Effects of Plastic Mulch on Soil Heat Flux and Energy Balance in a Cotton Field in Northwest China

Abstract

Surface energy balance (SEB) is a basic principle for all of the atmospheric circulation models, and surface soil heat flux (G₀) is an important component of energy balance. Plastic mulch is widely used in arid regions and alters land surface processes. In this study, hourly/daily G₀ and SEB over a mulched cotton field were analyzed in Northwest China. The net radiation beneath the mulch was simulated using transmitted down short-wave and down long-wave radiation through the mulch and up short-wave and up long-wave radiation from the soil beneath mulch. The results show that: (1) on the hourly time scale, G₀ in mulched soil is much smaller than that in non-mulched soil, day and night. This implies that mulch prevents energy going into the soil during the daytime and reduces soil heat loss at night. On the daily time scale, G₀ is similar in mulched and non-mulched soil. (2) During the seedling emergence period when the cotton is small, the energy balance closure over mulched soil (0.79) is slightly smaller than that over non-mulched soil (0.81). (3) Comparing to soil temperature in non-mulched soil, there’s a time offset on soil temperature in mulched soil at the same depth, which will bring a hysteresis on G₀ and SEB.

1. Introduction

Energy exchange between the surface and the atmosphere is one of the most crucial ecological processes in terrestrial ecosystems [1,2]. Surface energy balance is the basic principle for atmospheric circulation models [3] and is also essential for surface evapotranspiration estimation [4,5,6,7]. Surface soil heat flux (G₀) represents the energy absorbed or released by surface soil during a given time period [2]. It is an important component of the surface energy balance. Although G₀ is relatively small on daily and longer timescales, the instantaneous G₀ can contribute as much as 50% of the net radiation for dry and bare soil [8,9,10,11]. G₀ is also a significant component of the energy balance for vegetated surfaces [12,13] and plays an important role in estimating surface evapotranspiration [14]. Incorrect G₀ is supposed to be an important reason for energy imbalance [15]. Land surface change will affect G₀ and energy balance and impact atmospheric properties [16]. The land surface characteristics, such as soil moisture, surface albedo, leaf area index and roughness, are related to land surface fluxes and are important for the simulation of local and regional weather and climate variables, including air temperature, wind, humidity and precipitation [17,18,19,20,21]. The land cover change can influence the atmospheric circulation not only at meteorological timescales, but also at climatological timescales [22].

The plastic film mulching technique and drip irrigation method have been widely applied in Xinjiang in Northwest China [23]. The mulched area amounted to more than 1.2 million hectares in 2009 in Xinjiang [24]. The film mulch reduces heat convection and evaporation from the soil to the atmosphere. The mulch also changes surface albedo and soil moisture, which will impact the meteorological parameters. Moreover, the droplets on the film mulch and the water vapor in the air beneath the film mulch can absorb long-wave radiation, resulting in more soil heating due to the greenhouse effect [25]. The soil thermal and moisture conditions are improved by mulch to ensure crop germination and small plant growth [24,26,27,28]. The mulch promotes crop development and yield [29]. Polythene mulch increased soil temperature by approximately 6 °C at a 5-cm depth in northern Vietnam [29] and by 1.6 °C at a 15-cm depth in Xinjiang, China [30]. Mulch greatly retards the loss of moisture from soil, which reduces the irrigation frequency and amount [29]. The mulch is also beneficial for disease control [25].

The mulch alters matter and energy exchanges and affects land surface processes, which has attracted substantial research. Previous studies mainly focused on the effect of mulch on soil temperature, crop growth and yield, while few research studies address its effect on the soil heat flux and energy balance. The present study will explore how the plastic mulch affects the soil heat flux and energy balance with 3-year (2013–2015) continuous observations at the Korla cotton site in Xinjiang, Northwest China. The surface soil heat flux was calculated by the thermal conduction equation using the observed soil temperature and moisture profile. The air heat storage and atmospheric moisture change were taken into account for energy balance, together with the observed sensible and latent heat flux by an eddy covariance system (EC). The net radiation beneath the mulch is essential for the surface energy balance beneath the mulch. It was calculated using transmitted down short-wave and down long-wave radiation through the mulch and up short-wave and up long-wave radiation of soil beneath mulch.

