1a). The building outlines within the study area were extracted and digitized using Google Maps and ArcGIS 10.7 software [30], as shown in Figure 1b,c. The campus contains a dense building complex measuring roughly 1200 m in length, 1050 m in width, and 138 m in maximum height. In accordance with established guidelines for computational fluid dynamics (CFD) simulations, the computational domain was designed to ensure accurate airflow predictions. Specifically, the downstream boundary was extended 20 H from the tallest building, where H represents the maximum building height, while the lateral and upstream boundaries were positioned 5 H from the building complex [31]. The configuration of the computational domain is detailed in Figure 2. This setup ensures the reliable simulation of urban airflow patterns and bioaerosol dispersion within the study area.

2.2. WRF Model Setup

WRF model data were obtained from the National Center for Atmospheric Research (NCAR) in the United States. The meteorological fields generated by the WRF-ARW model served as boundary conditions for subsequent CFD simulations. A three-level nested grid captured scales ranging from the broader North China region down to the Beijing Institute of Technology (BIT), using horizontal resolutions of 9 km, 3 km, and 1 km. A total of 65 vertical layers were included, with 26 situated within the first kilometer above ground to capture the intricate interactions between the urban environment and atmospheric processes within the planetary boundary layer. The model was run from 00:00 on 22 June to 24:00 on 23 June 2022, with outputs recorded hourly. This high-resolution dataset supplied essential atmospheric parameters for CFD calculations, providing a more robust depiction of urban airflow and bioaerosol dispersion.

This study employs the Rapid Radiative Transfer Model (RRTM) schemes for both longwave and shortwave radiation [32]. The WDM-6 scheme available in the WRF model is used for microphysical processes, encompassing six distinct hydrophysical processes. Land surface processes are represented by the unified Noah land surface model, enabling accurate simulations of heat, water vapor, and momentum exchanges at the surface [33]. To capture planetary boundary layer turbulence and near-surface dynamics, the Yale University Scheme (YSU) is adopted. The YSU approach integrates the Mellor–Yamada–Janjic (MYJ) boundary layer model with the MYJ Monin–Obukhov scheme, ensuring a robust representation of boundary layer characteristics [34]. The details are shown in

Table 1. Runtime, domain configuration, and physical field settings used by WRF.

	Settings	Explanation
Running time	22 June 2022 00:00–23 June 2022 24:00
Grid size/km	9 × 9, 3 × 3, 1 × 1
Microphysical scheme	WDM-6	It simulates cloud water, rain, snow, ice, graupel, and hail. It includes advanced treatment for the processes of cloud formation, precipitation, and interactions between hydrometeors in varying atmospheric conditions.
Radiation scheme	RRTM	It provides efficient and accurate calculations of the radiative fluxes, especially in cloudy and clear sky conditions, by solving the radiative transfer equation.
PBL scheme	YSU	It is designed to represent turbulent mixing within the planetary boundary layer (PBL) and accounts for surface fluxes, turbulent exchanges, and stable/unstable atmospheric conditions, improving the simulation of wind, temperature, and moisture profiles near the surface.

.

The coupling mechanism between the WRF and CFD models is essential for accurately simulating aerosol dispersion and bioaerosol transport in urban environments. In this study, the WRF model provides meteorological inputs, such as wind profiles, temperature, and atmospheric stability, which serve as boundary conditions for the CFD model. Specifically, the RRTM (Rapid Radiative Transfer Model) in WRF calculates longwave radiation transfer, influencing surface thermal fluxes and atmospheric stability, which are crucial for the heat exchange process in the CFD model. The WDM-6 microphysics scheme in WRF provides precipitation and cloud data, which impact the atmosphere’s moisture content and bioaerosols’ dispersion, particularly under unstable thermal conditions, as well as the PBL scheme in WRF models turbulence within the planetary boundary layer, providing wind profiles and mixing height information that is directly transferred to the CFD model. The horizontal resolutions used in WRF are optimized to capture large-scale atmospheric features, ensuring that accurate wind and temperature profiles are generated for the CFD model. These inputs, such as wind speed, turbulence, and temperature, are then applied as boundary conditions in the CFD model, which simulates the movement and dispersion of aerosols within the urban environment.

