Agricultural and Forest Meteorology 260–261 (2018) 176–182

Contents lists available at ScienceDirect

Agricultural and Forest Meteorology

journal homepage: www.elsevier.com/locate/agrformet

Determination of soil ground heat flux through heat pulse and plate T
methods: Effects of subsurface latent heat on surface energy balance closure
⁎ ⁎
Sen Lua,b, , Hesong Wangc, Ping Menga,b, , Jinsong Zhanga,b, Xiao Zhangd
a
Key Laboratory of Tree Breeding and Cultivation of State Forestry Administration, Research Institute of Forestry, Chinese Academy of Forestry, Beijing, 100091, China
b
Co-Innovation Center for Sustaintable Forestry in Southern China, Nanjing Forestry University, Nanjing 210037, China
c
College of Forestry, Beijing Forestry University, Beijing, 100083, China
d
Institute of Desertification Studies, Chinese Academy of Forestry, Beijing, 100091, China

A R T I C LE I N FO A B S T R A C T

Keywords: The soil heat flux plate method is popularly applied in surface energy balance studies. Previous studies have
Soil heat flux shown that impervious plate blocks the flow of water and vapor within soil. Soil heat flux is generally commonly
Surface energy balance measured below surface, and its exact constitution is required in calculating surface energy balance. When
Subsurface latent heat subsurface evaporation occurs, subsurface latent heat sink constitutes an important proportion of the apparent
Heat pulse
ground heat flux. However, the plate method fails to detect such occurrence. In aboveground meteorological
measurement, evaporated vapor moving out of soil profile is also being detected and the subsurface latent heat
sink is recognized as part of turbulent latent heat flux. Thus, caution should be exercised when excluding the
potential error from double counting of subsurface latent heat sink in surface energy balance evaluation. In this
study, two common combination methods were used to determine the ground heat flux without latent heat sink
(G0). One method is a combination of gradient-based heat pulse measurements and calorimetric method
(GradC), and the other method is a combination of plate measurements at shallow depths and calorimetric
method (PlateC). Results demonstrated that, in contrast to the PlateC method, the GradC method minimized the
disturbance in soil structure and reduced the disruption in heat and water flow. Furthermore, the estimated G0
from the PlateC method was only 49.2% of that of the GradC method during daytime. Moreover, surface energy
balance closure (EBC) was evaluated using the estimated G0 and aboveground turbulent energy flux data. In
comparison with the PlateC method, the GradC method improved the surface EBC from 79.3% to 87.7% during
daytime. In summary, accurate knowledge on the composition of ground heat flux and the location of water
evaporation is necessary to calculate surface energy balance during micro-meteorology measurements.

1. Introduction surface–atmosphere turbulent fluxes of sensible heat and latent heat (W

Understanding the energy exchange between land surface and at-
mosphere is important in simulating hydrological, atmospheric, and
ecological processes (Baldocchi et al., 2001; Stoy et al., 2009). How-
ever, the surface energy imbalance in micrometeorological studies re-
mains unsolved (Wilson et al., 2002; Foken, 2008; Foken et al., 2011;
Stoy et al., 2013; Russell et al., 2015). In the past decades, imbalance in
energy budget has been widely studied, and the average energy balance
closure (EBC) from the eddy covariance (EC) measurement ranges be-
tween 0.75 and 0.87 at most flux sites (Stoy et al., 2013).
The energy balance equation is generally expressed as
Rn − G0 = H + LET
where Rn is the net radiation (W m−2); H and LET are the
m−2), respectively; G0 is the ground heat flux at soil surface without
latent heat sink (W m−2).
The left side of Rn−G0 is often termed as the available energy, and
EBC is denoted as the ratio of turbulent fluxes (H+LET) to available
energy. Notably, Eq. (1) assumes that evaporation occurs at the soil
surface (Wang and Bou-Zeid, 2012). This condition only occurs on soils
with large water content and that immediately following rainfall
(Heitman et al., 2008a). If the evaporation front migrates downward
into the subsurface, then the evaporation actually occurs at a certain
depth below the surface (Mayocchi and Bristow, 1995; Heitman et al.,
2010). Thus, accurate knowledge on the composition of soil ground
(1) heat flux is necessary because a large proportion of apparent ground
heat flux is contributed by subsurface latent heat sink. Meanwhile, in
aboveground EC measurement, soil water evaporated vapor combines

⁎
Corresponding author at: Address: No 1, Dong Xiao Fu, Beijing, 100091, China.
E-mail addresses: asen205@cau.edu.cn (S. Lu), mengping@caf.ac.cn (P. Meng).

https://doi.org/10.1016/j.agrformet.2018.06.008
Received 25 October 2017; Received in revised form 7 June 2018; Accepted 8 June 2018
0168-1923/ © 2018 Elsevier B.V. All rights reserved.
S. Lu et al. Agricultural and Forest Meteorology 260–261 (2018) 176–182