2. Methods and Materials

2.1. Experimental Site and Data Records

The Korla site (86°12′E, 41°36′N, 886 m a.s.l.) is located in the northeast Tarim River Basin in the Xinjiang Uygur Autonomous Region (Figure 1). It is characterized by a typical inland arid climate with scarce precipitation and strong diurnal temperature fluctuation [24]. The mean annual precipitation is approximately 60 mm, and the mean temperature is 11.48 °C [24]. The mean annual potential evaporation measured by a Φ20 evaporation pan (20 cm in diameter) is approximately 2788 mm [31]. The Korla site is covered by cotton planted in late April and harvested in September. The major soil type is silt loam. The cotton yield contributes to nearly 50% of the total lint yield of China, with approximately 3.2 million tons in 2012 [24]. The distribution of cotton planting, plastic mulch and drip pipe, shown in Figure 2, is referred to as a “one pipe, one film and four rows of cotton arrangement” [32]. The width of the mulch is approximately 110 cm, and the inter-mulch zone is approximately 40 cm.

The hourly datasets in 2013–2015 were used in the present study and included EC and automatic weather station (AWS) measurements (Figure 3).

Table 1. The parameters and depths/heights of sensors at the Korla site during 2013–2015.

Instrument	Parameters	2013 (m)	2014 (m)	2015 (m)
Eddy covariance system (EC)	Latent heat flux and sensible heat flux	2.25	2.25	2.25
	Atmospheric moisture density	2.25	2.25	2.25
Automatic weather station (AWS)	Soil temperature	0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8	0.1, 0.2, 0.3, 0.5, 0.6	0.05, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8, 1.0, 1.2
	Soil water content	0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8	0.1, 0.2, 0.3, 0.5, 0.6	0.05, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8, 1.0, 1.2
	Net radiation	2.25	2.25	2.25
	Air temperature	2.8	2.8	2.8

shows detailed information on the datasets. The EC consists of an EC150 open-path CO₂/H₂O gas analyzer (Campbell Scientific Inc., Logan, UT, USA), a CSAT3 three-dimensional sonic anemometer (Campbell Scientific Inc., Logan, UT, USA) and an air temperature/humidity sensor (HMP155A, Vaisala Inc., Woburn, MA, USA) installed at a 2.25-m height. The EC data were recorded by a data logger (CR3000, Campbell Scientific Inc., Logan, UT, USA) at 30-min time intervals. The energy and water vapor fluxes (latent heat flux LE and sensible heat flux H) and carbon dioxide flux were measured by EC. The AWS measurements contained wind speed and direction, air temperature and pressure, precipitation, net radiation, soil temperature and soil moisture profiles. The net radiation was measured at a height of 2.25 m by a net radiometer (CNR 4, Kipp&Zonen, Delft, The Netherlands). The soil temperature and moisture were measured by a Hydra Probe® sensor (Stevens Water Monitoring System, Inc., Beaverton, OR, USA) in mulched soil and non-mulched soil, respectively (Figure 3b).

The following post-processing procedures were implemented to control the quality of the EC data: linear de-trending, tilt correction through the yaw and pitch rotation, density fluctuation correction and sonic temperature correction [24]. The datasets from 2013–2015 with a 30-min time scale were converted to hourly values in this study. Furthermore, the missing data accounted for less than 1%, and the data collected on a few rainy days were also included.

2.2. Methods

The full surface energy balance equation can be expressed as follows [33,34,35]:

$$R_{\mathrm{n}}=LE+H+G_0+S_{\mathrm{c}}+S_p+S_a+S_q+S_d$$

(1)

where R_n is the net radiation, L (J·kg⁻¹) is the latent heat of vaporization and E (kg·m⁻²·s⁻¹) is the evaporation, the product of which is the latent heat flux. H is the sensible heat flux; G₀ is the surface soil heat flux; S_c is the canopy heat storage in the biomass and water; S_p is the photosynthesis flux; S_a is the air storage between the eddy covariance system and the ground surface; S_q is the atmospheric moisture change; and S_d is the canopy dew enthalpy change (all fluxes are in W·m⁻²). S_c, S_p and S_d were not included in this study, as cotton plants are very small during the period for the study of energy balance.

