Spatiotemporal Analysis of Ventilation Efficiency in Single-Span Plastic Greenhouses in Hot-Humid Regions of China: Using Validated CFD Modeling

Abstract

To characterize the spatiotemporal distribution of temperature and airflow in single-span plastic-film greenhouses, we coupled field experiments with three-dimensional computational fluid dynamics (CFD) simulations in a warm–temperate region of China. Model reliability and validity were evaluated against field measurements. The average and maximum relative errors between simulated and measured values were 6% and 9%, respectively. Significant spatial heterogeneity in both temperature and airflow was observed. Vertically, temperature rose with height; horizontally, it declined from the center toward the sidewalls. Under prevailing meteorological conditions, the daily maxima occurred at distinct elevations above the fan-vent outlets. Airflow was most vigorous near the vents, whereas extensive stagnant zones aloft reduced overall ventilation efficiency. These findings provide a quantitative basis for designing single-span plastic film greenhouses in China’s hot–humid regions, informing ventilation improvements, and guiding future optimization efforts.

1. Introduction

With the development of the plastics industry, plastic film greenhouses have gained widespread application globally due to their advantages of simple construction, convenient use, and low investment costs, and have become the primary type of greenhouse in China [1,2,3,4]. Airflow regulation within greenhouses is critical for plant growth, and efficient production relies on an appropriate microclimate environment [5,6]. In the relatively enclosed greenhouse environment, airflow is a key ecological factor influencing crop growth and development, directly affecting their physiological processes. Issues such as poor airflow circulation and uneven temperature distribution can significantly impact plant growth, leading to reduced air quality and increased disease risk [7,8]. Poor airflow circulation weakens plants’ photosynthetic capacity and suppresses their immune function, thereby reducing their resistance to plant pathogens [9]; uneven temperature distribution directly disrupts normal plant growth [10]. Improper greenhouse design or internal obstacles can restrict air exchange within the plant canopy, hinder humidity regulation, and create a high-humidity environment that is prone to fungal diseases [11]. Optimizing ventilation is beneficial for crop growth and development. Uneven temperature distribution can cause localized overheating or overcooling of plants [12]; high temperatures increase transpiration rates, leading to imbalances in water and nutrient availability, making plants more susceptible to stress [13]; while low temperatures inhibit metabolism and growth, increasing plant susceptibility to diseases [14,15]. In summary, poor airflow and uneven temperature distribution hinder air exchange and temperature/humidity regulation, increase disease risk, weaken photosynthesis and immune function, and ultimately impair plant health. Optimizing airflow and temperature distribution in greenhouse design and management is therefore critical.

Computational Fluid Dynamics (CFD) is a powerful tool for simulating flow fields. It solves the governing equations of fluid flow numerically, combining fluid mechanics such as turbulence models and computational methods with computer technology, and can be used to analyze and simulate fluid flow and heat and mass transfer processes within greenhouses [16,17]. In recent years, CFD simulations have been widely applied in greenhouse environmental research, aiding in the study of airflow, temperature, and humidity distribution within greenhouses, thereby guiding environmental optimization [18]. By simulating temperature and humidity fields in enclosed spaces using CFD, the efficiency of ventilation systems can be assessed, airflow organization optimized, and optimal design solutions determined [19]. Research examples include using CFD for temperature uniformity testing to determine appropriate airflow rates [20]; employing orifice plate airflow optimization design to improve airflow uniformity and avoid discomfort, analyzing the effects of different ventilation methods, shading, window opening, and fan layout on greenhouse temperature fields, and establishing a greenhouse CFD model to simulate internal temperature and light environments [21]; and combining experiments with simulations to optimize ventilation configurations in plastic greenhouses to enhance cooling efficiency.