2.3. CFD Settings

In this study, CFD simulations were conducted to explore airflow characteristics and bioaerosol transport. Previous research has demonstrated that RANS models can achieve close agreement with experimental measurements [35]. A steady RANS model was used to simulate the urban wind field, and subsequently, the DPM model was employed to conduct transient simulations of bioaerosol particle dispersion. Therefore, the realizable k − ε model within the RANS framework was selected to represent turbulence, with a standard wall function handling near wall effects. The governing equations were discretized using a second-order upwind scheme to enhance accuracy and numerical stability [36]. The specific control equations for the realizable $k-\epsilon$ model are as follows:

$${\displaystyle \frac{\mathsf{\partial }\left(\rho \phi \right)}{\mathsf{\partial }t}}+\nabla \cdot \left(\rho \phi {\displaystyle \frac{dy}{dx}}\stackrel{\rightharpoonup }{V}\right)=\nabla \cdot \left({\mathsf{\Gamma }}_{\phi }\nabla \phi \right)+S_{\phi }$$

(1)

where $\rho$ represents fluid density, $\stackrel{\rightharpoonup }{V}$ is the velocity vector, $\phi$ refers to each velocity component, ${\mathsf{\Gamma }}_{\phi }$ is the effective dispersion coefficient for $\nabla \phi$, and $S_{\phi }$ denotes the source term.

At the inlet boundary, a velocity inlet is applied, with symmetrical conditions at the sides and top, and the wind speed data derived from the WRF follow a power law distribution. The inlet wind speed is given by the following equation:

$$U\left(H\right)=U_{\mathit{ref}}{(H/H_{ref})}^{\alpha }$$

(2)

where U_ref represents the annual mean wind speed at the reference height H_ref, and this value is obtained based on the WRF fitting. Considering urban morphology and building density, the value of α was set to 0.3.

To evaluate how atmospheric vertical stability influences the flow field and bioaerosol dispersion in a dense urban setting, a dimensionless Richardson number (Ri) was introduced. The associated equations are given by:

$$Ri=-{\displaystyle \frac{Gr}{R{e}^2}}=-{\displaystyle \frac{gh\Delta T}{\overline{U_{\mathit{ref}}^2{T}_a}}}$$

(3)

where Gr and Re are the Grashof and Reynolds number, reflecting thermal buoyancy and wind strength in the urban boundary layer, respectively. The parameter h indicates the reference height, while U_ref represents the mean wind speed measured at h = 10 m, g is the gravitational acceleration and $\Delta T$ is the temperature difference between the ground surface and the ambient air within the urban environment. Furthermore, the stability of fluid flow can be quantified using the Richardson number (Ri), which provides insight into the relationship between buoyancy and turbulence. Specifically, when Ri < 0, the flow is unstable and characterized by high turbulent kinetic energy. In this case, vertical motion is enhanced, which promotes the dispersion of aerosols. When Ri = 0, the fluid is in a critical state, where instability may occur, and when Ri > 0, the flow is stable, with low turbulent kinetic energy, leading to reduced mixing and aerosol dispersion.

In this study, three atmospheric stability conditions—unstable, stable, and neutral—were considered to evaluate their effects on bioaerosol transport and dispersion. Unstable conditions occur when the surface air is warmer than the air above, promoting vertical mixing and strong turbulence, which enhances aerosol dispersion and results in lower surface concentrations and greater atmospheric spread. Stable conditions, on the other hand, are characterized by cooler surface air, which inhibits vertical mixing and reduces turbulence, leading to the trapping of pollutants and bioaerosols near the surface and decreased dispersion. Neutral conditions, where the temperature gradient is minimal, represent a scenario of limited mixing, resulting in intermediate dispersion between the extremes of unstable and stable conditions. These conditions were incorporated into the model to assess their influence on the migration and infection risk of bioaerosols in an urban setting.

Since the temperature does not vary much with time, a linear approximation is applied to the inlet boundary temperature during the WRF simulation period. Stable, unstable and neutral thermal conditions are assigned to 8:00, 14:00, and 18:00 on 23 June, respectively, with further meteorological details provided in

Table 2. Meteorological data at different times.