2.2.1. Surface Soil Heat Flux

The surface soil heat flux was calculated by a “TDEC” method (thermal diffusion equation and correction) proposed by [36], using soil temperature and moisture observations that do not require accurate thermal conductivity. The equation is as follows:

$$G(0,t)=G(z_{ref},t)+{\displaystyle \underset{z_{ref}}{\overset{0}{\int }}\frac{\partial {\rho }_s{c}_{s}T(z,t)}{\partial t}}dz$$

(2)

where G (0, t) (W·m⁻²) is the surface soil heat flux at time t (s) and G (z_ref,t) (W·m⁻²) is the soil heat flux at a reference depth z_ref at time t, assumed to be zero if z_ref is deep enough. ρ_s (kg·m⁻³) is the soil bulk density, and c_s (J·kg⁻¹·K⁻¹) is the heat capacity; the product of ρ_s and c_s is the soil volumetric heat capacity (J·m⁻³·K⁻¹). T (z,t) (K) is the soil temperature at depth z and time t.

The soil volumetric heat capacity can be obtained from the following equation:

$${\rho }_s{c}_s={\rho }_{solid}{c}_{solid}(1-{\theta }_{sat})+{\rho }_w{c}_w\theta$$

(3)

where ρ_solidc_solid is the volumetric heat capacity of the soil solids, with a value of 2.1 × 10⁶ J·m⁻³·K⁻¹. The uncertainty of the ρ_solidc_solid value brings little uncertainty of G₀. ρ_wc_w is the volumetric heat capacity of liquid water, with a value of 4.2 × 10⁶ J·m^–3·K⁻¹. θ_sat (%) and θ (cm³·cm⁻³) are soil porosity and observed soil water content, respectively. θ_sat is 42% observed at the Korla site.

The key issue in calculating G₀ is obtaining a reliable soil temperature profile from limited observed temperatures. Yang and Wang (2008) [36] proposed a new method for obtaining a preliminary soil temperature profile (80 layers in the present study) by the tridiagonal matrix algorithm with the T (z₀) and T (z_bot) as the upper and lower boundary conditions, respectively, and then adjusting the profile with the observed soil temperature. T (z₀) (denoted as “T_s”) for bare soil was calculated from the upward and downward long-wave radiation fluxes:

$$T({\mathrm{z}}_0)={(\frac{R_L^{\uparrow }-(1-{\epsilon }_{\mathrm{s}})R_L^{\downarrow }}{{\epsilon }_s\sigma })}^{1/4}$$

(4)

where $R_L^{\uparrow }$ and $R_L^{\downarrow }$ are the observed upward and downward long-wave radiation fluxes (W·m⁻²); εs is the land surface emissivity, which is 0.96 for silt loam soils [37]; and the Stefan–Boltzmann constant is σ = 5.67 × 10⁻⁸ W·m⁻²·K⁻⁴ T (z₀) for soil beneath mulch will be given in next section. T (z_bot) is the bottom soil temperature measurement. The soil heat fluxes at different soil depths can be calculated by integrating the adjusted soil temperature profile with the bottom soil heat flux (e.g., 1.6 m) as zero. The observed soil temperature is assimilated in the process of simulating the soil temperature profile, which leads to the TDEC method’s lack of sensitivity to soil thermal conductivity that is difficult to obtain. In addition, TDEC is also not sensitive to T (z₀), as it depends on multiple layers of soil temperature [38].

2.2.2. Net Radiation

The net radiation for bare soil (non-mulched) is as follows:

$$R_{\mathrm{n}}=(1-{\rho }_s)S+R_L^{\downarrow }-{\mathsf{\epsilon }}_s\mathsf{\sigma }{T}_s^4$$

(5)

where R_n (W·m⁻²) is the surface soil net radiation, S (W·m^–2) is solar radiation and r_s is the surface soil reflectivity, 0.15 for silt loam soils [37]. ε_s is the surface soil emissivity, given as 0.96 [37].

The net radiation in the surface soil beneath mulch is as follows [39]:

$$R_n^{\prime }=(1-{\mathrm{r}}_s){\mathsf{\tau }}_{m1}S+{\mathsf{\tau }}_{m2}{R}_L^{\downarrow }+{\mathsf{\epsilon }}_m\mathsf{\sigma }{T}_m^4-{\mathsf{\epsilon }}_s\mathsf{\sigma }{T}_{\mathrm{s}}^{\prime 4}$$