In response to the aforementioned issues of poor airflow circulation, uneven temperature distribution, and reduced ventilation efficiency in single-span plastic greenhouses under hot–humid climatic conditions, this study employs CFD technology with Chengdu as a representative case. By integrating local meteorological data, indoor–outdoor temperature differentials, evaporative cooling system effects, and solar radiation, the research aims to achieve the following: (1) establish a spatiotemporal model for temperature and airflow changes under typical sunny and cloudy conditions; (2) elucidate the dynamic variation patterns of temperature and airflow within the greenhouse at different times of day and under different solar radiation levels; and (3) propose quantitative optimization of structural and ventilation parameters to improve airflow uniformity, enhance temperature control efficiency, and mitigate microclimate-related plant stress. These objectives directly address the core microclimatic challenges identified above and provide a scientific basis for improved greenhouse design and management in hot–humid regions of China.

2. Materials and Methods

2.1. Construction of the Test Greenhouse Model

The test greenhouse uses forced ventilation with a wet pad-fan system, and a cooling pad is installed on the north side of the greenhouse in China’s humid and hot regions, a typical test greenhouse was selected. It is located at the Chengdu Campus of Sichuan Agricultural University in Wenjiang District, Chengdu City, Sichuan Province, China (30.70° N, 103.86° E), with its longitudinal axis oriented north–south, with a single-span arched structure, 8 m in length, with a span of 6 m, a shoulder width of 2.8 m, and a ridge height of 4.3 m. The greenhouse is covered with 0.15 mm thick polyethylene plastic film. Ventilated insect screens are installed on the east and west sides at heights of 0.5–1.6 m. A humidifier is installed at the same height on the north side of the greenhouse, and a fan is installed on the south exterior wall. The top view is shown in Figure 1. A 1:1 three-dimensional model of the experimental greenhouse was created in SolidWorks (2021SP3, USA) based on its three-dimensional dimensions. To reduce the computational load, the greenhouse model structure was simplified, as shown in Figure 2.

2.2. Computational Meshing

The computational domain consists of two parts: the indoor area and the outdoor area of the plastic greenhouse. The north direction of the greenhouse is defined as the positive direction of the Z-axis, the west direction as the positive direction of the X-axis, and the height direction as the positive direction of the Y-axis. The three-dimensional geometric modeling of the greenhouse is performed at a 1:1 scale. The greenhouse computational domain consists of an external fluid computational domain and an internal fluid computational domain. The height of the external fluid computational domain is set to three times the ridge height of the greenhouse (13 m).

A non-structured tetrahedral mesh is used to divide the geometry, with densification applied at the fan, evaporative cooling pad, insect screen, greenhouse roof, and ground. The maximum mesh size is 0.5 mm, with a total of 1.15 × 10⁵ mesh divisions. The model mesh is moderately densified to enhance prediction accuracy, with the mesh quality controlled according to the Equi Angle Skew standard, ensuring a quality factor of 0.75 or higher.

2.3. Simulation Equations

Based on the three fundamental control equations of mass conservation, momentum conservation, and energy conservation, the air inside the plastic film greenhouse is assumed to be a continuous, stable, incompressible Newtonian fluid, with the fluid flow field being entirely turbulent, satisfying the Boussinesq assumption. The radiative effects from internal heat transfer surfaces are neglected. The inner walls of the plastic film greenhouse, outlet fins, and other surfaces maintain uniform and constant temperatures (first-order boundary conditions). The door has good sealing properties, and air leakage is not considered [22,23,24].

2.3.1. Control Equations for Greenhouse Fluids

(1) 

Mass conservation equation:

$${\displaystyle \frac{\partial u}{\partial x}}+{\displaystyle \frac{\partial v}{\partial y}}+{\displaystyle \frac{\partial w}{\partial z}}=0$$

(1)

where u, v, and w are the components of velocity in the x, y, and z directions (m/s).

(2) 

Conservation of momentum in an inertial (non-accelerating) reference frame

$${\displaystyle \frac{\partial }{\partial \mathrm{t}}}\left(\mathsf{\rho }\overrightarrow{\mathrm{v}}\right)+\nabla \cdot \left(\mathsf{\rho }\overrightarrow{\mathrm{v}}\overrightarrow{\mathrm{v}}\right)=-\nabla \mathrm{p}+\nabla \cdot \left(\stackrel{̿}{\mathsf{\tau }}\right)+\mathsf{\rho }\overrightarrow{\mathrm{g}}+\overrightarrow{\mathrm{F}}$$

(2)

where $\mathrm{t}$ is time; $\overrightarrow{\mathrm{v}}$ is the velocity vector; $\mathsf{\rho }$ is the fluid density; $\mathrm{p}$ is the static pressure; $\stackrel{̿}{\mathsf{\tau }}$ is the stress tensor; and $\mathsf{\rho }\overrightarrow{\mathrm{g}}$ and $\overrightarrow{\mathrm{F}}$ are the gravitational force and external force, respectively.