Nomenclature LET Surface–atmosphere turbulent flux of latent heat

PlateC Combination of heat flux plate measurements and calori-
d Diameter of heat flux plate metric method
E Soil evaporation rate Pt Thickness of heat flux plate
EBC Energy balance closure Rn Net radiation
EC Eddy covariance T Soil temperature
ET Evapotranspiration rate t time
f Geometrical factor associated with the dimensions of heat z Soil depth
flux plate
G0 Ground heat flux without latent heat sink Greek symbols
Ga Apparent soil heat flux
Gp Corrected apparent soil heat flux determined with the λp Plate thermal conductivity
plate method λs Soil thermal conductivity
Gpu Uncorrected apparent soil heat flux measurement ob- ΔS Change in soil heat storage
tained using the plate method ΔT/Δz Soil temperature gradient
Gr Apparent soil heat flux at a reference depth
GradC Combination of gradient-based heat pulse measurements Subscripts
and calorimetric method
H Surface–atmosphere turbulent flux of sensible heat i Index variables for soil depth layers
L Volumetric latent heat of vaporization j Index variables for time steps

with transpiration vapor and is detected and counted as part of tur- heat storage (ΔS, W m−2) between the reference depth and the soil
bulent latent heat flux. Thus, G0 in Eq. (1) should exclude the propor- surface as follows:
tion of subsurface latent heat sink. An accurate understanding of the
G0 = Gr + ΔS (2)
subsurface latent heat sink dynamics is required to avoid double
counting error in EBC calculation. This research aims to determine the location and magnitude of
The heat flux plate method is commonly used for apparent soil heat subsurface evaporation by using the heat pulse technique and obtain
flux measurement at present (Sauer et al., 2003; Heusinkveld et al., the ground heat flux without latent heat sink by using the combination
2004; Sauer et al., 2007). The plate embeds a thermopile in a thin disk method. The contribution of subsurface latent heat sink on the apparent
with a fixed thermal conductivity, and the measured thermopile voltage heat flux at soil surface is illustrated and distinguished. We also use the
output is converted into heat flux. Heat flow divergence or convergence G0 of this study and aboveground turbulent energy data to improve the
may be induced in plate measurements because of differences between surface EBC. The effect of vapor flow blockage of the plate method on
plate and variable soil thermal conductivities (Philip, 1961; Sauer et al., surface energy balance is also demonstrated.
2003; Liebethal and Foken, 2006). This error can be corrected if soil
thermal conductivity (λs), plate thermal conductivity (λp), and plate
2. Materials and methods
dimensions are known (Philip, 1961). However, the continuous λs in-
formation in many applications is not commonly available under field
2.1. Experimental setup
conditions (Ochsner et al., 2006). Heat flux plates are generally placed
horizontally at depths between 2.5 and 10 cm (Sauer, 2002) but are
Measurements were conducted on a semi-arid grassland of Tongyu
mostly positioned at depths between 5 and 10 cm (Liebethal et al.,
Country, Jilin Province, northeastern China (44°59ʹ N, 122°52ʹ W) in
2005; Lindroth et al., 2010; Evett et al., 2012). In some cases, plates are
July 2015 (DOY 187–199). The annual mean air temperature of the
situated at shallow depths (e.g., ≤ 2 cm) to reduce the magnitude of
study site is 6.3 °C, with a range from −33.7 °C to 38.9 °C. The mean
heat storage correction in the layer above the plates (e.g., Heusinkveld
annual precipitation is 320 mm, and approximately 80% of rainfall
et al., 2004; Amiro, 2009; Evans et al., 2012; Higgins et al., 2013; Ma
occurs during summer. The soil at the site is chernozem and possesses a
et al., 2014; Chow et al., 2014). Owing to the impervious plate design,
sandy loam texture. The grassland is covered largely by Chloris virgate
the plate method exerts a large disturbance on soil structure and blocks
community and annual weed community with a spatially random dis-
heat and fluid flow (Ochsner et al., 2006). The latent heat sink occur-
tribution. Leymus chinensis plants are visible on occasion. In the past
ring within soil cannot be measured properly. Moreover, divergence of
decade, drought has exacerbated in the study region. Few bare patches
heat flow may be caused by thermal contact resistance at the plate–soil
are visible occasionally because of drought. During our measurement
interface in plate measurements (Sauer et al., 2003; Ochsner et al.,
period, grass was sparse and vegetation height was approximately
2006).
3–9 cm.
Unlike the plate method, the gradient method based on the heat
Turbulent heat flux was measured from the EC system in the site.
pulse technique usually possesses a small needle diameter and presents
The EC systems consisted of an ultrasonic anemometer (Model CSAT3,
advantages of minimized soil disturbance. Cobos and Baker (2003) and
Campbell Scientific Ltd.) and a Li-7500 open path CO2/H2O analyzer to
Ochsner et al. (2006) reported the good performance of the heat pulse
continuously measure CO2, H2O, and energy fluxes half-hourly. The
method on soil heat flux measurement both in the laboratory and field.
meteorological instruments on the tower included temperature and
Moreover, the heat pulse method provides a dynamic determination of
humidity measurements (HMP 45CL, Vaisala Inc.), wind speed and
soil temperature and thermal properties (e.g., soil thermal conductivity
direction (034AL and 014AL, Met One Inc.), and radiation measure-
and soil volumetric heat capacity) at various depths and allows calcu-
ment (CM21 and CG4, Kipp and Zonen Inc.). The measurement height
lation of subsurface soil evaporation at fine-scale depth increments
of the radiation and EC system on grassland was 2 m. The soil heat flux
(Heitman et al., 2008a,b; Deol et al., 2012).
measurements in the current study included four heat flux plate plots
G0 can be calculated by a typical combination method by combining
and two heat pulse sensor plots. During the measurement period, the
the apparent soil heat flux measurement at a reference depth (Gr, W
underlying surface was flat and all of the sensors were installed on bare
m−2) obtained using the heat pulse or plate method and the change in
soils between grass stems. Pre-experiment showed that the calculated