WRF Data	Unit	8:00	14:00	18:00
temperature	K	−0.0074y + 309.92	0.003443y + 305.2	−0.0009y + 300.09
10 m velocity	m/s	2.83	3.19	3.55

.

The discrete phase model employs a Lagrangian discrete stochastic particle wandering approach to track bioaerosol diffusion, governed by the following equations:

$${\displaystyle \frac{d\stackrel{\rightharpoonup }{u_p}}{dt}}=F_D\left(\stackrel{\rightharpoonup }{u}-\stackrel{\rightharpoonup }{u_p}\right)+{\displaystyle \frac{\stackrel{\rightharpoonup }{g}\left({\rho }_p-\rho \right)}{{\rho }_p}}+\stackrel{\rightharpoonup }{F}$$

(4)

where $\stackrel{\rightharpoonup }{u}$ and $\stackrel{\rightharpoonup }{u_p}$ represent the airflow and bioaerosol velocities, respectively; $\stackrel{\rightharpoonup }{F_D}$ indicates the drag force acting on the bioaerosol particles; $\rho$ and ${\rho }_P$ are the densities of airflow and bioaerosol, respectively; $\stackrel{\rightharpoonup }{g}$ is gravitational acceleration; and $\stackrel{\rightharpoonup }{F}$ denotes additional forces exerted on the bioaerosol, which include the thermophoretic force and the Saffman lift force.

The initial conditions of this study are based on a converged mean flow field (with residuals up to 10⁻⁶) generated by the RANS model and subsequently injected with particles for non-stationary calculations. The DPM model used has been validated in previous studies [37,38,39]. The bioaerosol release was by vertical surface release, with a particle size of 5 × 10⁻⁷ m, a density of 1000 kg/m³, and a mass flow rate of 3.392 × 10⁻⁶ kg/s [36]. The boundary conditions at the entrance and exit of the computational domain were set to escape, while the walls and floor were set to capture. The non-constant calculation was performed with a time step of 1 s, and each time step was iterated 20 times, giving a total time of 900 s for the release of the particles. By employing the above method, the accuracy of the turbulence simulation and the stability of the numerical calculations were ensured, and a reliable theoretical basis for the diffusion simulation of bioaerosols was provided.

2.4. Dose–Response Model

DPM is widely used to assess the risk of infection associated with bioaerosols [34]. For the study, z = 1.6 m was selected as the breathing height of pedestrians. The study area was divided equally into 8 × 8 rectangular regions for ease of analysis. In this study, the infection risk of different populations was analyzed using staphylococcus aureus bioaerosols [40]. The infection probability at varying pedestrian heights was determined using an exponential dose–response model [41], given by the following equations:

$$p=1-\exp (-d\bullet k)$$

(5)

where p denotes the probability of infection by bioaerosols; d is the inhaled bioaerosol dose for various populations, CFU; and the parameter k is pathogen-specific and has been identified in the literature [42] as the infectious dose of Staphylococcus aureus, with k = 8.05 × 10⁻⁸, CFU⁻¹. The dose, d, is computed according to the following equation:

$$d={\displaystyle \underset{t_0}{\overset{t}{\int }}{q}_{in}E{C}_i(\tau )d\tau }$$

(6)

where q_in denotes the inhalation volume, m³/h; t − t₀ is the residence time, h; and EC_i represents the bioaerosol concentration in each cell, CFU/m³. Further details of the dose–response model parameters are provided in

Table 3. Residence time settings for different populations.

Parameters	Unit	Numeric
Retention time	min	5, 10, 15, and 20
		adult
		male	female
Inhalation rate	m3/h	0.777	0.616

[43].

3. Result and Discussion

3.1. Grid-Independent Verification

Three different grid resolutions were established in this study to ensure the reliability of the numerical simulation results. Among them, the grid counts for fine (minimum size 0.5 m), medium (minimum size 1 m), and coarse (minimum size 2 m) resolutions are 14, 11, and 7.9 million, respectively. Figure 3 shows the relative velocity error at the location x = 600 m, y = 600 m. The results indicate that the relative velocity error decreases as the number of grids increases. When the minimum grid size decreases from 1 m to 0.5 m, the relative errors for medium and fine exhibit reduced fluctuations, remaining below 5% with increasing vertical height. Considering the balance between computational accuracy and efficiency, a medium-resolution grid with a minimum grid size of 1 m is ultimately selected to simulate the flow field distribution in urban areas.