(3) 

Conservation of energy

$${\displaystyle \frac{\partial }{\partial \mathrm{t}}}\left(\mathsf{\rho }{\mathrm{C}}_{\mathrm{a}}\mathrm{T}\right)+{\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{i}}}}\left(\mathsf{\rho }{\mathrm{u}}_{\mathrm{j}}{\mathrm{C}}_{\mathrm{a}}\mathrm{T}\right)-{\displaystyle \frac{\partial }{\partial {\mathrm{x}}_{\mathrm{j}}}}\left(\mathsf{\lambda }{\displaystyle \frac{\partial \mathrm{T}}{\partial {\mathrm{x}}_{\mathrm{j}}}}\right)={\mathrm{S}}_{\mathrm{T}}$$

(3)

2.3.2. Turbulence Modeling

Gas flow in plastic greenhouses typically exhibits turbulent conditions. Therefore, this study employs a high-precision k-ε turbulence model for simulation. Based on the model characteristics, the Realizable k-ε model is selected. The transport equations corresponding to the turbulent kinetic energy k and turbulent dissipation rate ε in the Realizable k-ε model are as follows:

$${\displaystyle \frac{\partial \mathsf{\rho }\mathrm{k}}{\partial \mathrm{t}}}+{\displaystyle \frac{\partial \mathsf{\rho }\mathrm{w}\mathrm{k}}{\partial \mathrm{z}}}={\displaystyle \frac{\partial }{\partial \mathrm{z}}}\left[\left(\mathsf{\mu }+{\displaystyle \frac{{\mathsf{\mu }}_{\mathrm{t}}}{{\mathsf{\sigma }}_{\mathrm{k}}}}\right){\displaystyle \frac{\partial \mathrm{k}}{\partial \mathrm{z}}}\right]+\mathrm{P}-\mathsf{\rho }\mathsf{\epsilon }$$

(4)

$${\displaystyle \frac{\partial \mathsf{\rho }\mathsf{\epsilon }}{\partial \mathrm{t}}}+{\displaystyle \frac{\partial \mathsf{\rho }\mathrm{w}\mathsf{\epsilon }}{\partial \mathrm{z}}}=-{\displaystyle \frac{\partial }{\partial \mathrm{z}}}\left[\left(\mathsf{\mu }+{\displaystyle \frac{{\mathsf{\mu }}_{\mathrm{t}}}{{\mathsf{\sigma }}_{\mathsf{\epsilon }}}}\right){\displaystyle \frac{\partial \mathsf{\epsilon }}{\partial \mathrm{z}}}\right]+\mathsf{\rho }{\mathrm{C}}_1^{\mathrm{*}}{\mathrm{E}}_{\mathrm{i}\mathrm{j}}-{\mathrm{C}}_2^{\mathrm{*}}{\displaystyle \frac{{\mathsf{\epsilon }}^2}{\mathrm{k}+\sqrt{\mathrm{v}\mathsf{\epsilon }}}}$$

(5)

where t is time; w is the time-averaged vertical flow velocity; z is the time-averaged velocity in the z direction; and C₁ and C₂ are empirical constants.

2.3.3. Radiation Model

During the summer, solar radiation is one of the key factors influencing the distribution of internal airflow and temperature fields. A discrete coordinate (DO) radiation model is selected based on radiation heat transfer conditions. During calculations, the FLUENT solar ray tracing method is employed. The primary parameters are as follows: E 103.86°, N 30.70°, and time zone + 8. The solar irradiation method is selected based on the outdoor meteorological conditions chosen for the computational model [25].

$$\nabla \left(I\left(r,s\right)\right)+\left(a,{\sigma }_s\right)I\left(r,s\right)={\displaystyle \frac{a{n}^2\sigma T^4}{ \pi }}+{\sigma }_s/{\int }_0^{4\pi }I\left(r,s\right)\varnothing \left(s,s’\right)d\Omega$$

(6)

where r, s, s′ is position vectors, direction vectors, and scattering direction vectors; a, σ_s, n is absorption coefficients, scattering coefficients, and refraction coefficients, 1/m; σ is the Stefan–Boltzmann constant, 5.67 × 10⁻⁸ W/(m²·K⁴); T is local temperature, K; I is solar radiation intensity, W/m²; Φ is phase function; and Ω is solid angle.