177
S. Lu et al. Agricultural and Forest Meteorology 260–261 (2018) 176–182

average relative deviations of soil thermal conductivity and tempera-
ture gradient measurements at 1 cm depth amongst these plots were
0.11 and 0.09, respectively. According to previous flux footprint ana-
lysis, approximately 85% of the measured turbulent fluxes of latent heat
and sensible heat originate within 600 m of the tower (Liu et al., 2008;
Wang et al., 2010). In the current study, all of the soil heat flux mea-
surement plots were located within 50 m at the southeastern direction
of the tower, which is the main source area according to prior footprint
analysis.
For the plate method (HFP01, Hukseflux Thermal Sensors,
Netherlands), the installment depth of soil heat flux plates (80-mm
diameter) was 4 mm (Fig. 1). In this study, we buried the plates at a
shallow depth of 4 mm for two reasons. First, subsurface evaporation
occurs mostly at the upper soil depth. The potential error from the
double counting of subsurface latent heat sink on EBC calculation can
be illustrated clearly only at shallow depths. Second, the effect of vapor
transfer disruption from the impervious plate can be illustrated clearly
at shallow depths. The heat pulse sensors were built following the study
of Zhang et al. (2012) and Deol et al. (2012) and each sensor consisted
of 11 stainless steel needles (1.3 mm diameter). A chromel-constantan
thermocouple was positioned at each needle for temperature mea-
surement (Fig. 1). The four longer needles contained a resistance heater
for producing a heat pulse. By calibrating the sensor in agar-stabilized
water (5 g per litre) before installment, the apparent needle-to-needle
spacing was obtained. Additional details on the heat pulse sensor design
are given by Zhang et al. (2012) and Deol et al. (2012). For the in-
stallation of each multi-needle heat pulse sensor, a small trench was dug
and the sensor was carefully inserted in a vertical orientation with its
needles parallel to the soil surface. The top needle of the sensors was
installed at 1 mm depth (Fig. 1). Finally, the trenches were back-filled.
During measurement, the heater needles were opened every 4 h and a
12 V battery was used to provide heating electrical current. For each
heater cycle, the pulse time was 8 s and the soil temperature was col-
lected for 180 s at 1 s intervals. Measured temperature increase was
corrected using the ambient temperature drift following the method of
Jury and Bellantuoni (1976). Soil thermal properties at various soil
depths (Fig. 1) were calculated with the nonlinear regression technique
by using the soil temperature with time data (Welch et al., 1996).
Moreover, soil temperatures were recorded at 5 min interval with 11
needles (Fig. 1) to obtain temperature gradients at different depths.
2.2. Calculation of apparent soil heat flux, ΔS, and subsurface evaporation
with the heat pulse method
N
Ti, j − Ti, j − 1
ΔS = ∑ Ci,j−1 t j − t j−1
(z i − z i − 1)
i=1 (5)
where C is the soil volumetric heat capacity (J m−3 K−1); the subscripts
i and j are the index variables for depth layers and time steps, respec-
tively.
2.3. Correction of plate measurements with the Philip method
Philip (1961) presented the following general method to correct the
heat flow distortion errors of soil heat flux plate measurements:
Gpu λ p / λs
=
Gp 1 + (λp / λs − 1) f (6)
where Gpu is the uncorrected apparent soil heat flux measurement ob-
tained using the plate method (W m−2), Gp is the corrected apparent
soil heat flux determined with the plate method (W m−2), and λp is the
plate thermal conductivity (0.8 W m−1 K−1, manufacturer-specified
property). In this study, λs measurements from the heat pulse method
were applied for the Philip correction calculation. f is a geometrical
factor associated with the dimensions of the plate. For circular plates, f
is calculated with the following equation (Philip, 1961; Liebethal and
Foken, 2006):
Pt
f = 1 − 1.92
d (7)
where Pt is the thickness of the plate (5 mm, manufacturer-specified
property), and d is the diameter of the plate (80 mm, manufacturer-
specified property).
2.4. Calculation of G0 with the combination methods
In this study, two common combination methods were used to de-
termine G0 by using Eq. (2). One method is a combination of gradient-
based heat pulse measurements and calorimetric method (GradC), and
the other method is a combination of plate measurements and calori-
metric method (PlateC).
For the PlateC method, Gr was obtained using the corrected heat
flux plate measurement at a depth of 4 mm, that is, Gp (4 mm). The
change in heat storage at 0–4 mm was calculated using Eq. (5), where
the soil volumetric heat capacity was determined through the heat