3.2. Validation of WRF Calculations

To verify the accuracy of the WRF model, this study adopts a comparative and analytical approach by comparing the meteorological data observed in the Zhongguancun area of Beijing, with the simulation results produced by the WRF model. The observed data were obtained from the Beijing Meteorological Monitoring Station (https://xihe-energy.com/, accessed on 10 July 2024) and included key meteorological variables, such as temperature, wind speed, humidity, and barometric pressure. The simulation domain covers Beijing and its surrounding regions, with a 1 km × 1 km resolution grid used to enhance the capture of meteorological characteristics within the urban environment. The WRF model generates meteorological data across multiple temporal periods, and the simulation outcomes, including temperature and wind speed, are systematically compared with the observed data on an hourly basis.

As demonstrated in Figure 4, the temporal variations in temperature and wind speed at a height of 10 m in the study area are compared between the observed values and those simulated by the WRF model. The prediction of wind speed, as depicted in Figure 4a, reveals some differences between the WRF model simulations and the observed data. While these differences exist, the WRF model still provides valuable wind speed data for analysis. Therefore, wind speed data are available. Furthermore, the time-dependent trend of the temperature in the study area is consistent with observed data, thereby demonstrating the WRF model’s enhanced accuracy in temperature simulation. In conclusion, the WRF model has been shown to accurately predict meteorological parameters, including wind speed and temperature, in the urban environment.

3.3. The Relative Velocity Distribution at Different Stabilization Levels

The relative velocity distribution at pedestrian height (z = 1.6 m) under varying thermal conditions is depicted in Figure 5. It was determined that atmospheric stability at different times exerted a significant influence on airflow distribution. It was observed that wind speed is reduced under stable thermal conditions [44]. Under these conditions, airflow around the building is weak, the airflow on the leeward side is slightly stagnant, and no apparent vortex phenomena are present. Conversely, under unstable thermal conditions, ground temperature rises, turbulence increases significantly, wind speed increases due to thermal buoyancy, and significant vortices appear around the building. In neutral conditions, as solar radiation diminishes and temperatures decrease, the atmosphere gradually regains stability, turbulence decreases, wind speed falls, the flow field stabilizes, airflow around the building becomes more uniform, and wind speed in the leeward zone exhibits less variation. The results above demonstrate that changes in atmospheric stability directly influence turbulence and airflow mixing, thereby determining the structure of the flow field around the building. Under stable conditions, airflow tends to be laminar, whereas, under unstable conditions, vertical and horizontal mixing is enhanced, resulting in a complex turbulent structure [45].

As illustrated in Figure 6, the wind speed distribution along the vertical section is influenced by temporal variations and atmospheric stability, which impact the flow field. During periods of instability, vertical airflow variations are minimal, wind speed remains relatively consistent, and the flow predominantly exhibits horizontal characteristics. Airflow above the building is restricted, updrafts are weakened, and the flow field is relatively smooth. Conversely, under unstable conditions, as atmospheric instability and intense solar radiation increase ground temperature, the airflow exhibits pronounced vertical movement [46]. Vertical wind speed around the building rises significantly, and upward airflow movement is strengthened, especially on the upwind side. Airflow disturbances are intensified, and the vortex structure becomes more complex. The distribution of vertical wind speed becomes uneven, with the development of distinct regions of updraft and downdraft. At 18:00, with the reduction in solar radiation and the decline in ground temperature, the thermal drive of the airflow is diminished, leading to a gradual return to a steady state in vertical wind speed. Consequently, the airflow around the building stabilizes, resulting in reduced vortices and turbulence. The disparity in wind speeds along the vertical section diminishes, and the airflow progressively restores a laminar flow state, leading to a tendency toward a uniform flow field [47].

The ensuing analysis elucidates the mechanism by which atmospheric stability affects the flow field. Under stable conditions, airflow is confined to the lower layers, exhibiting reduced wind speeds and fewer vortices. Conversely, under unstable thermal conditions, thermodynamic effects increase airflow turbulence, leading to higher wind speeds and augmented disturbances around the building.