2.4. Boundary Conditions and Material Parameters

The internal flow field is bounded by the plastic greenhouse walls. Details of the boundary conditions used in the calculations are shown in

Table 1. Boundary condition settings in calculation.

Fluid: Air

Indoor temperature: Actual measured temperature

Gravitational acceleration: 9.81 m s−2

Air inlet: Speed (using actual measured values) Temperature (using actual measured values)

Air outlet: Free flow

. Outdoor wind speed and direction data were collected from a meteorological station near the greenhouse.

Table 2. Table of surface wind speed and direction.

Outdoor Airflow Parameters on Cloudy Days			Outdoor Airflow Parameters on a Sunny Day
Wind Direction	Angle of Drift (°)	Air Velocity (m s−1)	Wind Direction	Angle of Drift (°)	Air Velocity (m s−1)
N	12	1.1	N	338	0.5
E	110	1.2	SE	126	1.4
E	88	1.8	S	191	2
W	281	1.7	SE	127	1.7
NW	334	1.8	S	188	1.2

shows the wind direction and speed obtained from the nearby meteorological station. We define the low-speed stagnation zone as the region where airflow velocity is below 0.5 m/s. Based on the collected experimental data and existing experience, the structural parameters for different materials were set according to

Table 3. Parameter settings.

Material	Density/(kg m−3)	Specific Heat/(J kg−1 K−1)	Thermal Conductivity/(W m−1·K−1)	Refractive Index
Air	1.225	1006.43	0.0242	1
Plastic film	1120	1850	3	1
Soil	1605	920	1.17	1
Wall-brick	2600	750	1.04	1

.

2.5. Validation Experiment

Outdoor meteorological parameters such as air temperature and humidity, solar irradiance, etc., were measured in the greenhouse. Indoor air parameters were tested using Pt100 temperature sensors. After testing, it was found that there was little difference in the indoor air temperature between the north–south and east–west directions within the greenhouse. Therefore, the measurement points were placed at the edges of the greenhouse, and the temperature sensor data were recorded by a data logger. Solar radiation was measured using two irradiance sensors: one installed outdoors to measure outdoor irradiance; and another installed at the plant canopy height (1.5 m) indoors to measure indoor irradiance. Indoor and outdoor irradiance data are also recorded by a data logger. The experiment was conducted from 4 January to 10 May 2023, under typical sunny and cloudy conditions with natural ventilation, with all fans, evaporative coolers, and greenhouse doors closed. The test points were located at a height of 1.5 m, with one each inside and outside, as shown in Figure 1. Therefore, the data measured by the sensors reflected the temperature and wind speed of the air at the test point’s height. After data collection, a simulation experiment was conducted to verify the model’s simulation effectiveness. The depth of the soil has been set at 0.3 m, consistent with the greenhouse’s typical soil configuration in the region. Five typical weather conditions were selected for simulation, and the simulated values inside the greenhouse were compared with the measured values during the same period to calculate the differences. The HK–ZYGLZ-A-type photosynthetic active radiation meter (±1%) was used to measure light intensity; the DT-618 handheld anemometer (±3% ± 0.1 m/s) was used to measure wind speed; temperature, humidity, and CO₂ measurements were taken using the tes1370NDIR CO₂ Meter Temperature, Humidity, and CO₂ Measurement Instrument (humidity: ±4% RH (25 °C), CO₂: ±3% ±50 ppm); and temperature measurements and recordings were taken using the Greenhouse Baby (domestic) (±0.2 °C). Indoor and outdoor air temperature, ground temperature, and light intensity data are automatically collected by the cloud-based greenhouse intelligent monitoring system, with data recorded every 5 min. Outdoor wind speed and direction are collected by a meteorological station near the greenhouse. To primarily study the effects of light intensity and wind speed on greenhouse gas flow, environmental data from 8:00 AM to 7:00 PM are primarily collected for analysis.