The apparent soil heat flux at various depths was calculated ac-
cording to Fourier’s law by employing soil thermal conductivity and
temperature gradient measurements as follows:

ΔT
Ga = −λs
Δz (3)

where Ga is the apparent soil heat flux at various depths obtained using
the heat pulse method (W m−2), λs is the soil thermal conductivity (W
m−1 K−1), T is the soil temperature (°C), z is the soil depth (m), and ΔT/
Δz is the soil temperature gradient (°C m−1).
For each soil layer, soil water vaporization could be estimated fol-
lowing the sensible heat balance method (Gardner and Hanks, 1966;
Heitman et al., 2008a) as follows:

LE = (Ga1 − Ga2) − ΔS (4)

−2
where Ga1 and Ga2 are the apparent soil heat fluxes (W m ) measured
at upper and lower boundaries of the soil layers, respectively; L is the
volumetric latent heat of vaporization (J m−3); E is the evaporation rate
(m s−1). The change in heat storage (ΔS, W m−2) for various soil layers Fig. 1. Schematic of the soil heat flux sensors installation depths (not to scale).
can be determined as (Heitman et al., 2008b): The dimensions are in millimeters.

178
S. Lu et al. Agricultural and Forest Meteorology 260–261 (2018) 176–182

pulse measurements. In the present study, we chose 4 mm as the depth

of plate placement to illustrate the effects of vapor transfer disruption
on G0 determination and EBC calculation clearly.
For the GradC method calculation, the measured soil volumetric
heat capacity at various depths from the heat pulse method was used to
determine ΔS through Eq. (5). The multi-needle heat pulse method can
be utilized to measure Ga and soil evaporation dynamics at various
depths. To determine G0 without latent heat sink, we obtained Gr by
using the measured Ga at a depth with no evaporation consumption in
accordance with E results.

2.5. EC flux calculation and data processing

All of the EC data were processed with EddyPro software (Li-Cor,

Inc., NE, USA), and data processing involved spikes removal, coordinate
rotation (double rotations), and frequency corrections. Webb, Pearman,
and Leuning (WPL) term (Webb et al., 1980), which accounts for errors Fig. 3. Temporal changes in measured apparent soil heat flux for the 4 mm soil
introduced by fluctuations in water vapor density and temperature, was depth obtained using heat pulse (Ga) and plate (Gp) methods.
applied to correct LET and H fluxes. During our measurement period
(DOY 187–199), a gap at night (within 2 h) due to rain on DOY 191
(Fig. 2) was filled using a linear interpolation method.

3. Results and discussion

3.1. Comparison of apparent soil heat flux between the heat pulse method
and the plate method

Heat flux measurements at multiple depths by using the heat pulse

method allow comparison of the plate measurements at 4 mm depth.
Fig. 3 shows that the maximum Ga observed from the multi-needle
gradient method at the 4 mm depth is nearly 450 W m−2 on DOY 189,
which is approximately 66.6% of the net radiation (Fig. 3). This large
proportion of net radiation contributing to the Ga (4 mm) is due to
sparse vegetation and measured apparent heat flux that includes energy
consumption for water evaporation deeper than 4 mm (Novak, 2010;
Heitman et al., 2010). However, even after Philip correction is carried
out, the observed Gp at a depth of 4 mm at the same time by using the
Fig. 4. Temporal changes in the measured apparent soil heat flux (Ga) at
plate method is only 47.7 W m−2, which is nearly 7.1% of the net ra-
multiple depth increments obtained with the multi-needle heat pulse method.
diation. Heat flow distortion correction shows small calculated devia-
tions of the plate measurements before and after the Philip correction is
applied (< 5 W m−2). Thus, the considerable discrepancy between Ga 3.2. Measured Ga at various depths and subsurface evaporation patterns
(4 mm) and Gp (4 mm) demonstrates that a significant heat loss occurs
because of water evaporation and that the plate method presents dif- The measured Ga dynamics at various depths from the multi-needle
ficulty in properly responding to subsurface water evaporation. More- heat pulse method is shown in Fig. 4. The results reveal that the mag-
over, the heat pulse method exhibits a small needle diameter of 1.3 mm, nitude of Ga varies largely with depth. For example, on DOY 189, the
thereby minimizing disturbance in soil structure and disruption in heat observed Ga maximum at 4 mm is 448.8 W m−2, which is 3.5 times of
and water flow. Ga (25 mm) and 7.5 times of Ga (46 mm). Divergence in Ga

Fig. 2. Time series of rainfall and net radiation during the measurement period.

