The surface-atmosphere exchange of 
carbon dioxide, water, and sensible heat 
across a dryland wheat-fallow rotation
Authors: Elizabeth S.K. Vick, Paul C. Stoy, Angela 
C.I. Tang, and Tobias Gerken 
NOTICE: this is the author’s version of a work that was accepted for publication in Agriculture, 
Ecosystems & Environment. Changes resulting from the publishing process, such as peer 
review, editing, corrections, structural formatting, and other quality control mechanisms may not 
be reflected in this document. Changes may have been made to this work since it was 
submitted for publication. A definitive version was subsequently published in Agriculture, 
Ecosystems & Environment, [VOL# 223 (September 2016)] DOI#: 10.1016/j.agee.2016.07.018 
Vick, Elizabeth S. K., Paul C. Stoy, Angela C. I. Tang, and Tobias Gerken. "The surface-
atmosphere exchange of carbon dioxide, water, and sensible heat across a dryland wheat-
fallow rotation." Agriculture, Ecosystems & Environment 223 (September 16, 2016): 129-140. 
DOI: 10.1016/j.agee.2016.07.018.
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
The surface-atmosphere exchange of carbon dioxide, water, and
sensible heat across a dryland wheat-fallow rotation
Elizabeth S.K. Vicka, Paul C. Stoya,*, Angela C.I. Tanga, Tobias Gerkena,b
aDepartment of Land Resources and Environmental Sciences, Montana State University, Bozeman, MT 59717, USA
bDepartment of Meteorology, The Pennsylvania State University, University Park, PA 16802, USA
A B S T R A C T
Summer fallow – the practice of keeping a field out of production during the growing season – is a
common practice in dryland wheat (Triticum aestivum L.) cropping systems to conserve soil water
resources. Fallow also depletes soil carbon stocks and thereby soil quality. The area of summer fallow has
decreased by tens of millions of hectares since the 1970s in the northern North American Great Plains as
producers have recognized that avoiding fallow usually confers both economic and soil conservation
benefits. Observed summertime cooling across parts of this region has coincided with fallow reduction,
suggesting that the role of fallow in atmospheric processes needs to be ascertained. We measured carbon
dioxide, latent heat, and sensible heat flux across a winter wheat – spring wheat – fallow sequence in
Montana, USA to determine the effects of dryland crop management on ecosystem carbon resources and
energy partitioning at the surface-atmosphere interface. Winter wheat and spring wheat fields were
carbon sinks (Fc = !203 " 52 g C!!CO2m!2 and !107 " 29 g C!!CO2m!2), respectively, during the April to
September study period, but the fallow field was a carbon source of 135 " 73 g C!!CO2m!2.
Evapotranspiration in the wheat crops was over 100 mm greater than the 275 " 39 mm observed in
the fallow field during the study period. Modeled maximum daily atmospheric boundary layer height
was on average 210 m higher and up to 900 m higher in fallow compared to the spring wheat field with
more crossings of the modeled atmospheric boundary layer and lifted condensation level, suggesting that
regional studies of the effects of fallow on near-surface temperature and moisture are necessary to
understand the effects of fallow reduction on regional climate dynamics. Results demonstrate that fallow
has a detrimental impact to soil carbon resources yet is less water intensive, with consequences for
regional climate via its impacts on atmospheric boundary layer development and global climate via its
carbon metabolism.
1. Introduction
Wheat (Triticum aestivum L.) provides more than 20% of the
calories and protein for the global population (Hawkesford et al.,
2013) and nearly 220 Mha of wheat was harvested globally in 2013
(FAO, 2013). Wheat therefore plays a central role in not only global
food production, but also in the global exchange of water, energy,
and climate-relevant trace gases like carbon dioxide between the
land surface and the atmosphere (West and Marland, 2002;
Buyanovsky and Wagner, 1998).
Our understanding of the interaction between wheat cropping
systems and the atmosphere remains incomplete. Wheat has a
very low canopy resistance to water vapor transport during its
main growth period (Bonan, 2008), and models tend to accurately
simulate latent heat flux (LE, see Table 1 for a list of abbreviations),
and sensible heat flux (H) during these periods (Ingwersen et al.,
2011). Aspects of the seasonal timing of crop development
including ripening (Ingwersen et al., 2011) and management
decisions like harvesting (Sus et al., 2010) on surface-atmosphere
fluxes continue to challenge ecosystem models. Management
practices including crop rotations have been identified as
important contributors to carbon metabolism at the field scale
(Béziat et al., 2009; Schmidt et al., 2012), but have largely been
studied in winter wheat crops and/or mesic cropping systems to
date (e.g. Billesbach et al., 2014; Moureaux et al., 2008). Surface-
atmosphere exchange in rotations common to dryland cropping
systems including spring wheat and chemical fallow (hereafter
‘fallow’) have been studied less-frequently to date.
Fallow is a common management practice in the dryland
wheat-growing regions of the northern North American Great
Plains to conserve water for subsequent crops (Lubowski et al.,
2006). Fallow however also increases erosion (Wischmeier, 1959)
and soil carbon loss (Cihacek and Ulmer, 1995), and fallow-small
grain management strategies are not considered sustainable from
the soil conservation perspective (Merrill et al., 1999). Manage-
ment practices are changing. The area of fallow in the Prairie
Provinces of Canada has decreased from over 15 Mha in the 1970s
to under 2 Mha at the present (Fig. 1) as producers have realized
that the water-savings benefit of fallow is outweighed by the
economic losses of not planting (Dhuyvetter et al., 1996). The area
under fallow in the United States has likewise decreased from
16 Mha to 6 Mha across the same time frame (Lubowski et al.,
2006), largely in the northern Great Plains and other areas of the
semiarid West (Fig. 1). Despite the decreasing trend in fallow area
across the North American northern Great Plains, fallow remains
common in many regions including major land resource area
(MLRA) 52 in north-central Montana – the largest wheat-growing
region in the state – where some 40% of agricultural lands may
remain in fallow in any given year. In contrast, fallow has been
reduced in northeastern Montana (MLRA 53) by hundreds of kha
over the past decade (Long et al., 2014, 2013) as producers have
adopted continuous cropping or alternate cropping practices
(Burgess et al., 2012; Miller et al., 2003, 2002).