3.4. Spatial and Temporal Distribution of Bioaerosol Dispersion at Different Stabilization Levels

Figure 7 illustrates the diffusion of bioaerosols at t = 300 s and t = 900 s under unstable, stable, and neutral thermal conditions. The effect of airflow on aerosol diffusion exhibits significant variations over time and across different thermal conditions. First, under stable thermal conditions, airflow laminarity is maintained, aerosol diffusion is slower, concentration distribution on the horizontal plane tends to be uniform, and diffusion distance is shorter. In contrast, under unstable thermal conditions, airflow disturbances are enhanced, leading to a significant increase in aerosol diffusion distance. Significantly at t = 300 s, under unstable conditions, airflow significantly enhances both vertical and horizontal aerosol diffusion, causing aerosols to disperse over a wider area and, thus, extending their propagation distance. In neutral thermal conditions, aerosol diffusion performance is intermediate between stable and unstable conditions due to the lack of significant upward or downward airflow tendencies. Aerosol diffusion distance increased compared to stable conditions but remained smaller than in unstable conditions. Therefore, it can be concluded that an unstable atmosphere is a significant factor contributing to the diffusion of urban bioaerosols, indicating that thermal environment stability plays a key role in aerosol dispersion [48].

Figure 8 further illustrates the concentration distribution of bioaerosols in six planes along the diffusion path from the release source under the three thermal conditions, with panels a–g showing the distribution based on proximity to the release source, respectively. The results indicate that aerosol concentration exhibits significant spatial variability with changes in diffusion direction. Under stable thermal conditions, aerosol concentration is higher near the release source, and as the diffusion path extends, concentration gradually decreases, with changes being smoother and exhibiting more minor. In contrast, under unstable thermal conditions, aerosol concentration distribution fluctuates more and is significantly higher in the downstream region of diffusion than under stable conditions. This phenomenon can be attributed to the strong perturbation of airflow in the unstable thermal environment, especially the promotion of aerosol diffusion by updrafts, which results in the broader distribution of aerosols in space [49]. The findings align with the general understanding that stable conditions lead to more concentrated, localized pollutant distribution, as in the case of particulate matter in winter months. However, bioaerosols, due to their biological properties and the turbulence-induced mixing in unstable conditions, behave differently from traditional pollutants. The broader diffusion of bioaerosols in unstable conditions highlights the potential for a wider spread of infection, making it a key consideration in scenarios involving sudden bioaerosol leaks in urban settings.

Overall, changes in thermal conditions and building layout have a significant impact on aerosol dispersion processes. Under stabilized thermal conditions, the dispersion of aerosols was limited, and the concentration distribution was relatively uniform. Near source concentrations were significantly higher under stabilized conditions than under unstable conditions, while under unstable thermal conditions, the diffusion range of aerosols increased significantly, especially in the downstream area, where the concentration was higher and fluctuated more.

3.5. Exposure Risk Assessment

To assess and analyze the infection risk of individuals exposed to bioaerosols in urban environments, this study divided the pedestrian-level area at the height of H = 1.6 m into equal subareas (175 m × 175 m) and calculated the infection probability within these subareas under various atmospheric conditions. Based on the dose–response model, the infection risk under different exposure conditions was further evaluated by incorporating wind direction with aerosol concentration distribution.

As illustrated in Figure 9, the probability of infection for individuals exposed to bioaerosols in urban environments varies across various atmospheric stability levels. The data demonstrate that the exposure risk is highest near the source in all three scenarios, reaching up to 4.71% in unstable atmospheres and 1.967% in stable atmospheres. Except for the relatively high exposure risk near the release source, the probability of infection in each area exhibited a decreasing trend from upstream to downstream along the wind direction. A comparison of the infected area ratio to the total study area reveals ratios of 42.19% under unstable conditions and 37.5% under stable conditions. Infected areas under unstable conditions were found to be greater than those under stable conditions. This phenomenon can be attributed to the rapid vertical transport and wider spatial dispersion of pollutants under unstable conditions caused by strong thermal perturbations and inhomogeneity in vertical airflow, resulting in a higher risk of infection at distant sources than in stable conditions. Conversely, under stable conditions, airflow is characterized by smoothness, and contaminants are primarily concentrated in specific areas with a slower diffusion process, resulting in a more extensive high-risk area [50]. Consequently, contaminants propagate more rapidly under unstable conditions, while stabilized conditions have a higher concentration of high-risk areas. This highlights the need for dynamic evacuation strategies that adapt to real-time atmospheric conditions. In unstable conditions, broader evacuation zones should be established to account for the larger dispersion area, while in stable conditions, focused evacuations in areas downwind of the release point are critical, and crowds near the source should be promptly evacuated to either side of the vertical wind direction. Additionally, using air filtration systems in regions of high aerosol concentration, particularly in stable conditions, can significantly reduce exposure risks by trapping harmful particles.