179
S. Lu et al.
measurement at various depths indicates that phase changes occur
within the soil profile (Gardner and Hanks, 1966; Heitman et al.,
2008a). By employing the fine-scale temperature and thermal property
measurement performed using the multi-needle heat pulse method,
sensible heat balance of soil layers was computed for estimating sub-
surface evaporation (Heitman et al., 2008b; Deol et al., 2012). Fig. 5
reveals the water evaporation dynamics at various soil depths. For ex-
ample, on DOY 189, the evaporation rates at 4–11 mm reach a peak of
0.66 mm h−1. Thereafter, the evaporation rate decreases gradually
until a new rainfall occurs at DOY 191. When soil water is not limiting,
such as immediately following rainfall event, evaporation occurs at or
near the soil surface in general (stage 1 evaporation). The relatively
lower E on DOY 192 (Fig. 5) does not suggest less total evaporation,
rather it indicates that evaporation occurs mostly near the soil surface.
Thereafter, evaporation occurring at deep soil depths increases from
DOY 193. As time increases, the evaporation zone proceeds downward
gradually (Fig. 5). Thus, understanding the composition of Ga at various
depths is important because this composition varies with the evapora-
tion zone location.
3.3. Surface energy balance evaluation by using the two methods
Accurate knowledge about apparent ground heat flux component is
required in the surface energy balance calculation. In the aboveground
EC measurement, evaporated vapor moves out of the soil profile and is
recognized as a part of the turbulent latent heat flux. In soil heat flux
measurement, the evaporation energy within the soil profile is counted
as a part of the apparent ground heat flux. If the energy consumed by
the water evaporation and its location within soil are not properly
determined, then double counting may occur (Ochsner et al., 2006).
Therefore, accurate separation of latent heat contribution consumed in
subsurface evaporation from the apparent ground heat flux is required
in surface energy balance studies. Only the ground heat flux without
latent heat sink (G0) can be incorporated into Eq. (1) in calculating the
surface EBC. However, in the commonly used plate method, the im-
pervious plate (80 mm diameter in this study) may block the liquid and
vapor transport. Evaporated vapor at deep depths may condense at the
low surface of plate disk and is detected, but the latent heat transfer
cannot be measured properly (Ochsner et al., 2006). On the contrary,
the gradient-based heat pulse method possesses advantage of negligible
disturbance in soil structure and thus provides accurate measurement.
The accurate determination of G0 is associated with the subsurface
evaporation process. According to the heat pulse method in this study,
no significant evaporation is observed at soil layers deeper than 38 mm
(Fig. 5). This observation agrees with that of Heitman et al. (2008b) and
Zhang et al. (2012), in which water evaporation occurred mostly at
upper soil depth. By employing the measured Ga at a deep layer with no
latent heat consumption and ΔS up to the soil surface, G0 can be ob-
tained using Eq. (2). In the current study, we used the measured Ga at
53 mm (as Gr) and ΔS (0–53 mm) to calculate G0 (GradC method).
Moreover, an alternative combination method that combines the mea-
sured Gp (4 mm) and ΔS at 0–4 mm to calculate G0 was included (PlateC
method). Fig. 6 reveals that ΔS at 0–53 mm ranges from −31 W m−2 to
71 W m−2. This result demonstrates that a large error will be produced
if the change in heat storage is ignored in the G0 calculation. Further
comparison of G0 estimates between the GradC and PlateC methods is
provided in Fig. 7. Evidently, the G0 values from the PlateC method are
constantly less than 60 W m−2, which is significantly lesser than that
with the GradC method. This remarkable discrepancy may arise from
the heat flow distortion, water flow disruption, and large thermal
contact resistance of the heat flux plate at the plate–soil interface (Sauer
et al., 2003). Ochsner et al. (2006) reported that impervious plate disks
caused the overlying soil to dry more rapidly than the surrounding soil.
As a result, the upper drier soil reduces soil thermal conductivity and
produces low heat flux estimations.
Fig. 8 shows the relationship between the available energy (Rn−G0)
180
Agricultural and Forest Meteorology 260–261 (2018) 176–182
and the sum of turbulent fluxes of latent heat and sensible heat (LET +
H) on DOY 193, where G0 was obtained from the GradC and PlateC