(6)

where R′_n (W·m⁻²) is the net radiation over the soil beneath the mulch; τ_m1 = 0.84 and τ_m2 = 0.82 are the transmittances of mulch for solar radiation and thermal radiation, respectively [37]; εm is the mulch emissivity, given as 0.05 [37]; T_m (K) is the mulch temperature, and T_s′ (K) is the surface soil temperature beneath the mulch. According to the experiment, Tm was taken as the same as T_s′ in the present study. The observed $R_L^{\uparrow }$ above the mulch contains three parts: reflective $R_L^{\downarrow }$ by mulch, transmitted thermal radiation of soil beneath the mulch through the mulch (${\mathsf{\tau }}_{m2}{\mathsf{\epsilon }}_s\mathsf{\sigma }{T}_{\mathrm{s}}^{\prime 4}$), and transmitted $R_L^{\downarrow }$ to soil beneath the mulch reflected by the soil and then transmitted to the air through the mulch. Thus, T_s′ can be calculated by the following equation:

$${\mathsf{\epsilon }}_s\mathsf{\sigma }{T}_{\mathrm{s}}^{\prime 4}=\frac{R_L^{\uparrow }-(1-{\mathsf{\tau }}_{m2}-{\mathsf{\epsilon }}_m)\times R_L^{\downarrow }-(1-{\mathsf{\epsilon }}_s)\times {\mathsf{\tau }}_{m2}^2\times R_L^{\downarrow }}{{\mathsf{\tau }}_{m2}}$$

(7)

2.2.3. Air Heat Storage

The air heat storage between the ground surface and the eddy covariance system is:

$$S_a={\rho }_a{c}_a{\displaystyle {\int }_0^h\frac{\partial T_a}{\partial t}}dz$$

(8)

where ρ_a (kg·m⁻³) is the air density; c_a (J·kg⁻¹·K⁻¹) is the specific heat capacity of moist air; T_a (K) is the air temperature at 2.8 m in the present study; and h = 2.25 m is the height of the EC.

2.2.4. Atmospheric Moisture Change

The atmospheric moisture change is:

$$S_q=L{\displaystyle {\int }_0^h\frac{\partial q}{\partial t}}dz$$

(9)

where q (kg·m⁻³) is the moist air density and the latent heat of vaporization L = 2.5 × 106 J·kg⁻¹.

3. Results and Discussions

3.1. The Effect of Mulch on Soil Temperature and Energy Fluxes

The soil temperature at depths of 5 cm, 10 cm and 20 cm, as well as R_n, H and LE during the seedling emergence period, are shown in Figure 4. The seedling emergence period is from 23 April–15 May, when the cotton plants are small and there is no drip irrigation. The impact of plastic mulch on land energy/mass exchange is most obvious in this period, and the effect of vegetation on G₀ and energy balance can be ignored. The soil temperature at a 5-cm depth in mulched soil is on average 2.9 °C higher than that in non-mulched soil, which shows the thermal insulation of plastic mulch. In addition, the difference of soil temperatures between non-mulched soil and mulched soil decreases as the soil depth increases. The average soil temperature differences at 10 cm and 20 cm are 1.9 °C and 1.3 °C, respectively. The amplitude of soil temperature decreases with deeper soil and the phase shift delays. LE is small, as the cotton plants are very small with less transpiration. Moreover, only soil evaporation in non-mulched soil contributes to LE. The main turbulent flux is H in this period.

There is a hysteresis between soil temperature in non-mulched soil and in mulched soil during the 3-year mean 24-h period, with a counterclockwise distribution (Figure 5). The same soil temperature in non-mulched soil corresponds to two different values in mulched soil in the morning and afternoon, respectively. It shows that the mulch delays soil temperature variation, which affects the phase of G₀ in mulched soil.

3.2. The Effect of Mulch on Surface Soil Heat Flux

3.2.1. Damped Diurnal Variation of Surface Soil Heat Flux beneath Mulch

The hourly surface soil heat fluxes over non-mulched and mulched soil were calculated by the TDEC method during the growth periods (23 April–18 September) in 2013–2015. To show the evolution of G₀ over the growing season, variations of the 3-year average of hourly G₀ from April–September are shown in Figure 6. Not only the positive G₀ (daytime), but also the negative G₀ (nighttime) in mulched soil are smaller than in non-mulched soil. The sign represents the heat flux direction; the positive G₀ represents heat flows from the surface to deeper soil; and the negative is opposite. The G₀ values in mulched soil decrease obviously in late June. The slope of regression between G₀ in non-mulched soil and mulched soil during the growing season (from 16 May–18 September) is 0.52, smaller than the value of 0.72 in the seedling emergence period (Figure 7), because the cotton grows, and the dense canopy attenuates solar radiation entering into the soil. Furthermore, drip irrigation creates higher soil moisture, and the soil temperature beneath the mulch varies slowly, bringing about low G₀. However, G₀ in non-mulched soil increases about on 21 June, 4 July and 14 July. That is because the soil temperature varies largely in these days, which determines the G₀ value. There is no drip irrigation in these days, and soil moisture is small. Furthermore, R_n usually reaches the maximum in July, later than the solar radiation.