Adult males face a significantly higher risk of infection than females for the same exposure duration due to higher inhalation rates [51]. This elevated risk is attributable to excellent lung ventilation and higher respiratory rates, which are characteristic of males. Consequently, males inhale a greater number of pollutants or pathogens under the same environmental conditions, thereby increasing their risk of infection. Specifically, the inspiratory rate of males is 20–30% higher than that of females, resulting in the inhalation of a more significant of pollutants during the same exposure duration, thereby increasing the probability of infection. This is consistent with the findings of existing studies [43]. Furthermore, larger body sizes and wider airways in males exacerbate the differences in inhalation rates. Finally, gender differences must be considered when developing evacuation strategies. Our study highlights that there may be differences in the mobility and response time between males and females during emergencies, which could impact the overall efficiency of the evacuation process. It is important to design gender-sensitive evacuation plans that account for these differences, ensuring that all individuals can evacuate safely and promptly. For instance, prioritizing the evacuation of vulnerable groups, providing accessible routes, and ensuring clear communication tailored to the needs of different demographics can enhance the overall effectiveness of the emergency response. Consequently, when formulating urban emergency evacuation strategies, it is imperative to consider not only atmospheric stability but also gender-related differences and geographical distance. This is crucial to minimize the risk of personnel exposure in environments with high concentrations of bioaerosols [52].

4. Conclusions

The spread of pathogens in the form of bioaerosols has emerged as a critical concern due to their rapid dispersal capabilities and potential hazards, posing significant challenges to urban biosecurity. This study systematically investigates the mechanisms driving bioaerosol dispersion under different atmospheric stability conditions, focusing on flow field evolution, aerosol diffusion, and infection risk assessment. The study’s primary findings are as follows:

(1)

The WRF model showed strong agreement with measured data in simulating temperature, particularly in capturing diurnal and nocturnal temperature variations, with high consistency observed.

(2)

The structure of the flow field exhibits significant variations across different thermal conditions. Under stable conditions, the airflow remains predominantly laminar, with minimal wind speeds and limited vertical exchange. In contrast, under unstable conditions, thermal disturbances are amplified, promoting enhanced vertical and horizontal mixing of the airflow.

(3)

Thermal stability has a significant impact on both the spatial extent and concentration variability of aerosol dispersion. Contaminants propagate more rapidly under unstable conditions, while stabilized conditions have a higher concentration of high-risk areas. Under stable thermal conditions, the range of aerosol diffusion is limited, with a relatively uniform concentration distribution, and aerosol concentration remains predominantly localized near the source. In contrast, under unstable thermal conditions, the increased turbulence of the airflow significantly extends the aerosol diffusion distance. As a result, the concentration in the downstream areas rises and exhibits more significant fluctuations.

(4)

The probability of infection exhibited a positive correlation with proximity to the release source under all three thermal conditions. Under unstable thermal conditions, the infection probability decreased more rapidly with distance, suggesting an increased potential for long-distance transmission. In contrast, under stable thermal conditions, the infection risk remained concentrated near the source, with a relatively limited range of spread. Additionally, gender-based analysis revealed that adult males had a significantly higher probability of infection than females, particularly under unstable conditions, due to their higher inhalation rates.

In summary, the present study reveals the key effects of thermal conditions on aerosol dispersion and infection risk by quantitatively analyzing the flow field and infection probability under varying atmospheric stability conditions. This scientific foundation can inform urban biosecurity prevention and control strategies. Subsequent studies may utilize experimental validation and parameter optimization to explore aerosol propagation mechanisms and risk assessment further, thereby establishing the foundation for the development of more accurate biosafety emergency response strategies.