methods. The results reveal that the average EBC values during daytime
from 8:00 to 16:00 are 96.9% and 84.8% for the GradC and PlateC
methods on DOY 193, respectively. This result indicates that the GradC
method significantly improves the surface energy balance over the
PlateC method (Fig. 8). In general, soil water evaporation occurs sig-
nificantly at daytime, especially around noon. Therefore, the im-
provement of the surface energy balance closure obtained with heat
pulse method is significant at daytime. At nighttime, the evaporation
rate is low, and the effect of vapor transfer disruption from the im-
pervious plate is reduced. Thus, the improvement of the GradC method
over the PlateC method is weakened (Fig. 8).
Fig. 9 shows the calculated (LET + H) versus (Rn−G0) at half-
hourly during the entire measurement period (daytime and nighttime),
where G0 was obtained from GradC and PlateC methods. Linear re-
gression on the total day data shows that the slope of GradC method
(1.13) is closer to the 1:1 line than the slope of the PlateC method
(1.26). Further test conducted on daytime reveals a similar conclusion
that the GradC method improves the EBC from 79.3% to 87.7% during
daytime compared with that of the PlateC method. Moreover, the EBC
improvement of GradC method over the PlateC method increases as
subsurface evaporation increases after rainfall. According to these re-
sults, the heat pulse method is a good alternative to heat flux plates.
This method not only offers a direct measurement of soil thermal
property and temperature gradient but also presents the advantage of
slight disturbance in soil structure (Heitman et al., 2008b). Our results
suggest that increased adoption of the GradC method can improve the
accuracy of soil heat flux measurement and surface EBC.
Moreover, for the plate method, the improved G0 estimations are
produced if heat flux plates are buried in a deep layer (e.g., 10 cm) with
heat storage adjustment. At a depth of 10 cm where no evaporation
occurs, our previous studies revealed that the differences in Ga (10 cm)
measurements between heat pulse and plate methods were small (Lu
et al., 2016). Similar G0 estimations are produced because the sub-
sequent heat storage adjustment is the same for GradC and PlateC
methods (Eq. (2)). Thus, in the energy balance study, the heat flux plate
should be placed relatively deep in the soil profile (e.g., 10 cm) to
minimize the errors of subsurface latent heat sink on G0 determination
under most field conditions.
4. Conclusions
Accurate determination of G0 is required in surface energy balance
studies. The heat flux plate method is commonly used in accomplishing
the measurement. However, the presence of heat flux plates causes
Fig. 5. Soil water evaporation rates of various soil layers determined with the
multi-needle heat pulse method.
S. Lu et al.
Fig. 6. Apparent soil heat flux at 53 mm (Ga), change in heat storage at
0–53 mm, and ground heat flux without latent heat sink (G0) determined with
the multi-needle heat pulse method.
Fig. 7. Times series of ground heat flux without latent heat sink (G0) obtained
using the multi-needle heat pulse (the GradC method) and heat flux plate (the
PlateC method) methods.
perturbations in soil heat and fluid flow and thus increases measure-
ment errors. Moreover, the plate method exhibits difficulty in de-
scribing the magnitude and location of the subsurface latent heat sink,
which must be considered in accurate G0 determination. In this study,
we presented the Ga results of the heat pulse method using the gradient
approach at various depths. In addition, heat flux plate method was
performed at a 4 mm depth to illustrate the effect of vapor transfer
disruption from the impervious plate. Ground heat flux without latent
heat sink could be obtained by employing the Ga measurement at a
deep layer with no latent heat consumption and ΔS up to the soil sur-
face using the heat pulse method. The results showed that a large error
was produced if the change in soil heat storage was ignored in the G0
calculation. Further calculation was conducted on the surface EBC in
which the turbulent energy flux measurements from the aboveground
EC system were used. In contrast to the combination of plate mea-
surements (4 mm depth) and calorimetric adjustment, the combination
method based on the heat pulse measurement significantly improved
the surface EBC. Our results suggest that the adoption of G0 measure-
ment in EBC calculation should exercise caution in excluding the effect
of subsurface latent heat sink. In summary, the heat pulse method can
