(A) Unlined (control) structure. (B) 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) 5 cm-thick cement blocks buried at a depth of 30 cm.
The residual iteration curve was calculated at 0800 and 1500 HR in the control (CK) test group (left) and structure test group (right) with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60). MSE = mean square error.
Test of simulated and measured values of soil temperature at 0800 and 1500 HR in the structure test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (left) and the structure test group with a lining that included cement blocks buried at a depth of 30 cm (CB30) (right). MSE = mean square error.
Test of simulated and measured values of soil moisture content at 0800 and 1500 HR in the control (CK) test group (top left), and the structure test groups with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60) (top right), a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (bottom left), and a lining that included cement blocks buried at a depth of 30 cm (CB30) (bottom right). MSE = mean square error.
(A) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the control test group. (B) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included cement blocks buried at a depth of 30 cm.
Click on author name to view affiliation information
To investigate the variations in temperature and humidity of the marginal soil on the southern side of a greenhouse during the winter in a cold, arid region of northern China, four linings were compared: an unlined (control) structure, a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm, a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm, and 5 cm-thick cement blocks buried at a depth of 30 cm. In addition, a computational fluid dynamics (CFD) method was used to simulate the soil temperature and humidity in the test area. The results showed that the relative error between the simulated and measured soil temperatures for different linings was less than 10%, the absolute error was 0.06 to 0.49 °C, the relative error of soil moisture contents was less than 1.2%, and the simulation results were more accurate. The linings blocked water and heat movement inside and outside the greenhouse. In summary, we provide a theoretical basis and experimental reference for methods of thermal insulation of marginal soil on the southern side of a greenhouse. To save engineering and material costs, it is recommended to bury a lining of extruded polystyrene boards at depths of 30 and 60 cm when planting shallow-rooted and deep-rooted crops in the greenhouse, respectively.
Soil temperature and moisture have a direct impact on plant root growth and seed germination for crops grown in greenhouses in cold and arid environments (Li et al. 2025; Zhang 2022). During the winter at a location in the northern region of China, the temperature and moisture of the marginal soil inside the greenhouse were greatly affected by the low-temperature soil outside the greenhouse, the soil temperature and humidity in the greenhouse differed during the day and night, and the effects on surface soil were particularly intense. The temperature changes affected the infiltration, redistribution, and evaporation of moisture in the soil. When the soil surface temperature is higher, there is more evaporation of moisture from the soil (Hei et al. 2025; Teng et al. 2016). In addition, changes in soil moisture affect its thermal characteristics, which in turn affect soil temperature (Roxy et al. 2014; Sumatran et al. 2014; Zhang et al. 2025; Zhao 2020). Many scholars have studied different climatic conditions in various regions and have shown that there is a linear relationship between the soil temperature at the surface (within 10 cm of the surface) and the change in soil moisture (Hu et al. 2019; Lakshmi et al. 2003; Liu et al. 2019; Mekki et al. 2013). Scholars have also established a variety of numerical models for the coupled transport of soil water and heat, such as Klute’s improved nonisothermal diffusion equation (Klute 1952). A model of the migration of the two-phase flow of liquid and gaseous water in soil under the combined action of a hydrothermal gradient, established by Philip (1957) and Philip and De Vries (1957), used water, heat, and salt transport equations and continuity equations to establish a water–heat–solute coupled transport model based on the Philip model (Nassar and Horton 1989, 1992). Domestic scholars (Guo and Li 1997; Meng and Xiao 2005) proposed a soil water–heat coupled transport model for water and heat in soil under changing temperature conditions based on the mutual coupling of water and heat transport in farmland soil. In recent years, with the development of computational fluid dynamics (CFD) technology, CFD has been widely used in simulations of the relationship between the greenhouse environment and crop growth (Roy et al. 2017; Salazar et al. 2017; Zhang et al. 2017b), and the spatial and temporal distribution of the greenhouse environment (Li et al. 2006; Molina-Aiz et al. 2017; Roy et al. 2014; Zhao et al. 2014). Zhang (2017) and Zhang et al. (2017a) used the CFD method to simulate the temperature and humidity of the marginal soil on the southern side of a greenhouse. According to the t test and F test of variances, the model was proved to be accurate. Saiyin (2019) used the CFD method to simulate the soil temperature numerically in a greenhouse and obtained a soil temperature prediction model with a relative error of less than 5%.