The widespread decline of fallow in agricultural areas of the
Canadian Prairie Provinces (Fig. 1) has coincided with a summer-
time cooling trend since the 1970s (Betts et al., 2013a, 2013b;
Gameda et al., 2007; Mahmood et al., 2014). Extreme temperature
events now occur less frequently than in the recent past, maximum
summer temperatures have decreased by ca. 2 #C, relative
humidity has increased by some 7% (Betts et al., 2013b), and
summer precipitation has increased by an average of 10 mm/
decade across parts of the Canadian Prairie Provinces (Gameda
et al., 2007). A remarkable 6 W m!2 summer cooling has been
observed (Betts et al., 2013a); for reference, anthropogenic
greenhouse gasses are responsible for a ca. 2.5 W m!2 warming
globally since the dawn of the Industrial Era (IPCC, 2007). These
climate benefits have only occurred during the growing season;
fall, winter, and early spring temperatures have followed global
trends (Betts et al., 2013b).
Studies suggest that these climate changes are the result of land
management, specifically the trend away from leaving fields fallow
Table 1
A list of abbreviations with their definitions.
Abbreviation Definition
a Initial slope of the light response curve
aPT Priestley-Taylor coefficient
b Gross ecosystem productivity at light saturation
g Psychrometric constant
g* Virtual temperature inversion strength
Dh Change in height of the atmospheric boundary layer
ePT Intercept term of the Priestley-Taylor-type model
l Latent heat of vaporization
ra Density of air
ABL Atmospheric boundary layer
cp Specific heat capacity
E0 Activation energy
ET Evapotranspiration
Fc Carbon dioxide flux
G Soil heat flux
GEP Gross ecosystem productivity
h Height of the atmospheric boundary layer
hLCL Height of the lifting condensation level
H Sensible heat flux
Hv Virtual heat flux
LE Latent heat flux
P Atmospheric surface pressure
r Water vapor mixing ratio
RE Ecosystem respiration
Rn Net radiation
R10 Ecosystem respiration at an air temperature of 10 #C
s Slope of the vapor pressure-temperature relationship at saturation
SWin Incident shortwave radiation
Ta Air temperature
T10 Air temperature of 10 #C
TLCL Air temperature at the lifting condensation level
Tt Reference temperature (227.13 K)
u* Friction velocity
Year
1920 1940 1960 1980 2000
Su
m
m
er
 fa
llo
w 
(M
ha
)
0
2
4
6
8
10
12
14
16 Canada
United States
Fig. 1. Trends in summer fallow area from 1920 until the present in Canada (red
dashed line) and the United States (blue solid line) using data from Statistics Canada
and the United States Department of Agriculture Economic Research Service. (For
interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
during summer (Betts et al., 2013a; Gameda et al., 2007; Mahmood
et al., 2014). These large-scale declines in fallow area have resulted
in less sensible heat and more water vapor entering the
atmosphere and a moister, shallower atmospheric boundary layer
(ABL) and lower lifted condensation level (LCL) with an increased
probability of convective precipitation (Betts et al., 2013b; Gameda
et al., 2007). In other words, the observed regional climate cooling
is broadly consistent with the effects of fallow avoidance on
climate processes.
Using a conceptual model, Gameda et al. (2007) suggested that
fallow supports an ABL height (h) on the order of 2–3 km, but that h
is only some 0.5 km over crops. These estimates have not been
supported by observations or models, in part because the impacts
of field-scale management changes on H and LE, which help
determine h, have not been studied to date in wheat-fallow
rotations. Doing so is an important step in understanding the role
of human management on climate processes in the North
American Northern Great Plains.
Cropping systems that include wheat also influence land
surface biogeochemistry. Many studies to date have used the
eddy covariance technique to find that wheat tends to be a strong
carbon sink at the field scale during the growing season, but is
infrequently a net annual carbon sink. For example, Anthoni et al.
(2004a,b) found that a central European winter wheat field was a
sink for atmospheric carbon dioxide during the growing season,
but a source of carbon annually, with a net biome productivity of
45–105 g C m!2 y!1, indicating a loss of C to the atmosphere when
the end products of the wheat crop were accounted for. Crop
rotations are also important for carbon metabolism. Béziat et al.
(2009) measured the carbon balance of a three-crop rotation and
found that the net biome productivity was !161 "66 g C m!2 y!1 in
winter wheat, but a maize crop at the same site a year earlier was a
carbon source of 372 " 78 g C m!2 y!1 due in part to residue from
previous crops. These observations emphasize the importance of
studying crop rotations for a comprehensive understanding of the
impacts of agriculture on surface-atmosphere exchange.
Here, we use the eddy covariance technique to measure the
surface-atmosphere exchange of carbon dioxide (Fc), LE, and H
across a dryland winter wheat – spring wheat – fallow sequence in
central Montana, USA. We quantify differences in surface fluxes
during the April–September growing period and use models of h
and the height of the lifted condensation level (hLCL) driven by
observed surface-atmosphere energy flux and near-surface air
temperature and humidity to help us understand the physical
mechanisms that underlie the observed effects of agricultural
management on ABL, LCL, and regional climate processes (Gameda
et al., 2007). We additionally discuss the impacts of wheat and
fallow on carbon and water resources to provide information to
producers for land management decisions.
2. Materials and methods
2.1. Site description
Eddy covariance and micrometeorological measurements were
made in two fields in central Montana, USA. Measurements were
made on the first field during 2013 when it was planted with
winter wheat, and during 2014 when it was planted with spring
Fig. 2. A map of the winter wheat (WW 2013) and spring wheat (SW 2014) tower location and the fallow (2014) tower location in central Montana, USA.
(Image date: 7/25/2014, Google Earth).
wheat. Measurements in the second field were made during 2014
when it was held in fallow.
The first field was planted with winter wheat in autumn, 2012
following a year of chemical fallow. In March 2013, a 3 m tower
with eddy covariance and micrometeorological instrumentation
was installed in this field at 46# 590 41.100 N,109# 360 49.500 W (Fig. 2).
Eddy covariance instrumentation was mounted at 1.8 m with a
fetch of 200 m from the nearest field edge to the west. The winter
wheat crop was harvested on August 7, 2013 and instruments were
left on the tower until September 2013, when they were removed
for calibration and reinstalled. The tower was moved to
accommodate the planting of spring wheat on May 5, 2014 and
it measured spring wheat during the 2014 growing season. It was
removed in September 2014 following the spring wheat harvest on
August 18, 2014.