improve the accuracy of G0 measurement and surface energy balance
181
Agricultural and Forest Meteorology 260–261 (2018) 176–182
Fig. 8. Available energy (Rn−G0) and the sum of turbulent fluxes of latent heat
and sensible heat (LET + H) on DOY 193. Residuals (Rn−G0−LET−H) be-
tween the available energy and the sum of turbulent fluxes of latent heat and
sensible heat are included. Rn is the net radiation, G0 is the ground heat flux
without latent heat sink, and H and LET are the surface–atmosphere sensible
heat and latent heat fluxes obtained using the eddy covariance technique, re-
spectively. GradC and PlateC indicate that G0 was obtained through multi-
needle heat pulse and heat flux plate methods, respectively.
Fig. 9. Comparison of the available energy (Rn−G0) and the sum of turbulent
fluxes of latent heat and sensible heat (LET + H) at half-hourly during the
measurement period. Rn is the net radiation, G0 is the ground heat flux without
latent heat sink, and H and LET are the surface–atmosphere sensible heat and
latent heat fluxes obtained using the eddy covariance technique, respectively.
GradC and PlateC indicate that G0 was obtained through multi-needle heat
pulse and heat flux plate methods, respectively. The solid line corresponds to
the 1:1 line.
closure.
Acknowledgements
The authors gratefully thank Dr. Ning Zheng for his technical as-
sistance. We also thank the associate editor and anonymous reviewers
for helpful comments to improve the manuscript. This research was
supported by the Fundamental Research Funds of CAF (No.
CAFYBB2017QC001) and the Natural Science Foundation of China (No.
41371240).
References
Amiro, B., 2009. Measuring boreal forest evapotranspiration using the energy balance
residual. J. Hydrol. 366, 112–118.
S. Lu et al.
Baldocchi, D., Falge, E., Gu, L., Olson, R., Hollinger, D., Running, S., Anthoni, P.,
Bernhofer, C., Davis, K., Evans, R., Fuentes, J., Goldstein, A., Katul, G., Law, B., Lee,
X., Malhi, Y., Meyers, T., Munger, W., Oechel, W., Paw U, K.T., Pilegaard, K., Schmid,
H.P., Valentini, R., Verma, S., Vesala, T., Wilson, K., Wofsy, S., 2001. FLUXNET: a
new tool to study the temporal and spatial variability of ecosystem-scale carbon di-
oxide, water vapor and energy flux densities. Bull. Am. Meteorol. Soc. 82,
2415–2434.
Chow, W.T., Volo, T.J., Vivoni, E.R., Jenerette, G.D., Ruddell, B.L., 2014. Seasonal dy-
namics of a suburban energy balance in Phoenix, Arizona. Int. J. Climatol. 34,
3863–3880.
Cobos, D.R., Baker, J.M., 2003. In situ measurement of soil heat flux with the gradient
method. Vadose Zone J. 2, 589–594.
Deol, P., Heitman, J., Amoozegar, A., Ren, T., Horton, R., 2012. Quantifying non-
isothermal surface soil water evaporation. Water Resour. Res. 48, W11503.
Evans, J.G., McNeil, D.D., Finch, J.W., Murray, T., Harding, R.J., Ward, H.C., Verhoef, A.,
2012. Determination of turbulent heat fluxes using a large aperture scintillometer
over undulating mixed agricultural terrain. Agric. For. Meteorol. 166, 221–233.
Evett, S.R., Agam, N., Kustas, W.P., Colaizzi, P.D., Schwartz, R.C., 2012. Soil profile
method for soil thermal diffusivity, conductivity and heat flux: comparison to soil
heat flux plates. Adv. Water Resour. 50, 41–54.
Foken, T., 2008. The energy balance closure problem: an overview. Ecol. Appl. 18,
1351–1367.
Foken, T., Aubinet, M., Finnigan, J.J., Leclerc, M.Y., Mauder, M., Paw, U.K.T., 2011.
Results of a panel discussion about the energy balance closure correction for trace
gases. Bull. Am. Meteorol. Soc. 92, ES13–ES18.
Gardner, H.R., Hanks, R.J., 1966. Evaluation of the evaporation zone in soil by mea-
surement of heat flux. Soil Sci. Soc. Am. Proc. 32, 326–328.
Heitman, J.L., Horton, R., Sauer, T.J., DeSutter, T.M., 2008a. Sensible heat observations
reveal soil-water evaporation dynamics. J. Hydrometeorol. 9, 165–171.
Heitman, J.L., Xiao, X., Horton, R., Sauer, T.J., 2008b. Sensible heat measurements in-
dicating depth and magnitude of subsurface soil water evaporation. Water Resour.
Res. 44, W00D. http://dx.doi.org/10.1029/2008WR006961.
Heitman, J.L., Horton, R., Sauer, T.J., Ren, T., Xiao, X., 2010. Latent heat in soil heat flux
measurements. Agric. For. Meteorol. 150, 1147–1153.
Heusinkveld, B.G., Jacobs, A.F.G., Holtslag, A.A.M., Berkowicz, S.M., 2004. Surface en-
ergy balance closure in an arid region: role of soil heat flux. Agric. For. Meteorol. 122,
21–37.
Higgins, C.W., Pardyjak, E., Froidevaux, M., Simeonov, V., Parlange, M.B., 2013.
Measured and estimated water vapor advection in the atmospheric surface layer. J.