The 3-year mean diurnal variation of G₀ in the seedling emergence period (23 April–15 May), the blooming and bolling period (1 July–31 August) and the growing season (April–September) were studied separately. The G₀ in mulched and non-mulched soil shows significant differences during a 24-h period. Moreover, the absolute value of G₀ in mulched soil is smaller than that in non-mulched soil in both day and night, with slopes of regression between G₀ in non-mulched soil and mulched soil of 0.72, 0.40 and 0.55 in these three growing periods (Figure 8d,e,f). The plastic mulch not only weakens solar radiation into the soil during the day, but also hinders soil heat into the air during the night. In addition, the difference becomes large in the blooming and bolling periods (Figure 8b), owing to the aforementioned dense vegetation and drip irrigation in addition to the plastic mulch effect. The daily maximum G₀ in mulched soil is 106 W·m⁻² in the seedling emergence period and 47 W·m⁻² in the blooming and bolling period. In addition, the time when heat transmission changes direction (from negative to positive) in the morning in non-mulched and mulched soil is the same, approximately 9:00 a.m. (Figure 8a). However, the heat transmission in non-mulched soil changes earlier in the afternoon than that in mulched soil, around 18:30 and 19:30, respectively (Figure 8b). The plastic mulch delays the change of direction in heat flow in the afternoon.

According to the thermal conduction equation, G₀ values are related to the variation of temperature in the topsoil at adjacent hours. In the present study, the daily maximum soil temperature observed in the topsoil minus the daily minimum value (5 cm in 2013 and 2015, 10 cm in 2014) represents the time variation of soil temperature, which controls G₀ values. The 3-year averages of the daily amplitude (maximum minus minimum) of the observed soil temperature in mulched and non-mulched zones during April–September are shown in Figure 9. The variations of the amplitudes of soil temperature are consistent with G₀ variations. The amplitude of temperature in non-mulched soil is larger than that in mulched soil from April–September. The biggest difference between the amplitudes of temperature in non-mulched and mulched soil appears at the beginning of July, in agreement with G₀ variation, which is caused by solar energy, dense vegetation and drip irrigation.

Although the plastic mulch increases soil temperature, the amplitude of soil temperature beneath the mulch is smaller than that in non-mulched soil in the seedling emergence period (Figure 9), which leads to smaller G₀. It follows that the plastic mulch keeps soil temperature stable.

3.2.2. Similar Daily Variation of G₀

The solar energy absorbed by the soil is not always equal to the released energy, which depends on the underlying surface and seasons. Thus, the daily G₀ (net energy entering the soil) is not always equal to zero [2]. Since the hourly G₀ beneath the mulch is much smaller than that in non-mulched soil both in day and night, what about the daily values? The daily G₀ in mulched and non-mulched soil are similar, and both monotonically decrease from April–September (Figure 10), changing from positive to negative values. It shows that the soil absorbs energy in the spring and releases energy in the autumn. The G₀ values are positive before July, alternately positive and negative from July to early August and almost negative starting from early August. The energy absorbed by the soil beneath the mulch and by bare soil between two mulch zones is larger than the released energy from April–June, which makes for positive G₀, and is opposite from early August–September. The maximum G₀ beneath the mulch was 20.5 W·m⁻² on 7 May, and the minimum is −14 W·m⁻² over bare soil on 8 July.