A second tower of identical height and instrumentation height
was installed during spring, 2014, at 46# 590 44.800 N, 109# 370 46.300
W in a no-till chemical fallow field that was planted with spring
wheat the previous year. Chemical treatments were applied to
discourage weedy growth during the study period, with the
exception of the immediate vicinity of the tower to avoid damaging
sensors. The fallow field was two fields to the west of the first
(wheat) field, and the fallow field tower was 1.2 km from the wheat
field tower (Fig. 2). The minimum fetch was 200 m to the
southwest. For the purposes of this study, measurements from
all fields are studied for the April-September period – hereafter the
‘study period’ – that encompasses the main growing season.
Mean annual temperature and precipitation were 6.6 #C and
388 mm, respectively, in Moccasin, MT where the Montana State
University Central Agricultural Research Center is located some
25 km from the study fields. The slope of both fields was less than
0.5# on Judith and Danvers clay loams.
2.2. Measurements
2.2.1. Meteorological measurements
Micrometeorological variables were measured at both towers
as summarized in Table 2. Canopy heights and snow depths were
measured using a SR50A-L sonic depth sensor (Campbell Scientific
Inc., Logan, UT, USA) at the wheat tower. Incident shortwave (SWin),
outgoing shortwave, incident longwave, outgoing longwave
radiation, and thereby net radiation (Rn) were measured at both
towers using NR01 four-component net radiometers (Hukseflux,
Delft, The Netherlands) mounted at 1.8 m above the soil surface. Air
temperature (Ta) and relative humidity were measured using
HMP45C temperature and relative humidity probes (Vaisala,
Vantaa, Finland) at 2 m. Soil heat flux (G) was measured using
self-calibrating HFP01 heat flux plates (Hukseflux) at 5 cm below
the soil surface. Soil moisture was measured at 5 cm and 10 cm
using two CS616 time domain reflectometer sensors (Campbell
Scientific) in the wheat field and two CS650 sensors (Campbell
Scientific) in the fallow field. Soil temperature was measured at
multiple depths using copper-constantan thermocouples in the
wheat field and the CS650 sensors in the fallow field. Measure-
ments were made every minute and half-hour averages were
stored using CR3000 and CR1000 data loggers (Campbell
Scientific).
2.2.2. Turbulent flux measurements
Fc and LE were measured using the eddy covariance technique.
This involved the coupling of CSAT-3 sonic anemometers (Camp-
bell Scientific) with LI-7200 CO2/H2O enclosed infrared gas
analyzers (LiCor, Lincoln, NE) installed 1.8 m above the ground
surface on both towers. H was measured using the CSAT-3 sonic
anemometers. Data were recorded at 10 Hz on CR3000 data loggers
(Campbell Scientific) and processed into half-hourly flux sums
using EddyPro (LiCor). Eddy covariance processing steps involved
double rotation rather than the planar fit method given the rapid
changes in wheat canopy height (Kaimal and Finnegan, 1994),
covariance maximization for time lag detection (Mauder and
Foken, 2004), and plausibility ranges of five standard deviations
from the mean for vertical velocity and 3.5 standard deviations
from the mean for CO2 and H2O concentrations. Negative values
refer to fluxes from the atmosphere to the surface following the
micrometeorological convention.
Surface-atmosphere fluxes are often underestimated by eddy
covariance measurements, especially during periods of insufficient
turbulent exchange that usually occur at night (Aubinet et al.,
2000; Falge et al., 2001a; Gu et al., 2005; Papale et al., 2006;
Reichstein et al., 2005). A friction velocity (u*) filter following
Reichstein et al. (2005) was applied to the data set to identify and
filter periods of insufficient turbulence. Fc measurements made
during night, defined as periods for which the solar zenith angle
exceeded 90#, were binned into six Ta classes for three month
periods and further binned into twenty u* classes. The u* filter was
chosen to be the value at which the mean value of the Fc
observations in a given u* class first exceeded 95% of the remaining
Fc observations at higher values of u*. u* threshold values
determined using this approach exhibited trivial differences (not
shown) from a statistical approach in which the u* threshold was
determined as the u* class at which the Fc observations were not
statistically different than Fc observations at higher u* as
determined by a one-sided t-test following Papale (2012).
Nighttime Fc observations that occurred under insufficient u*,
and Fc observations that exceeded logical upper bounds of
20 mmol m!2 s!1 and lower bounds of !35 mmol m!2 s!1 deter-
mined using probability distribution functions were excluded from
the data record.
2.3. Gapfilling and data processing
Missing data occurred due to power outages and equipment
malfunction, as well as damage to wires by animals. Missing SWin,
Ta, relative humidity, and soil temperature data were gapfilled
using linear regression from the neighboring tower if data were
available. If data from the neighboring tower were also missing,
data from the Moccasin Soil Climate Analysis Network (SCAN) site,
located 25 km from the study fields, were used in the gapfilling
procedure.
2.3.1. Gapfilling soil heat flux and net radiation
Missing G observations were gapfilled by establishing a linear
relationship with Rn on a daily basis, relating the resulting slope
and intercept parameters with canopy height, and gapfilling using
regression parameters that varied as a function of canopy height. If
measured Rn data were not available from either tower, a linear
relationship for each day between tower Rn and SWin from the
Moccasin SCAN site was established and used for gapfilling Rn,
which in turn was used to gapfill G.
Table 2
Variables measured at each site (see Table 1) along with sensor type. W: Wheat
(winter wheat in 2013, spring wheat in 2014), F: Fallow (2014).
Measurement Sensor(s) Site
SWin NR01 net radiometer W, F
Rn NR01 net radiometer W, F
Canopy height SR50 sonic distance sensor W
Ta and relative humidity HMP-50 temperature/relative humidity probe W, F
G HFP01 heat flux plate W, F
H CSAT-3 W, F
LE CSAT-3 and LI-7200 W, F
Fc CSAT-3 and LI-7200 W, F
2.3.2. Gapfilling carbon dioxide flux
Our gapfilling model for Fc needed to simultaneously account
for photosynthetic and respiratory processes due to rapid changes
in canopy biomass during the growth period and for non-vegetated
conditions before plant growth and after harvest. To gapfill Fc, a
rectangular hyperbolic model for ecosystem production and the
respiration model of Lloyd and Taylor (1994) were combined:
Fc ¼ ! abSWinaSWin þ bþ R10exp E0
1
T10 ! Tt !