2.6. Data Analysis

All collected data were entered into Excel 2019 for centralized processing, and Origin 2021 was used for image plotting.

3. Result and Analysis

3.1. Model Validation

For simulation verification under natural ventilation conditions, 7 April was selected as a representative cloudy day and 14 March as a typical sunny day. Boundary conditions for the model included outdoor temperature, wind speed, and ground temperature measured at five time points (9:00, 11:00, 13:00, 15:00, and 17:00). The simulated temperatures at these time points were compared with corresponding measured values, as presented in Figure 3. Under cloudy conditions (Figure 3a), the absolute error between simulated and measured temperatures ranged from 1.0 to 2.1 °C, with a maximum relative error of 8.89% and an average of 6.53%. On the sunny day (Figure 3b), absolute errors ranged from 1.1 to 1.7 °C, with a maximum relative error of 6.29% and an average of 5.56%. The close agreement between simulated and observed temperatures—both in magnitude and temporal trends—with relative errors consistently below 10%, demonstrates the effectiveness of the established CFD model.

The statistical evaluation of the model performance showed strong agreement between simulated and measured temperatures under both weather conditions (

Table 4. Model validation statistics.

Condition	Correlation Coefficient (R)	RMSE (°C)	MAE (°C)
Cloudy	0.98	0.84	0.66
Sunny	0.99	0.94	0.90

). For the cloudy day scenario, the correlation coefficient (R) was 0.984 (p < 0.01), with an RMSE of 0.84 °C and an MAE of 0.66 °C. For the sunny day scenario, the R value reached 0.993 (p < 0.01), with an RMSE of 0.94 °C and an MAE of 0.90 °C. These results indicate that the model captures the temporal variation patterns of greenhouse air temperature with high accuracy, and the residual errors fall well within the acceptable range for engineering applications.

3.2. Temperature Field Analysis

On a cloudy day (Figure 4a), at a height of 1.8 m above the ground, the indoor temperature at 9:00 a.m. is the lowest of the day, with an average minimum temperature of 17 °C. At 11:00, as outdoor temperature and solar radiation increase, the greenhouse exhibits a significant temperature gradient, with temperatures around the interior being notably higher than those near the ground. The average temperature near the film is 19.5 °C, while the average temperature near the ground is 17.2 °C, resulting in a maximum temperature difference of 6.7 °C within the greenhouse, indicating significant temperature variability. At 1:00 p.m. and 3:00 p.m., the indoor temperature in the greenhouse showed a gradient distribution, with temperatures gradually increasing from the outer perimeter toward the interior. The average indoor temperature was 24.8 °C, with relatively consistent temperature uniformity. At 5:00 p.m., as outdoor temperatures and solar radiation decreased, indoor temperatures began to drop gradually from the outer perimeter toward the interior.

On a sunny day (Figure 4b), at a height of 1.8 m above ground level, the indoor temperature at 9:00 a.m. is the lowest of the day, with an average minimum temperature of 10.7 °C. At 11:00 a.m., on a sunny day, the outdoor temperature and solar radiation increase rapidly, causing the indoor temperature to rise quickly. The average temperature near the top of the film is 19.5 °C, and the average temperature near the ground is 16.5 °C, with a maximum temperature difference of 3.2 °C inside the greenhouse. At 1:00 p.m. and 3:00 p.m., the film absorbs heat through heat exchange and then heats the surrounding air through convection, causing the indoor air temperature to rise, with relatively consistent temperature uniformity. The average temperature at the crop canopy level (Y = 0.5–1.6 m) inside the greenhouse was 23.8 °C, and the average temperature above the crop canopy (Y = 0.5–1.6 m) was 26.4 °C. However, at 17:00, the outdoor temperature and solar radiation decreased, but the indoor temperature remained around 21.0 °C.