In summary, scholars in China and other nations have conducted in-depth studies on water and heat transfer and coupling in soil. However, there have been few studies of the effects of different lining structures on the movement of water and heat in soil inside and outside greenhouses, and there have been few studies on the use of the CFD method to simulate the transfer of water and heat in soil inside and outside greenhouses. We studied the transfer of water and heat in soil inside and outside a greenhouse from 0800 to 1500 HR during the day, and used a simulation analysis to compare effective measures suitable for the thermal insulation of the marginal soil on the southern side of the greenhouse.
The greenhouse faced south, its length was 70 m from east to west, and its span from north to south was 8.5 m. To reduce the influence of the boundaries of the greenhouse on the experiment, the central area of the greenhouse was selected for this experiment—namely, the control area, the area in which a 5 cm-thick polystyrene extruded board was buried at a depth of 60 cm (XPS60), the area in which a 5 cm-thick polystyrene extruded board was buried at a depth of 30 cm (XPS30), and the area in which 5 cm-thick cement blocks were buried at a depth of 30 cm (CB30). Each test area was 2.4 m long from east to west. The lining structures were buried outside the foundation on the south side of the greenhouse, and the thickness of the lining materials was 5 cm. Indoor and outdoor soil temperature and humidity sensors (hereinafter referred to as soil sensors) were buried at vertical depths of 5 cm, 15 cm, 25 cm, and 55 cm belowground. The outdoor soil sensors were located 20 cm from the lining structure, and the indoor soil sensors were positioned 10, 65, and 120 cm from the foundation of the south side of the greenhouse. The soil temperature and humidity sensors were produced by Hebei Oudu Technology Co, Ltd (Hebei, China); the soil temperature measurement range was –30 to –70 °C, with a measurement accuracy of ±0.2 °C; and the soil moisture measurement range was 0% to 50%, with a measurement accuracy of ±2% (measured in cubic meter per cubic meter).
For seepage in porous media, the flow is a laminar motion of an incompressible fluid. When the source term is not considered, the main governing equations are the continuity, momentum, and energy equations.
The continuous equation is [1] the momentum equation is [2] and the energy equation is [3] where ε* is the porosity of the porous media, Si is the momentum source term added to the standard momentum equation in the fluent porous media model, ρ is fluid density, μi is The velocity components (i represents the spatial coordinate directions, such as the x, y, and z directions), characterize the fluid’s movement speed in different directions, t is time, xi and xj is spatial coordinates (i, j) respectively represent different spatial directions and are used to describe spatial positions, P is pressure; μ is the dynamic viscosity of a fluid, ρt is density of the fluid phase, ct is specific heat capacity of the solid phase, ρf is density of the fluid phase, cf is the specific heat capacity of the fluid phase, T is temperature, Kt is thermal conductivity of the solid phase, ρtct and kt are the total heat capacity and total thermal conductivity of the porous medium, respectively. Si, ρtct, and kt can be determined according to the following equations: [4] [5]
and [6] where is the viscosity loss term; is inertia loss; Kd is the permeability coefficient; ρfcf and kf represent the heat capacity and thermal conductivity of the liquid phase in the porous medium, respectively; and ρscs and ks represent the heat capacity and thermal conductivity of the solid phase in the porous medium, respectively. Scholar pointed out the following: The local sewage seepage velocity is relatively small, the inertial loss term can be ignored (i.e., C2 is 0), and the additional momentum source term only considers the viscous loss term.
The water content in saturated soil was measured in the range of 0.47% to 0.51%. During the study period, the soil water content inside and outside the greenhouse was less than the soil saturated water content; therefore, the soil pores were unsaturated, and the soil pores contained water in liquid and gaseous states. In our study, the main components in the soil porous medium were liquid water and gaseous water vapor. Activating the moisture transport model and defining the component materials as liquid water and water vapor, the moisture evaporation option was selected under the influence of temperature (Zhao 2020), and the calculation was as follows: [7] where Ci is the volume concentration of component i in the mixed gas, ρCi is the mass concentration of component i, Di is the mass diffusivity of component i, and Si is the extra generation rate of the generalized source.
Porosity is the ratio of the skeleton pore volume to the total volume in a representative unit. We used the ring knife method to measure the soil porosity: [8] where ε* is the soil Porosity, which is the percentage of the soil pore volume in the total volume; G1 is the total mass of the ring knife and wet soil; g is the quality of the ring knife; G2 is the total mass of the ring knife and the dried soil; ρ is density of water; and V is the volume of the ring knife.
Permeability is the permeability of porous media to fluids, as demonstrated by [9] where Kd is soil permeability and d is the average diameter of soil particles.
The viscous resistance coefficient (Vrc) and inertial resistance coefficient (Irc) are important resistance coefficients of the porous media model, and they are determined as follows: [10] and [11] According to [Eqs. 8–11], the main parameters of the porous media model of each layer of soil were obtained, as shown in Table 1.