1
Ta ! Tt
! "! "
ð1Þ
where a is the initial response of the light response curve (mmol
CO2 J!1), b is gross ecosystem productivity (GEP) at light saturation
(mmol CO2 m!2 s!1), R10 (mmol CO2 m!2 s!1) is ecosystem
respiration (RE) at a Ta of 10 #C (T10), E0 is the activation energy
parameter (K), and the reference temperature Tt is set to 227.13 K
following Falge et al. (2001b). Parameters were fit using least
squares regression for observations within a seven-day moving
window of Fc, SWin, and Ta observations. Parameter sets during
periods for which the parameter estimation routine did not
converge to an optimal solution or for which observations were not
available were estimated using the previous day’s values for the
case of small gaps of two days or less, and using linear interpolation
of parameter values from preceding and subsequent periods for
large gaps of more than two days.
2.3.3. Gapfilling latent heat flux
Evapotranspiration (ET) was calculated as LE divided by the
latent heat of vaporization with units set to equal mm per half hour
eddy covariance measurement period. Missing LE (and thereby ET)
observations, including those that exceeded logical bounds of
!100 W m!2 and 600 W m!2, were gapfilled using the Priestley-
Taylor model (Priestley and Taylor, 1972) adjusted to include an
uncertainty term, ePT, that functions as an intercept parameter:
ET0 ¼ 1l
s Rn ! Gð Þ
s þ g aPT þ ePT : ð2Þ
Here, l is the latent heat of vaporization, s is the slope of the
saturation vapor pressure-temperature relationship, g is the
psychrometric constant, and aPT is the Priestley-Taylor coefficient.
aPT and ePT were determined for the winter wheat, spring wheat,
and fallow fields individually for each day of the study period using
linear regression against measured ET. aPT and ePT for days during
which insufficient data were available were estimated using the
previous day’s parameters for the case of small gaps of two days or
less, and linear interpolation from the parameters of preceding and
subsequent periods for large gaps of more than two days.
2.3.4. Gapfilling sensible heat flux
Missing H values, including those that exceeded logical bounds
of !100 W m!2 and 500 W m!2, were gapfilled by fitting a daily
linear regression between observed H and Rn. The linear model
included a slope and an intercept parameter similar to equation 2,
and missing H values were filled using the results of this
regression. As with the other flux gapfilling routines, parameters
which were unable to be calculated due to missing data were
estimated using the previous day’s parameters for the case of small
gaps of two days or less, and linear interpolation from the
parameters of preceding and subsequent periods for large gaps of
more than two days.
2.4. Uncertainty analysis
Uncertainty in eddy covariance observations is often on the
order of 10–15% (e.g. Goulden et al., 1997), and varies as a function
of flux magnitude (Richardson et al., 2008, 2006). The total
uncertainty in eddy covariance measurements is a function of
observational uncertainty (Moncrieff et al., 1996), gapfilling
uncertainty (Falge et al., 2001a, 2001b), and spatial uncertainty
due to the assumption that a spatially variable flux can be
expressed as an average on a square meter basis (Oren et al., 2006).
We applied the approach of Richardson et al. (2008) to estimate
the random uncertainty of Fc, LE, and H. Briefly, data from the same
half hour period of consecutive days were screened for similar
micrometeorological conditions defined as those for which Rn
differed by less than 75 W m!2, air temperature differed by less
than 3 #C, and average wind speed differed by less than 1 m s!1
under the assumption that fluxes measured during these
conditions on consecutive days should be similar and any
differences are related to the random error of eddy covariance
measurements. Differences in Fc, LE, and H identified using this
approach were found to be linearly related to the mean flux
(Richardson et al., 2008, 2006), the model for which is taken to be
the random uncertainty of the flux measurements. Random
uncertainty of the seasonal sums of Fc, LE, and H was taken to
be the mean absolute value of these fluxes multiplied by the
percent random uncertainty following Stoy et al. (2006a,b).
Random uncertainty was propagated through the gapfilling
routines following the recommendations of Motulsky and Ransnas
(1987) by perturbing the input flux observations with a random
value drawn from a normal distribution multiplied by the random
uncertainty calculated using the Richardson et al. (2008) routine.
This procedure was repeated 100 times for each day of
observations, and 100 parameter sets for the gapfilling model
were subsequently fit for each day using least squares regression.
Each of the corresponding parameter sets for the case of Fc were
tested for convergence of the optimization routine, and all
parameter sets were adjusted for missing values using the routines
for large and small gaps described above. Gapfilling uncertainty
was then determined as the standard deviation of the April–
September seasonal flux sum. Total uncertainty for each flux over
the study periods was calculated by summing the variance
attributable to random uncertainty and that attributable to
gapfilling parameter uncertainty – noting the propagation of
random error into the estimate of gapfilling uncertainty – in order
to obtain a conservative value of total flux uncertainty for each field
for each study period.
2.5. Energy balance
The surface-atmosphere energy balance is rarely closed using
the eddy covariance and micrometeorological measurements of Rn,
G, LE, and H at a single tower (Franssen et al., 2010; Stoy et al., 2013;
Wilson et al., 2002), although 26 of the 173 tower sites studied by
Stoy et al. (2013) had an energy balance closure greater than 100%.
Energy balance closure for the winter wheat, spring wheat, and
fallow fields was calculated as the percent of the sum of available
energy during the April–September study periods (i.e. Rn! G) that
is realized as the measured sum of sensible and latent heat fluxes
(H + lE), excluding periods for which turbulent fluxes were
gapfilled.
2.6. Atmospheric boundary layer height model
We used surface-atmosphere exchange observations and a
simple one-dimensional ABL model to test the conceptual model of
Gameda et al. (2007), which assumes that fallow results in h on the
order of 2–3 km and vegetated surfaces result in h on the order of
500 m in the Canadian Prairies. h was modeled for each of the
management practices following Luyssaert et al. (2014):
Dh ¼ HvDtÞ=racphgv
# ð3Þ
where Dh is the change in h in meters per unit time (Dt), ra is the
air density in kg m!3, cp is the specific heat capacity in J kg!1 K!1, gv
is the virtual temperature inversion strength in K m!1 set to
0.003 K m!1, and Hv is the virtual heat flux in W m!2 obtained by
measured H and LE:
Hv ¼ H þ 0:07LE: ð4Þ
Eq. (3) was calculated for each day during the study periods and h
was set to 150 m at the beginning of each day, also following
Luyssaert et al. (2014). We focus our comparisons on model
predictions of maximum daily h for the different agricultural
treatments.