The spatial distribution patterns of indoor temperatures vary across different time periods, with uneven temperature distribution. The internal temperatures of the greenhouse exhibit a gradient where temperatures around the periphery are significantly higher than those in the interior. Vertically, temperatures are highest at the greenhouse roof, followed by the crop canopy layer, and lowest near the ground. On cloudy days, the temperature at the greenhouse roof ranges from 21.8 to 23.3 °C, with most areas between 19.4 and 21.3 °C. On sunny days, the temperature at the greenhouse roof ranges from 23.6 to 26.3 °C, with most areas between 22.7 and 24.8 °C.

3.3. Airflow Field Analysis

Figure 5 shows the dynamic changes in outdoor wind speed under cloudy and sunny conditions. As shown in Figure 5a, there is a positive correlation between outdoor airflow velocity and time under cloudy conditions. During the first period from 9:00 to 11:00, the wind speed gradually increases from an initial value of 1.1 m/s to 1.2 m/s, followed by a sharp increase from 11:00 to 13:00, rising from 1.2 m/s to 1.8 m/s, after which the area stabilizes. Figure 5b shows the trend of outdoor airflow velocity on a sunny day. We can see that the velocity follows a “U”-shaped trend, starting at 0.5 m/s at 9:00 a.m., followed by a sharp increase from 9:00 a.m. to 1:00 p.m., rising by 1.5 m/s to reach 2.0 m/s, after which the outdoor wind speed begins to decrease, finally reaching 1.2 m/s at 17:00.

In the greenhouse ventilation structure, due to the low position of the ventilation openings, the wind speed in the upper part of the greenhouse is relatively low, and even stagnant airflow zones may form, resulting in low ventilation efficiency. Based on the direction of the airflow, the airflow around the crop area forms one or more circulating airflow zones of varying sizes. In these areas, the airflow speed is low, and the airflow mixing and flow characteristics are poor. As shown in Figure 6, the airflow velocity is relatively high at the ventilation openings, but it gradually decreases after entering the crop area. There are certain stagnant zones within the greenhouse, primarily located above the ventilation areas. However, these stagnant zones decrease in size as external wind speeds increase, and the ventilation conditions in the crop planting area—the primary ventilation zone—are good.

The percentage of flow regions under different typical weather conditions is shown in the Figure 7, with detailed data in

Table 5. Percentage of area of different flow rate zones in the greenhouse on cloudy days.

Time	Outdoor Flow Velocity/m s−1	Area Proportion
		<0.5 m s−1	0.5–2.5 m s−1	>2.5 m s−1
9:00	1.1	19.9%	61.2%	19.3%
11:00	1.2	12.7%	70.9%	17.2%
13:00	1.8	8.1%	75.3%	17.3%
15:00	1.7	19.2%	67.6%	13.1%
17:00	1.8	7.3%	72.5%	20.9%

and

Table 6. Percentage of area of different flow rate zones in the greenhouse on a sunny day.

Time	Outdoor Flow Velocity/m s−1	Area Proportion
		<0.5 m s−1	0.5–2.5 m s−1	>2.5 m s−1
9:00	0.5	21.2%	57.7%	21.1%
11:00	1.4	12.5%	66.3%	21.2%
13:00	2.0	31.5%	47.2%	21.3%
15:00	1.7	13.2%	66.4%	20.4%
17:00	1.2	20.3%	58.3%	21.4%

. On cloudy days, the percentages of low-speed stagnant zones in the five time periods were 19.9%, 12.7%, 8.1%, 19.2%, and 7.3%, respectively, while the percentages for high-speed airflow zones were 19.3%, 17.2%, 17.3%, 13.1%, and 20.9%, respectively. The percentages for zones with suitable wind speeds were 61.2%, 70.9%, 75.3%, 67.6%, and 72.5%, respectively. On sunny days, the proportions of low-speed stagnation zones in the five time periods were 21.2%, 12.5%, 31.5%, 13.2%, and 20.3%, respectively, while the proportions of high-speed airflow zones were 21.1%, 21.2%, 21.3%, 20.4%, and 21.4%, respectively. The proportions of zones with suitable airflow speeds were 57.7%, 66.3%, 47.2%, 66.4%, and 58.3%, respectively.