The boundary conditions included soil temperature and humidity values at different depths outside the greenhouse, at different depths inside the greenhouse (120 cm from the south side of the foundation), at the soil surface inside and outside the greenhouse, and under the foundation of the greenhouse. The soil was a porous medium with laminar movement. The thermal performance parameters of each material are shown in Table 2.
The winter weather was cold, dry, and mostly sunny, and there was not much rain or snow. This test started 11 Nov 2020. The test data for a sunny day, 13 Jan 2021, was selected for analysis. Seventeen days after the last watering, because of the short solar radiation time and low radiation value in the northern winter, the evaporation from the soil on the south side of the greenhouse was slow. The soil moisture content at each test point in the greenhouse was 15% to 30%, and the soil function rate at the test point outside the greenhouse was 6% to 10%. According to the analysis of the test data, the daily soil temperature was at a minimum between 0730 and 0900 HR, and the soil temperature was at a maximum at 1500 to 1600 HR. The data measured at 0800 and 1500 HR were selected as the initial values for the simulation of soil water and heat transfer.
The mean square error (MSE) reflected the overall error between the measured value and the simulated value. The smaller the mean square error, the greater the measurement accuracy. The coefficient of determination (R2) evaluated the accuracy of the constructed transfer function, reflecting the degree of agreement between the measured value and the simulated value. These values were derived as [12] and [13] where ymi is the measured value of the model, ypi is the estimated value of the model, is the average of the measured values, N is the total number of samples, SSE is sum of squared errors, and SSQ is sum of squared total.
A model was established by dividing the test area vertically into three sections: 1) outside the greenhouse, 2) under the foundation, and 3) inside the greenhouse. The soil was divided horizontally into areas 5, 15, 25, and 55 cm belowground. The depth of the foundation was 15 cm below the surface on average, and the simulated area was 120 cm long inside the greenhouse and 60 cm outside the greenhouse. ICEM CFD software ver. Ansys 2018 (ICEM CFD Engineering, Berkeley, CA, USA) was used to divide the mesh. The mesh had a regular quadrilateral unstructured structure. The maximum grid size was 0.1 cm. The number of grids in the four regions was more than 13,000. The mesh calculation quality was more than 0.95 and the grid quality was good, which met the calculation requirements, as shown in Fig. 1. A steady-state method was used to solve the control equations, the numerical calculation was performed with the second-order upside style finite volume method, and the pressure and velocity coupling momentum equation used the coupling algorithm. Figure 2 shows the residual error iteration diagram of the XPS60 lining test group.
Citation: HortTechnology 35, 6; 10.21273/HORTTECH05778-25
Citation: HortTechnology 35, 6; 10.21273/HORTTECH05778-25
Because the number of convergence steps of the experimental residual errors of each group is similar, and considering the length of this article, a set of residual error iteration diagrams at 0800 and 1500 is now described.
To verify the accuracy of the CFD simulation, the actual measured values of soil temperature and humidity at locations 10 and 65 cm from the south side of the foundation at soil depths of 5, 15, 25, and 55 cm were selected for the different lining test groups. A straight line was established 10 and 65 cm from the south side of the foundation through CFD-Post (ANSYS Inc, Canonsburg, PA, USA), and the corresponding points on the straight line were used to simulate temperature and humidity values of the soil at different depths. A linear equation was established between the measured values (x-axis) and the simulated values (y-axis). The closer the scattered points from the model to the 1:1 line, the more accurate the model simulation results. The smaller the MSE value, the smaller the MSE between the simulated and measured values.
Figure 3 shows the degree of dispersion of the simulated and measured soil temperature changes on both sides of the 1:1 line at each time point. The scattered points for soil temperature were distributed on both sides of the 1:1 line, the degree of fit of the scattered points was greater than 0.95, and the MSE value was small. Table 3 shows the relative errors between the simulated and measured values of the soil temperature at 0800 and 1500 HR. The relative errors of the simulated and measured values in each test group were less than 10% at different times and locations. The simulated values of the model were compared with the measured values. The absolute error of the value was 0.06 to 0.49 °C, and the simulation was more accurate. Figure 4 shows the degree of dispersion between the simulated and measured values of soil moisture change on both sides of the 1:1 line. Because of the small soil moisture changes, the scattered points were more dispersed, but the MSE value was small, both less than 0.015, showing that the absolute error of soil moisture was less than 0.012. In summary, it was more accurate to use the CFD method to simulate changes in soil temperature and humidity for different linings, and this model can be used to analyze changes in soil temperature and humidity.