Modeled h is compared to the modeled height of the lifted
condensation level (hLCL), which was calculated by dry-adiabati-
cally raising an air parcel with near-surface temperature and
moisture to its saturation point approximated by (Stull, 2012):
TLCL ¼ 2840
3:5ln Tað Þ ! ln Pr0:622þr
$ %
! 7:108
þ 55 ð5Þ
with atmospheric pressure at the surface (P) in kPa, the mixing
ratio r, and Ta in K. We focus our discussion of h and hLCL on the 2014
growing season when spring wheat was measured by the first
tower and fallow was measured by the second tower.
3. Results
3.1. Prevailing weather conditions
On average, 2014 was cooler, cloudier, and wetter than 2013.
Mean annual Ta during 2013 was 7.0 #C and mean April-September
Tawas 14.3 #C. Mean annual Ta (6.5 #C) and April–September mean
Ta (13.7 #C) were about 0.5 #C cooler in 2014 than 2013. Annual
cumulative SWinwas 3% greater in 2013 (5290 MJ m!2 year!1) than
in 2014 (5140 MJ m!2 year!1), and the April-September cumulative
incident SWin was 5% greater in 2013. Cumulative precipitation
measured at the Moccasin SCAN site was 452 mm in 2013 and
503 mm in 2014. The study period cumulative precipitation was
360 mm in 2013 and 391 mm in 2014,162 mm of which fell during a
large precipitation event during Aug. 21–24 shortly after the
harvest of the spring wheat field on Aug. 18, 2014.
3.2. Plant height growth
Canopy height was near zero after the 2012 fallow growing
season that preceded winter wheat planting. Measureable winter
wheat growth began in early May 2013 (Fig. 3). The winter wheat
crop reached its maximum height of 0.64 m on July 9, 2013. The
canopy was reduced to stubble measured at 0.2 m after harvest on
August 7, 2013. With the exception of variation due to snow cover,
the plant canopy remained at 0.2 m during the 2013/2014 winter,
and was effectively 0 m after spring wheat planting on May 2, 2014.
Measurable spring wheat crop growth began on May 26 and
reached 0.83 m on July 25, ca. two weeks after winter wheat
reached its maximum height in 2013. Canopy height was likewise
was reduced to 0.2 m stubble after harvest on Aug. 18, 2014 (Fig. 3).
3.3. u* Thresholds and data availability
The Reichstein et al. (2005) algorithm selects a unique u*
threshold for three month periods, which correspond to April
through June and July through September during the study period.
The identified u* thresholds were 0.077 m s!1 for 2013 winter
wheat, 0.048 m s!1 for 2014 spring wheat, and 0.073 m s!1 for the
fallow field during the April through June period, and 0.054 m s!1
for 2013 winter wheat, 0.092 m s!1 for 2014 spring wheat, and
0.057 m s!1 for the fallow field during the July through September
period. Turbulent flux observations taken during periods where
measured u* was less than these values during nighttime were
removed from the observational record and filled using the
gapfilling procedures described in Sections 2.3.2–2.3.4. The
percent of missing data for each treatment that resulted from
the u* filter and logical filters, as well as power system and
instrumental failures, is described in Table 3. Uncertainty in flux
sums that result from missing data is propagated through the data
gapfilling routines to result in monthly and seasonal uncertainty
estimates as described in Section 2.4.
3.4. Carbon dioxide fluxes
The cumulative sum of Fc in units of carbon in CO2 (C!!CO2) in
all fields during the study period is shown in Fig. 4A. The 2013
winter wheat field became a net sink of carbon (since April 1, 2013)
around May 8, 2013, and the 2014 spring wheat field became a net
sink of carbon (since April 1, 2014) around May 23, 2014. The
winter wheat field had a cumulative C!!CO2 uptake of !208 " 53
g C m!2 during the 2013 April ! September study period
(expressed as the mean of the gapfilling model simulations plus
or minus one standard deviation of random plus gapfilling
uncertainty), and the spring wheat field took up !107 " 29 g C m!2
during the 2014 study period (Fig. 4B). The fallow field lost
135 " 73 g C!!CO2 to the atmosphere during the 2014 study period
with two brief periods of net C!!CO2 uptake of ca. 10–20 g C m!2
corresponding to periods of weedy growth near the tower during
late June and weedy growth following the large rain event after the
spring wheat harvest in August. The larger uncertainty for Fc in the
fallow field is attributable to greater uncertainty in net respiratory
April May June July Aug Sept
Ca
no
py
 H
ei
gh
t (m
)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Winter wheat (2013)
Spring wheat (2014)
Winter wheat harvest
Spring wheat harvest
Fig. 3. The daily maximum canopy height or snowpack height for the winter wheat
and spring wheat rotations measured by a sonic distance sensor (Table 2). Canopy
height was assumed to be near zero before winter wheat growth began in earnest in
early May, 2013. Stubble left on the winter wheat field in 2013 measured ca. 0.2 m,
and was assumed to drop to 0 m following the planting of spring wheat on May 2,
2014.
Table 3
The percent of available turbulent flux data during the entire April–September
study period for each study site. Daytime refers to periods when the solar zenith
angle was less than 90# .
Site Period Fc LE H
Winter wheat (2013) Daytime 52% 52% 53%
Total 48% 49% 49%
Spring wheat (2014) Daytime 48% 48% 48%
Total 45% 45% 45%
Fallow (2014) Daytime 44% 50% 51%
Total 42% 49% 50%
fluxes calculated by the Richardson et al. (2008) algorithm, and the
larger uncertainty in Fc at the winter wheat field versus the spring
wheat field is due in part to the larger uncertainties attributable to
larger average flux magnitude.
Maximum monthly Fc occurred at the winter wheat field in June
at !186 " 47 g C!!CO2 month!1 (Fig. 4B) and in July at the spring
wheat field at !113 " 30 g C!!CO2 month!1, which is of greater
magnitude than Fc in June at the spring wheat field at !100 " 27 g
C!!CO2 month!1 as identified by a two-sided t-test (p < 0.05,
Fig. 4B). Monthly Fc from the fallow field peaked in July at 39 " 22 g
C!!CO2 month!1.