As shown in Figure 8, the ventilation openings are located on the east and west sides of the greenhouse. Regardless of whether it is cloudy or sunny, the outdoor airflow enters the greenhouse through the west-side ventilation openings, where the wind speed gradually decreases. The middle and upper parts of the greenhouse are influenced by the airflow from the ventilation zones inside the greenhouse, forming a clockwise vortex in the middle and upper parts of the greenhouse. Therefore, the wind speed is relatively low, the airflow mixing is poor, and the heat exchange efficiency is low.

4. Discussion

Plastic greenhouses offer advantages such as low cost and high efficiency, and are widely used in most regions of the world [26]. CFD optimization of greenhouse ventilation plays a crucial role in design optimization [27]. To validate the reliability of the established CFD model under natural ventilation conditions, this study selected representative meteorological days for comparison: 7 April as a typical cloudy day and 14 March as a typical sunny day. At five key time points—9:00 a.m., 11:00 a.m., 1:00 p.m., 3:00 p.m., and 5:00 p.m.—the measured outdoor temperature, outdoor wind speed, and ground temperature data were input into the model as boundary conditions. The simulated indoor temperature values at these corresponding times were compared with the measured values (Figure 3). The analysis indicates that the model has high accuracy in predicting temperature. Under overcast conditions (Figure 3a), the absolute error range between the simulated and measured temperatures at each measurement point is 1.0 °C to 2.1 °C, with a maximum relative error of 8.89% and an average relative error of 6.53%. Under sunny conditions (Figure 3b), the absolute error range further narrowed to 1.1 °C to 1.7 °C, with a maximum relative error of 6.29% and an average relative error of 5.56%. This indicates that the relative error of temperature prediction under all conditions is below the common engineering threshold of 10%, demonstrating that the model has acceptable engineering accuracy. The simulated values and measured values show a highly consistent trend over time, reflecting that the model can reasonably capture dynamic thermal environment characteristics [28]. The model performs stably under both typical weather conditions of cloudy and sunny days, further proving its applicability. However, the relative error between simulated and measured values may result from multiple factors. In this simulation, wind speed was set as a constant value at a specific time, whereas outdoor wind speed varies unpredictably in reality [29]. Additionally, the simulation employed a steady-state flow field assumption, which may lead to inaccurate internal flow fields [30]. Due to solar radiation, the actual location of the greenhouse being shaded, or the low temperature of the temperature measurement instruments, the actual measured temperature was lower than the simulated value. In summary, the simulation results and measured data exhibit good consistency in terms of error range and trend, validating the effectiveness of the established CFD model in predicting the thermal environment of naturally ventilated buildings, and providing a reliable numerical tool foundation for subsequent parameterization studies and design optimization.

Further observations indicate that the spatial distribution of indoor temperatures exhibits significant heterogeneity and temporal dynamics. This study, through a comparative analysis of typical cloudy and sunny days, reveals the core distribution patterns of the thermal environment in naturally ventilated greenhouses; vertically, a stable thermal stratification structure of “roof > crop canopy > near-ground” is consistently observed, and this structure is significantly regulated by solar radiation intensity—on sunny days, the roof temperature is 1.8–3.0 °C higher than on cloudy days. The horizontal distribution exhibits dynamic diurnal variation characteristics, as follows: at 9:00 a.m., the temperature field is relatively uniform; at 11:00 a.m., during the enhanced radiation period, the outer film region forms a high-temperature zone due to the thermal boundary effect, while the near-surface region maintains a low temperature, resulting in a horizontal temperature difference of up to 6.7 °C on cloudy days, significantly greater than the 3.2% on sunny days; during the midday radiation peak period (13:00–15:00), convective heat transfer intensifies, improving temperature uniformity, but crop canopies still exhibit significant thermal lag. In the evening, as energy input declines, temperatures decrease from the periphery toward the center, with sunny days maintaining higher residual temperatures due to thermal inertia. These findings validate the reliability of the previous model verification and clarify that solar radiation modulates convective intensity and boundary heat exchange, serving as the core physical mechanism driving thermal heterogeneity and vertical stratification within the greenhouse space. This provides a critical theoretical basis for optimizing ventilation strategies and mitigating crop thermal stress.