Citation: HortTechnology 35, 6; 10.21273/HORTTECH05778-25
Citation: HortTechnology 35, 6; 10.21273/HORTTECH05778-25
Figure 5A shows that in the control test area at 0800 HR, the outdoor low-temperature soil migrated into the greenhouse, and the high-temperature soil in the greenhouse migrated to the low-temperature area. There was an S-type temperature gradient, and the south–north span was within 30 cm of the foundation. The soil temperature at different depths was greatly affected by the external low-temperature soil, the temperature gradient was an I type, and the marginal low-temperature effect was obvious. At 1500 HR, the soil temperature inside and outside the greenhouse rose (compared with 0800 HR), the soil depth in the greenhouse was more than 10 cm, and the surface soil temperature rose faster. The soil temperature gradient within 30 cm of the foundation was of the reverse-C type, and the temperature was relatively low. The reason for this is that the heat transfer was mainly through the solid soil framework and the fluid in the soil pores, but these methods of transfer were relatively slow, and the low-temperature soil outside the greenhouse continued to move into the greenhouse. The soil temperature gradient 30 cm from the foundation gradually changed from an S type at 0800 HR to an I type, because in the low-temperature environment of the greenhouse, the soil was a heat storage body and released heat continuously to the environment outside the greenhouse. The stored heat decreased greatly. As the ambient temperature in the greenhouse rose, the surface soil absorbed a large amount of heat, resulting in the longitudinal temperature resembling the ambient temperature. Compared with 0800 HR, the soil moisture was less at 1500 HR within 80 cm from the south side of the foundation. The main reason for this was that, as the soil temperature rose, the soil moisture evaporated. The soil moisture increased outside, 80 cm from the south side of the foundation. Combined with the actual situation in the greenhouse at 1200 HR, as the ambient temperature in the greenhouse rose, the frost on the plastic film in the greenhouse gradually melted and dripped, which caused the soil moisture in the local area to increase.
Citation: HortTechnology 35, 6; 10.21273/HORTTECH05778-25
Figure 5B shows that in the XPS60 test area, the XPS60 lining structure effectively prevented the low temperature outside the greenhouse from moving into the greenhouse at 0800 HR, and it also prevented the indoor soil temperature from moving outward. The soil temperature gradient in the greenhouse was S type. Near the edge, the temperature gradient followed a C-type trend. Because of the barrier of the lining structure, the soil temperature below 15 cm was less affected by the external low temperature, so the heat release was less for the deep soil than the surface soil. At 1500 HR, as the soil temperature rose, the surface layer of the soil heated faster than the deep layer, so the soil temperature gradient had a reverse-C–type trend. At 1500 HR, the maximum average temperature difference between the XPS60 group and the control group at different depths in the greenhouse was 0.8 °C. At 1500 HR, as the soil temperature increased, the soil moisture was affected by evaporation, which was less compared with that at 0800 HR, with an average decrease of 0.2%. Because of the low crop coverage in the marginal area, the change in soil moisture was mainly a result of soil evaporation. As the surface soil moisture decreased, the soil suction effect gradually increased, and the deep soil moisture continued to rise through the soil capillary pores when the soil depth was less than 15 cm. The soil moisture tended to increase compared with that noted at 0800 HR.
Figure 5C shows the XPS30 test area. At 0800 HR, the soil temperature gradient was like that of the XPS60 lining structure; but, below the 30-cm lining, the low-temperature soil outside the greenhouse tended to migrate into the greenhouse. However, because of the small variation in water and heat transport in the soil below 25 cm, the migration of the low temperature from the outside was concentrated mainly in the deep soil under the foundation, which caused the average temperature of the deep soil to be less than that of the XPS60 group in the long term. At 1500 HR, the surface temperature of the soil was high, the soil temperature moved downward, and the temperature gradient had a reverse-C shape. With the increase in soil temperature compared with that at 0800 HR, the soil moisture followed a decreasing trend as a result of evaporation. Like the XPS60 text area, the soil depth was below 15 cm, and the soil moisture tended to increase compared with that at 0800 HR.
Figure 5D shows the CB30 test area. At 0800 HR, the low-temperature soil outside the greenhouse penetrated the CB30 lining structure and moved into the greenhouse. Because the low temperature had weakened after penetration, it had not moved into the greenhouse, and the soil temperature gradient was similar to that of the XPS30 test area. Compared with the XPS30 lining at 1500 HR, 10 and 65 cm away from the south side of the foundation, the soil temperature above 30 cm was lower by 0.5 and 0.2 °C, respectively. The change in soil moisture was similar to that in the XPS30 group. With the increase in soil temperature compared with 0800 HR, the soil moisture followed a decreasing trend, resulting from the evaporation of soil moisture.