3.5. Evapotranspiration
Cumulative ET from the 2013 winter wheat field was
410 " 36 mm during the study period (Fig. 5A), significantly less
than the 437 "44 mm observed for the 2014 spring wheat field
during the cooler and wetter 2014 (p < 0.05), but only by some 6%
and reflective of the differences in April-September precipitation
observed between years. The rate of increase in cumulative ET in
the 2014 fallow field was not as steep as that in the winter wheat
and spring wheat fields, and the Apr.–Sept. total reached
275 " 39 mm, ca. 2/3 of observed ET from the spring wheat field.
Monthly total ET from the winter wheat field peaked in June at
130 " 10 mm and monthly total ET from the spring wheat field
peaked in July at 147 " 10 mm (Fig. 5B), mirroring the monthly
patterns of Fc uptake (Fig. 4B). Monthly total ET in the Fallow
treatment peaked in June and July at ca. 69 " 10 mm.
3.6. Sensible heat fluxes and energy balance closure
H for all treatments before 2013 winter wheat and 2014 spring
wheat growth began was similar at ca. 2.5 MJ m!2 day!1. After crop
growth initiated, the increase in cumulative H diverged between
the spring wheat and fallow fields (Fig. 6A). The winter wheat and
spring wheat fields had similar totals of H of 445 " 66 and
465 " 67 MJ m!2 for the respective study periods. H at the fallow
field reached 759 " 98 MJ m!2 for the study period, an increase in H
proportional to the decrease in ET (Figs. Fig. 55A and Fig. 66A).
Monthly H was greater at the fallow field than the wheat fields
until harvest in August, especially from the winter wheat field
which ceased growth and was harvested sooner than the spring
wheat field (see e.g. Fig. 6B). Energy balance closure during the
study period was calculated to be 86% at the winter wheat field,
105% at the spring wheat field, and 107% at the fallow field.
3.7. Modeled atmospheric boundary layer height and lifted
condensation level
Modeled h using surface flux data from the spring wheat and
fallow fields that were both measured in 2014 (Fig. 7A) demon-
strates a seasonal pattern that follows the seasonal trend in H
(Fig. 6B), which is the largest contributor to Hv (Eq. (4)). Maximum
modeled h was 2070 m for the spring wheat field and 2440 m for
the fallow field (Fig. 7A), and the maximum daily difference
between the two fields was over 900 m (Fig. 7B). The mean
difference in modeled maximum daily h between the spring wheat
April May June July Aug Sept
evitalu
mu
C
F c
m 
C g(
-2
)
-300
-200
-100
0
100
200 A
Winter wheat (2013)
Spring wheat (2014)
Fallow (2014)
Winter wheat harvest
Spring wheat harvest
April May June July Aug Sept
F c
m 
C g(
-2
htno
m
-1
)
-250
-200
-150
-100
-50
0
50
B
Winter wheat (2013)
Spring wheat (2014)
Fallow (2014)
Fig. 4. The cumulative (A) and monthly (B) sums of the net ecosystem exchange of
carbon in CO2 (Fc) between the winter wheat, spring wheat, and fallow fields and
the atmosphere during the April-September study period. Negative values denote
carbon uptake by the biosphere. Error bars represent one standard deviation from
the cumulative or monthly sum.
April May June July Aug Sept
ET
htno
m
m
m(
-1
)
0
20
40
60
80
100
120
140
160
180
BWinter wheat (2013)
Spring wheat (2014)
Fallow (2014)
Fig. 5. The cumulative (A) and monthly (B) sums of evapotranspiration (ET) from
the winter wheat, spring wheat, and fallow study fields during the the April–
September study periods. Error bars represent one standard deviation from the
cumulative or monthly sum.
and fallow fields was 210 m during the April–September study
period.
Differences in modeled hLCL between fallow and spring wheat
were minor and averaged 20 m over the study period. As a
consequence, hLCL modeled using data from the spring wheat and
fallow fields were averaged for subsequent analyses. The mean
value of maximum daily hLCL was ca. 1600 m but exceeded 3500 m
during the month before the spring wheat harvest (Fig. 7A). A
comparison of modeled h and hLCL for both spring wheat and fallow
indicates that the growing convective ABL height reaches the LCL
(h > hLCL) on average some 30 min earlier for fallow, and such
crossings occurred during 81 days with fallow and 74 days with
spring wheat (Fig. 8).
4. Discussion
It is important to recall that the 2014 study period was cooler
and wetter than 2013 and comparisons between the winter wheat
field (measured in 2013) and the spring wheat and fallow fields
(measured in 2014) are subject to these differences in prevailing
weather conditions. We first discuss differences in surface-
atmosphere carbon dioxide exchange with a focus on the winter
wheat and spring wheat fields – noting that the fallow field lost
carbon to the atmosphere as expected – followed by an analysis of
the impacts of spring wheat versus fallow management on the
turbulent fluxes of latent and sensible heat and consequences for
modeled atmospheric boundary layer development.
4.1. Carbon dioxide exchange
The spring wheat and fallow fields – both measured during
2014 – lost a similar amount of C!!CO2 during April and May before
crop growth and during initial stages of spring wheat emergence as
quantified as measureable height growth by the sonic distance
sensor (Fig. 2; 37 g C m!2 for April and May for the spring wheat
field and 43 g C m!2 for April and May for the fallow field). Mean Ta
(8.2 #C) and soil temperature (9.0 #C in the spring wheat field and
9.1 #C in the fallow field) were characteristically low during this
period, and carbon fluxes during the largely fallow period in both
fields behaved in a similar manner. Following crop emergence, the
spring wheat field turned into a net carbon sink (since April 1,
2014) in mid-June. Outside of brief periods of weedy growth, the
consistent C source from the fallow field resulted in net April-
September C!!CO2 losses (135 " 73 g C m!2 study period!1) that
were of similar magnitude to C sinks in the spring wheat field
(!107 " 29 g C m!2 study period!1).