Hydrometeorol. 14, 1966–1972.
Jury, W., Bellantuoni, B., 1976. A background temperature correction for thermal con-
ductivity probes. Soil Sci. Soc. Am. J. 40, 608–610.
Liebethal, C., Foken, T., 2006. On the use of two repeatedly heated sensors in the de-
termination of physical soil parameters. Meteorologische Zeitschrift 15, 293–299.
Liebethal, C., Huwe, B., Foken, T., 2005. Sensitivity analysis for two ground heat flux
calculation approaches. Agric. For. Meteorol. 132, 253–262.
Lindroth, A., Mölder, M., Lagergren, F., 2010. Heat storage in forest biomass improves
energy balance closure. Biogeosciences 7, 301–313.
Liu, H., Tu, G., Fu, C., Shi, L., 2008. Three-year variations of water, energy and CO2 fluxes
of cropland and degraded grassland surfaces in a semi-arid area of Northeastern
China. Adv. Atmos. Sci. 25, 1009–1020.
182
Agricultural and Forest Meteorology 260–261 (2018) 176–182
Lu, S., Ma, C., Meng, P., Zhang, J., Zhang, X., Lu, Y., Yin, C., 2016. Experimental in-
vestigation of subsurface soil water evaporation on soil heat flux plate measurement.
Appl. Therm. Eng. 93, 433–437.
Ma, J., Liu, R., Tang, L.S., Lan, Z.D., Li, Y., 2014. A downward CO2 flux seems to have
nowhere to go. Biogeosciences 11, 6251–6262.
Mayocchi, C.L., Bristow, K.L., 1995. Soil surface heat flux: some general questions and
comments on measurements. Agric. For. Meteorol. 75, 43–50.
Novak, M.D., 2010. Dynamics of the near-surface evaporation zone and corresponding
effects on the surface energy balance of a drying bare soil. Agric. For. Meteorol. 150,
1358–1365.
Ochsner, T.E., Sauer, T.J., Horton, R., 2006. Field tests of the soil heat flux plate method
and some alternatives. Agron. J. 98, 1005–1014.
Philip, J.R., 1961. The theory of heat flux meters. J. Geophys. Res. 66 (2), 571–579.
Russell, E.S., Liu, H., Gao, Z., Finn, D., Lamb, B., 2015. Impacts of soil heat flux calcu-
lation methods on the surface energy balance closure. Agric. For. Meteorol. 214–215,
189–200.
Sauer, T.J., 2002. Heat flux density. In: Dane, H., Topp, G.C. (Eds.), Methods of Soil
Analysis. Part 4. SSSA, Madison, WI pp. 1233–1248.
Sauer, T.J., Meek, D.W., Ochsner, T.E., Harris, A.R., Horton, R., 2003. Errors in heat flux
measurement by flux plates of contrasting design and thermal conductivity. Vadose
Zone J. 2, 580–588.
Sauer, T.J., Ochsner, T.E., Horton, R., 2007. Soil heat flux plates: heat flow distortion and
thermal contact resistance. Agron. J. 99, 304–310.
Stoy, P.C., Richardson, A.D., Baldocchi, D.D., Katul, G.G., Stanovick, J., Mahecha, M.D.,
Reichstein, M., Detto, M., Law, B.E., Wohlfahrt, G., Arriga, N., Campos, J.,
McCaughey, J.H., Montagnani, L., Paw U, K.T., Sevanto, S., Williams, M., 2009.
Biosphere-atmosphere exchange of CO2 in relation to climate: a cross-biome analysis
across multiple time scales. Biogeosciences 6, 2297–2312.
Stoy, P.C., Mauder, M., Foken, T., Marcolla, B., Boegh, E., Ibrom, A., Altaf Arain, M.,
Arneth, A., Aurela, M., Bernhofer, C., Cescatti, A., Dellwik, E., Duce, P., Gianelle, D.,
van Gorsel, E., Kiely, G., Knohl, A., Margolis, H., McCaughey, H., Merbold, L.,
Montagnani, L., Papale, D., Reichstein, M., Saunders, M., Serrano-Ortiz, P.,
Sottocornola, M., Spano, D., Vaccari, F., Varlagin, A., 2013. A data-driven analysis of
energy balance closure across FLUXNET research sites: the role of landscape scale
heterogeneity. Agric. For. Meteorol. 171–172, 137–152.
Wang, Z., Bou-Zeid, E., 2012. A novel approach for the estimation of soil ground heat flux.
Agric. For. Meteorol. 154–155, 214–221.
Wang, Z., Xiao, X., Yan, X., 2010. Modeling gross primary production of maize cropland
and degraded grassland in northeastern China. Agric. For. Meteorol. 150, 1160–1167.
Webb, E.K., Pearman, G.I., Leuning, R., 1980. Correction of flux measurements for density
effects due to heat and water vapour transfer. Q. J. R. Meteorol. Soc. 106, 85–100.
Welch, S.M., Kluitenberg, G.J., Bristow, K.L., 1996. Rapid numerical estimation of soil
thermal properties for a broad class of heat pulse emitter geometries. Meas. Sci.
Technol. 7, 932–938.
Wilson, K., Goldstein, A., Falge, E., Aubinet, M., Baldocchi, D., Berbigier, P., Bernhofer,
C., Ceulemans, R., Dolman, H., Field, C., Grelle, A., Ibrom, A., Law, B.E., Kowalski, A.,
Meyers, T., Moncrieff, J., Monson, R., Oechel, W., Tenhunen, J., Valentini, R., Verma,
S., 2002. Energy balance closure at FLUXNET sites. Agric. For. Meteorol. 113,
223–243.
Zhang, X., Heitman, J.L., Horton, R., Ren, T., 2012. Measuring subsurface soil-water
evaporation with an improved heat-pulse probe. Soil Sci. Soc. Am. J. 76, 876–879.