The design of low-positioned ventilation openings results in a significant airflow attenuation zone in the upper space of the greenhouse (Figure 6). Outdoor airflow primarily originates from the western ventilation openings with higher initial velocities, but kinetic energy is rapidly dissipated upon entering the crop area, creating a persistent low-speed stagnation zone with wind speeds < 0.3 m/s in the greenhouse roof region. Subsequently, vortex generation and heat exchange inhibition cause the airflow to flow around the crop canopy, forming a clockwise circulating vortex, resulting in a 32–41% reduction in turbulence intensity in the upper space, severely hindering heat and mass exchange between hot air and fresh air. The proportion of stagnant zones shows significant weather-related differences—7.3–19.9% on cloudy days, with an average of 13.4%, while on sunny days it ranges from 12.5% to 31.5%, with an average of 19.7%. During the 1:00 p.m. time slot on sunny days, the stagnant zone area abnormally increased to 31.5%, which is closely related to the enhanced thermal pressure gradient and increased upper air stagnation caused by strong solar radiation [31]. Notably, the ventilation efficiency in the planting area remains relatively stable, with the proportion of suitable speed zones ranging from 61.2% to 75.3% on cloudy days and 47.2% to 66.4% on sunny days. For every 1 m/s increase in external wind speed, the stagnation zone decreases by 4.8–5.6%, suggesting that enhanced driving forces can partially mitigate structural defects. Therefore, optimizing the high-position configuration of ventilation openings to disrupt vortex structures and increase upper wind speeds is a priority direction for improving the overall environmental uniformity of greenhouses.

While the above analyses provide technical pathways for improving ventilation efficiency, their adoption by greenhouse operators depends on both economic feasibility and operational practicality [3]. In typical single-span plastic greenhouses in hot–humid regions of China [32], structural modifications such as raising ventilation openings or adding supplementary roof vents involve an estimated one-time cost of approximately CNY 30–50 per m² (including materials and labor). For a standard 48 m² greenhouse, this equates to CNY 1440–2400 [31], which is substantially lower than the cost of retrofitting mechanical cooling systems. Based on reported yield increases of 8–12% in similar microclimate optimization studies and current vegetable market prices in Sichuan, this investment could be recouped within 1–2 growing seasons [33]. Moreover, these passive ventilation improvements require minimal additional energy input, thereby reducing long-term operational costs. For operators with limited budgets, phased implementation—starting with partial vent elevation in the most stagnant zones identified by CFD analysis—can provide incremental benefits while spreading out capital expenditure. These considerations suggest that the recommended design adjustments are economically accessible and offer a favorable cost–benefit ratio for small- to medium-scale producers in the region.

5. Conclusions

This paper employs CFD modeling methods to establish a spatio-temporal model of temperature and airflow changes in single-span plastic film greenhouses representative of the Chengdu Plain region, targeting the humid–hot zones of China. Simulations were conducted to analyze temperature and airflow changes under typical sunny and cloudy weather conditions, and computational analyses were performed on the internal climate conditions of the greenhouse at different time intervals. Considering the influence of external environmental factors, the following conclusions were drawn:

(1) A three-dimensional CFD model of the plastic greenhouse was constructed using the Realizable k-ε turbulence model. After validation with multi-time-period experimental data, it was confirmed that this model can accurately simulate the dynamic spatiotemporal distribution patterns of temperature and airflow fields within the greenhouse.

(2) Under sunny/cloudy conditions, the temperature trends inside and outside the greenhouse are highly consistent. The greenhouse exhibits stable thermal stratification in the vertical direction: the highest temperature is at the greenhouse roof, followed by the crop canopy layer, and the lowest temperature is near the ground.

(3) Reducing the area of low-speed zones can suppress the extent of recirculation vortices; lateral air intake structures cause large low-speed stagnation zones to form above the ventilation zone, significantly reducing airflow mixing efficiency and heat and mass transfer performance; and raising or adding roof-level vents can improve overall ventilation. This provides reliable optimization criteria for the design of single-span plastic greenhouses in China’s humid and hot regions.