Both wheat fields were a net sink of C!!CO2 during the study
period on the order of 200 g C m!2 for the winter wheat field and
100 g C m!2 for the spring wheat field. It has been suggested that
non-woody C3-dominated ecosystems should have similar net
ecosystem productivity due to metabolic and organism size
constraints (Peichl et al., 2013), bringing into question why the
winter wheat field had two-fold greater C!!CO2 uptake than the
spring wheat field, especially when canopy height growth was
greater in the spring wheat field than the winter wheat field
(Fig. 3). Potential explanations include 1) differences in climato-
logical variables between growing seasons; 2) net C!!CO2 efflux in
the spring wheat field during April and May after planting (i.e. the
difference among fields during the April-September study period is
a result of planting date); or 3) respiratory losses from organic
matter left from winter wheat field in the spring wheat field noting
that chemical fallow preceded winter wheat. We briefly discuss
each sequentially.
To test the effects of the cooler and wetter conditions on
determining 2014 fluxes, we parameterized the daily Fc model
using 2014 flux and micrometeorological observations and forced
it with 2013 SWin and Ta observations. This analysis follows the
approach of Richardson et al. (2007) to interpret the role of climate
versus biological and management responses to climate variability
of surface-atmosphere fluxes. The cumulative sum of Fc during the
study period using the 2014 models forced by 2013 meteorology
was !120 g C!!CO2m!2, a difference of only a ca. 10 g C!!CO2m!2
for the study period, suggesting that differences in Fc are not
primarily due to differences in growing season climate.
The second explanation – differences in crop emergence timing
– contributes to the differences in CO2 uptake between the winter
wheat and spring wheat crops. The spring wheat field lost ca.
40 g C m!2 during April and May before the spring wheat field
became a net daily sink for C!!CO2 around May 29, 2014. Due to
earlier planting and thereby crop emergence, the winter wheat
field became a net C!!CO2 sink at the beginning of May. Differences
in seasonal Fc even among winter wheat treatments has been
attributed to differences in early season crop development
(Dufranne et al., 2011).
The third explanation – that there is a greater respiratory source
in the spring wheat field – can be confirmed by analyzing the
parameters of the gapfilling model (Eq. (1)). The mean R10
parameter during the period of peak growth in the winter wheat
field is ca. 2.5 mmol CO2 m!2 s!1, but over 3 mmol CO2 m!2 s!1 in
the spring wheat field. Using mean R10 and E0 parameters in the
respiratory component of Eq. (1) with observed Ta – noting that
Eq. (1) is designed to gapfill Fc rather than partition it into its
components (Reichstein et al., 2012; Wehr et al., 2016) – results in a
modeled RE of 263 g C m!2 during the study period in the winter
April May June July Aug Sept
evitalu
mu
C
H
(M
J 
m
-2
)
0
100
200
300
400
500
600
700
800
900
A
Winter wheat (2013)
Spring wheat (2014)
Fallow (2014)
Winter wheat harvest
Spring wheat harvest
April May June July Aug Sept
H
(M
J 
m
-2
m
on
th
-1
)
0
20
40
60
80
100
120
140
160
180
200
BWinter wheat (2013)
Spring wheat (2014)
Fallow (2014)
Fig. 6. The cumulative (A) and monthly (B) sums of sensible heat flux (H) from the
winter wheat, spring wheat, and fallow study fields during the the April–September
study periods. Error bars represent one standard deviation from the cumulative or
monthly sum.
wheat field and 488 g C m!2 during the study period in the spring
wheat field. This difference may seem large, but using these
modeled RE values to model GEP as the difference between Fc and
RE results in !466 g C m!2 during the study period in the winter
wheat field and !595 g C m!2 in the spring wheat field, a difference
of about 25% that is proportional to the difference in maximum
canopy height (Fig. 3). Greater GEP in the spring wheat field results
in more labile C available for RE (Högberg et al., 2001; Ryan and
Law, 2005; Stoy et al., 2007), but differences in modeled RE are
greater than differences in GEP among fields, pointing to a
respiratory source in the spring wheat field that is not present in
the winter wheat field. These results are consistent with the
findings of Schmidt et al. (2012) and Waldo et al. (2016), who
demonstrated the importance of plant residues from previous
crops to the crop carbon balance. The plant residues that were left
after the 2013 winter wheat harvest (e.g. Fig. 3) that remained in
the field during the 2014 spring wheat crop – noting that chemical
fallow preceded winter wheat – cannot be excluded as a cause of
the differences in C!!CO2 uptake between our winter wheat and
spring wheat observations. These findings are in agreement with
the findings of Peichl et al. (2013) in that differences in carbon
metabolism in these non-woody C3 systems was dominated by the
impacts of land management.
4.2. Evapotranspiration and water management
Daily ET sums quantified here are similar to or greater than
other studies in wheat cropping systems. Maximum daily ET was
ca. 7 mm day!1 for the winter wheat crop (in June) and the spring
wheat crop (in July), which corresponds to low surface resistance
to water by wheat (Bonan, 2008) during periods with sufficient soil
moisture and the high incident radiation load of the study sites.
That being said, the maximum value of aPT calculated by the
gapfilling models was 0.75 for winter wheat and 0.85 for spring
wheat, far less than the theoretical maximum of 1.26 described by
Priestley and Taylor (1972) and values found for irrigated wheat
crops (e.g. 1.17–1.26; Zhang et al., 2004), suggesting limitations to
the transport of water between soil and atmosphere that may be
due to either stomatal closure or limited evaporation due to dry
soil surface conditions. Future work should seek to identify the
source of this water limitation, perhaps using advanced algorithms
for estimating the contribution of evaporation and transpiration to
ET using high frequency eddy covariance data (Scanlon and Kustas,
2010; Scanlon and Sahu, 2008; Sulman et al., 2016).
There were notable differences amongst ET observed here and
in other wheat cropping systems. Anthoni et al. (2004a,b) found
maximum rates of ET on the order 3–4 mm day!1 in mid-May in a
winter wheat crop in Germany, far lower than the 7 mm day!1
maximum observed here. We can estimate SWin from PPFD
observations at the Gebesee study site of Anthoni et al. (2004a,
b) and approximate annual SWin to be on the order of 3500–
4010 MJ m!2 year!1, which is approximately 2/3 of the annual SWin
received by the Judith Basin study sites. In other words, differences
in maximum ET among our observations and those of Anthoni et al.
(2004a,b) are due in part to lower SWin during the growing season
in central Europe.
ET was predictably higher in the winter wheat and spring wheat
fields than the fallow field (Fig. 5A), confirming that fallowing
results in soil water savings if lateral water transport and drainage
losses can be assumed to be minimal during the growth period.