For different lining structures, the soil temperature gradient moves from S type at 0800 HR to reverse-C type at 1500 HR. The main reason for this was that there was a certain delay in the soil heating. As the ambient temperature rose, the temperature of the surface soil rose faster than that of the deep soil, which was consistent with the results of most scholars (Guo 2016; Liu et al. 2010; Teng 2017). According to the simulation results, the change in the soil temperature gradient was also related to the moisture content in the soil. Compared with other reverse-C–type phenomena, the soil temperature gradient was smaller and similar to reverse-L–type behavior. The main reason for this was that the soil moisture content was low, and the heat transfer between the soil skeletons was relatively fast, which is similar to the results of Teng (2017). With increasing temperature, irregular movement of soil moisture occurred, primarily because the increase in system temperature reduced the surface tension and viscosity coefficient of soil moisture, which increased the water conductivity and changed the soil water potential. This result is similar to the simulation results of Ren et al. (2017).
The soil data in the test area showed that the soil mechanical composition was similar in each area of the greenhouse; the soil quality was loamy sandy soil with sand grains of diameter 0.02 to 2 mm, accounting for more than 95%; soil porosity at depths of 25 cm or more in the greenhouse was relatively large; the measured porosity was 0.45% to 0.47%; and the soil porosity at a depth of 55 cm was 0.375%. Outdoors, the soil porosity was 0.34% to 0.38% at depths of 5 and 15 cm, respectively, and 0.42% to 0.44% at depths of 25 and 55 cm, respectively. As a result of test errors, there were also errors in the calculation of the porous media parameters. To improve the accuracy of the simulation in follow-up research, the porosity parameters of different-lining test groups should be measured, and the CFD model should be subdivided into regions.
CFD is an accurate means to simulate changes in soil water and temperature. This method can be used to simulate different forms of lining and to identify optimal marginal insulation that can prevent changes in high-temperature soil in greenhouses. External heat transfer affects the loss of soil heat in greenhouses. Extruded polystyrene foam board lining material provides better thermal insulation than cement block lining material. To save engineering and material costs, an extruded polystyrene foam board lining buried at a depth of 30 cm is recommended when planting shallow-rooted crops in a greenhouse; an extruded polystyrene board lining buried at a depth of 60 cm is recommended for deep-rooted crops.
(A) Unlined (control) structure. (B) 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) 5 cm-thick cement blocks buried at a depth of 30 cm.
The residual iteration curve was calculated at 0800 and 1500 HR in the control (CK) test group (left) and structure test group (right) with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60). MSE = mean square error.
Test of simulated and measured values of soil temperature at 0800 and 1500 HR in the structure test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (left) and the structure test group with a lining that included cement blocks buried at a depth of 30 cm (CB30) (right). MSE = mean square error.
Test of simulated and measured values of soil moisture content at 0800 and 1500 HR in the control (CK) test group (top left), and the structure test groups with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60) (top right), a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (bottom left), and a lining that included cement blocks buried at a depth of 30 cm (CB30) (bottom right). MSE = mean square error.
(A) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the control test group. (B) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included cement blocks buried at a depth of 30 cm.
Contributor Notes
This work was supported by the “Mirror Lake Talents” Project of Huzhou Vocational and Technical College (JH2023007).
X.Z. is the corresponding author. E-mail: z_xu1987@163.com.
(A) Unlined (control) structure. (B) 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) 5 cm-thick cement blocks buried at a depth of 30 cm.
The residual iteration curve was calculated at 0800 and 1500 HR in the control (CK) test group (left) and structure test group (right) with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60). MSE = mean square error.
Test of simulated and measured values of soil temperature at 0800 and 1500 HR in the structure test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (left) and the structure test group with a lining that included cement blocks buried at a depth of 30 cm (CB30) (right). MSE = mean square error.
Test of simulated and measured values of soil moisture content at 0800 and 1500 HR in the control (CK) test group (top left), and the structure test groups with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm (XPS60) (top right), a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm (XPS30) (bottom left), and a lining that included cement blocks buried at a depth of 30 cm (CB30) (bottom right). MSE = mean square error.
(A) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the control test group. (B) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 60 cm. (C) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included a 5 cm-thick polystyrene extruded board buried at a depth of 30 cm. (D) Simulation of soil temperature and humidity migration from 0800 to 1500 HR in the test group with a lining that included cement blocks buried at a depth of 30 cm.