Compared to the hourly G₀ in non-mulched soil and that of mulched soil with a slope of 0.72 (Figure 7a), the daily G₀ are nearly same, with a slope of 1.07 in the seeding and germination period (Figure 11a). Although n = 23 in Figure 11a and the p-value is 5.88 × 10⁻¹² < 0.05, they are significantly correlated at a confidence level of 0.05. It shows that the effect of mulch on G₀ is diminished with increasing time scale and that mulch has a great impact on diurnal G₀. However, the daily G₀ in non-mulched soil is larger than that in mulched soil from May–September when the cotton grows, with a slope of 0.83 (Figure 11b). That is not only because of the positive G₀, but also because the negative G₀ in non-mulched soil is larger than that in mulched soil. The dense vegetation and mulch constitute two masks over the soil, which leads to less energy entering the soil and also reduces soil heat loss, leading to smaller positive and negative G₀ values.

The sum of positive values of daily G₀ beneath the mulch in the seedling emergence period is 242.5 W·m⁻², and the sum of negative values is −3.6 W·m⁻². These sums of G₀ in non-mulched soil are 200 and −3.6 W·m⁻², respectively (

Table 2. The sum of positive, negative and all 3-year mean daily G₀ values over non-mulched and mulched soil from 23 April–15 May and from 16 May–18 September at the Korla site in 2013–2015.

	Non-mulched	Beneath Mulch	Non-mulched	Beneath Mulch
	23 April–15 May (W·m–2)		16 May–18 September (W·m–2)
Sum of positive	200.0	242.5	409.0	446.5
Sum of negative	−3.6	−3.6	−281.7	−202.5
Sum	196.4	238.8	127.4	244.0

). The cumulative energy is positive (absorbing energy), with a value of 238.8 W·m⁻² in mulched soil, and larger than that in non-mulched soil, with 196.4 W·m⁻² in the seedling emergence period. Furthermore, the heat income to soil beneath mulch is larger than that to bare soil (242.5 > 200 W·m⁻²), and heat losses from the soil to air are less affected by the presence of the mulch, with the same value of 3.6 W·m⁻², similar to Liakatas’s research [39]. This demonstrates that the plastic mulch gathers energy for seed germination and growth. Although the vegetation and mulch reduce soil heat gain and loss, the cumulative energy is positive, with a value of 244.0 W·m⁻², larger than 127.4 W·m⁻² in non-mulched soil.

3.3. The Effect of Mulch on Energy Balance

3.3.1. Smaller Net Radiation beneath Mulch

In addition to atmospheric long-wave radiation, the mulch also contributes downward long-wave radiation to the net radiation of the soil beneath the mulch. The hourly mulch temperature and surface soil temperature beneath the mulch (8:00–20:00) were measured using an infrared thermometer (FLUKE Inc., Everett, WA, USA) for seven days (27–30 April and 3–5 May) in 2016. Ten data points per hour were measured for mulch and the soil beneath mulch. The average of the 10 data points was adopted as the final hourly mulch and soil temperatures. The mulch temperature is nearly the same as the soil temperature beneath mulch (Figure 12a) with a slope of regression between mulch temperature and soil temperature of 0.96 (Figure 12b), which is consistent with Liakatas’s study [39].

The 3-year mean hourly net radiation in mulched soil is smaller than that in non-mulched bare soil (Figure 13a), in agreement with other results, such as those of [39,40,41]. The slope of regression between R_n in non-mulched and mulched soil is 0.80, and the determination coefficient is 0.99 (Figure 14a). Primarily because the net solar radiation is reduced by 16% while the net long-wave radiation is approximately equal (Figure 13b). The slopes of regression between net solar radiation in non-mulched soil and that in mulched soil and between net long-wave radiation in non-mulched soil and that in mulched soil are 0.84 and 1.03, respectively (Figure 14b,c). The humid air held below the mulch, which can absorb the downward long-wave radiation penetrating through mulch and upward long-wave radiation emitted by the soil beneath the mulch, will contribute to R′_n and SEB. Nevertheless, the humid air is not included in the present study, as the mulch is close to the ground and the space is very small.

3.3.2. Energy Balance in Mulched and Non-Mulched Soil

To show the effect of plastic mulch on land surface energy balance, data from 23 April–15 May were selected along with the air heat storage and atmospheric moisture change. To show the energy balance closure (EBC), the sum of 3-year mean hourly turbulent energy fluxes (H and LE) and heat storage were plotted against the 3-year mean hourly available energy (R_n – G₀) in Figure 15, using 3-year mean hourly data from 2013–2015 at the Korla site. The EBC in surface soil beneath mulch is 0.79 (Figure 15b), smaller than the 0.81 in non-mulched soil (Figure 15a). That may be because the transpiration is likely not included in mulched soil, as the cotton has emerged on approximately 5 May and has reached a height of 10 cm on 15 May. Moreover, the uncertainty of parameters (emissivity, transmittance of the mulch and the emissivity of soil) in R′_n calculation may bring uncertainty of EBC in mulched soil.

There is a hysteresis phenomenon over non-mulched soil (Figure 15a) that appears with an annular distribution with a determination coefficient of 0.94, smaller than the 0.96 over mulched soil. That is mainly because G₀ is out of sync with R_n over non-mulched soil (Figure 15c), and G₀ generally reaches a diurnal maximum earlier than R_n [4]. In addition, H, LE and S are almost in sync with R_n during the 24-hour period. The available energy is out of sync with turbulent fluxes over non-mulched soil. However, G₀ is more in sync with R′_n in mulched soil (Figure 15d), as the soil temperature in mulched soil delays, which leads to the hysteresis over G₀. That leads to the available energy being in sync with the sum of turbulent energy fluxes and heat storage over mulched soil.

4. Conclusions

Plastic mulch has been widely used in agriculture in Northwest China because it can not only promote soil temperature, but can also reduce water loss from the soil surface, which is beneficial to seed germination. In addition, the mulch also modifies the thermal microclimate of the soil. Nevertheless, few studies address its effect on the soil microclimate and energy balance. This study focused on how the mulch affects G₀ and SEB. However, the climate parameters’ change (e.g., air temperature, wind, precipitation, convection cloud, etc.) caused by changed G₀ and SEB in mulched soil is not included in the present study. In this study, the hourly surface soil heat flux was calculated from April–September in 2013–2015 using a heat conduction equation based on soil temperature and moisture measurements in mulched and non-mulched soil at the Korla experimental site planted with cotton in the Xinjiang Uygur Autonomous Region. The variations of G₀ at different time scales were analyzed. In addition to the surface net radiation beneath mulch, the energy balance in mulched and non-mulched soil were also discussed. The main conclusions obtained from the investigation are summarized as follows:

The 3-year mean hourly G₀ in mulched soil is smaller than that in non-mulched soil both day and night from April–September. The plastic mulch not only prevents solar radiation into the soil in the daytime, but also reduces soil heat loss at night. The hourly G₀ over mulched soil decreases obviously in late June, as solar energy is blocked by the grown cotton and because the drip irrigation increases soil moisture, making soil temperature vary slowly and bringing about low G₀. Moreover, the mulch delays the moment when heat flow changes direction in the afternoon.

Although the hourly G₀ are largely different, the 3-year mean daily G₀ in non-mulched and mulched soil are similar, with a slope of regression of 1.07 in the seeding and germination period. This shows that mulch has a great impact on instantaneous G₀ and less impact on G₀ at a daily time scale. During the whole growing season, the daily G₀ decreases monotonically. G₀ values are positive before July and almost negative starting from early August, which indicates that soil absorbs energy in spring and releases energy in autumn. The sum of positive values of daily G₀ in mulched soil is 242.5 W·m⁻², larger than 200 W·m⁻² in non-mulched soil in the seeding and germination period. The sum of negative G₀ over mulched and non-mulched soil is the same, −3.6 W·m⁻². Thus, the mulch cumulates more energy into the soil than bare soil.

The surface net radiation beneath the mulch is 0.88 times that over non-mulched soil as less solar radiation enters into the soil through the mulch, while the net long-wave radiation is almost the same as that over non-mulched soil.

The EBC in mulched soil is smaller than that over non-mulched soil, 0.79 versus 0.81, based on 3-year mean hourly R_n, G₀, H, LE, air heat storage and atmospheric moisture changes. LE is not included in mulched soil, which may lead to smaller EBC. The uncertainty of R’n coming from the uncertainty of parameters may also affect EBC in mulched soil. A hysteresis phenomenon appears in non-mulched soil between H + LE + S and R_n – G₀, as G₀ is out of sync with Rn. However, there is no hysteresis in mulched soil. That is mainly because the soil temperature delays in mulched soil and leads to G₀ delays, which makes G₀ is sync with R′_n. The average EBC is approximately 0.8 in the present study, which is reasonable and consistent with the study of Wilson et al. (2002) [42]. The 20% imbalance may be due to the accuracies of R_n, G₀, the mismatch between the turbulent flux footprint and other components of the energy balance, or the absence of advection.