Winter wheat ET was 410 mm during the April–September study
period, which represented ca. 90% of the total precipitation based
on the data from the Moccasin SCAN site in 2013 for the same time
frame, noting that precipitation maps suggest that the study fields
receive slightly more precipitation than the Moccasin SCAN site on
average due to topographical influences. Total ET for the spring
April May June July Aug Sept
)
m( thgie
H
0
500
1000
1500
2000
2500
3000
3500
A
Spring wheat h
Fallow h
h
LCL
Spring wheat harvest
April May June July Aug Sept
yliad 
mu
mixa
M
h
)
m( ecnereffid
-600
-400
-200
0
200
400
600
800
1000
B
Fig. 7. (A) The maximum daily height of the modeled atmospheric boundary layer
(h) calculated using surface virtual heat flux measurements from the spring wheat
and fallow fields and a one dimensional boundary layer model following Luyssaert
et al. (2014) (Eq. (3)) and the height of the lifted condensation level (hLCL) for
maximum daily air temperature modeled by lifting air parcels dry adiabatically to
the saturation point temperature (Eq. (5)) using meteorological observations from
2014 from the spring wheat and fallow study sites. Daily values were smoothed
using a seven-day digital filter. (B) The difference in simulated h among the spring
wheat and fallow fields. Dots represent daily maximum modeled h and lines
represent the output of a seven day digital filter.
Local time
Cl
ou
dy
N
on
e
6:
00
9:
00
12
:0
0
15
:0
0
18
:0
0
21
:0
0
N
um
be
r o
f c
ro
ss
in
gs
0
10
20
30
40
50
60
70
80
# total:
# cloudy:
# none: 
# cross:
t
avg (h):
183
34
74
75
16:08
183
32
70
81
15:37
Fig. 8. Histogram of times during the April–September 2014 study period when
modeled daytime lifted condensation level (hLCL) first exceeds the height of the
atmospheric boundary layer (h). ‘None’ and ‘cloudy’ refers to days when daytime
hLCL never exceeds h and when hLCL at sunrise was below 150 m, respectively.
wheat and fallow fields during the 2014 study period was 440 mm
and 315 mm, which was ca. 90% and 60% respectively of Moccasin
SCAN precipitation observations in 2014. In other words, wheat
cropping returned almost all incident precipitation to the
atmosphere, while the fallow field left over 100 mm of water
available for a subsequent crop, or for deep drainage.
It is likely that the no-till treatment reduced ET in the fallow
field versus a situation in which it was tilled; Chi et al. (2016) for
example demonstrated that tillage tended to increase soil
evaporation. It is important to note that ET may be underestimated
in the winter wheat field and overestimated in the spring wheat
and fallow fields owing to imperfect energy balance closure, but it
is unclear if radiometric or turbulent flux observations are the
cause of the imperfect closure and current practices suggest that
turbulent flux observations should not be adjusted to match
radiometric observations (Baldocchi, 2008).
The amount of water that would need to be replenished to equal
the water savings conferred by fallow during the study period is on
the order of 100 mm. Stubble height in our case is 0.2 m (Fig. 3), and
snow tended to accumulate to stubble height in the study fields. If
we assume a characteristic snow density in MT to be on the order
of 0.08 g cm!3 (Jacobson, 2010; Welch et al., 2016), the depth of
water that corresponds to 0.2 m of snow of this density is 16 mm of
water, an order of magnitude less than the amount of water that
needs to be replenished to account for water use of the wheat
fields. In other words, changes in stubble management are unlikely
to make up the difference in water lost to the atmosphere among
the wheat and fallow fields. If water in the soil column during the
non-growing season is replenished such that soil moisture is at or
near field capacity before crop growth regardless of prior
agricultural treatment, the water savings benefits of fallow should
be questioned.
4.3. Impacts of surface conditions on atmospheric boundary layer &
lifted condensation level
Cumulative H at both wheat fields was ca. 450 MJ m!2 during
the study period, with important seasonal differences due to crop
development and harvest that have implications for energy inputs
into the ABL and its diurnal growth (Fig. 7A). The choice to fallow is
among these management decisions with implications for ABL
dynamics as evidenced by large differences in modeled maximum
daily h between the spring wheat and fallow fields, especially
during the peak crop growth period in June and July in our case
(Fig. 7B).
Gameda et al. (2007) studied the impacts of fallow reduction on
regional climate in the Canadian Prairie Provinces and used a
conceptual model to argue that fallow results in (modeled) h on the
order of 2–3 km but h resulting from vegetated surfaces is on the
order of 0.5 km. Their reasoning is that Rn in fallow is greater due to
lower albedo and is preferentially partitioned into H, which results
in greater virtual heat flux following Eq. (4), while vegetated
surfaces partition Rn preferentially into LE resulting in a cooler and
shallower ABL. Fig. 7A agrees with Gameda et al. (2007) in that the
modeled h is higher as a result of observed surface fluxes from the
fallow field versus the spring wheat field, but our results differ in
magnitude. Our results suggest that the conceptual model of
Gameda et al. (2007) underestimates h in wheat scenarios. It is
important to note that the one-dimensional simulations presented
here simply estimate the influence of a single surface (spring
wheat versus fallow) on a modeled vertical column of the
atmosphere and do not seek to simulate realistic atmospheric
flows.
The intersection of the convective ABL height with the LCL is a
necessary, albeit not sufficient, condition for the initiation of local
convection (e.g. Juang et al., 2007a; Juang et al., 2007b). Maximum
daily hLCL differed on average by ca. 20 m between the two plots,
which can be explained by their close proximity. Low relative
humitidies and hot air temperatures during the month preceding
spring wheat harvest led to modeled hLCL greater than 2500 m,
which exceeded maximum h and suggests that convective
precipitation during this period is unlikely. There was a tendency
to reach hLCL sooner and more often in the fallow simulation due to
higher H and thus h above fallow, (Fig. 8). These findings, however,
need to be interpreted with caution for two reasons. (i) Differences
in surface meteorology between the experimental plots are small,
and (ii) h as well as hLCL are calculated from local quantities, which
do not account for feedbacks between the regional surface energy
balance and the thermodynamic state of the atmosphere as well as
non-local effects such as the development of mesoscale circu-
lations between patches with different surface characteristics (e.g.
Taylor et al., 2007).
Overall, large-scale reduction of fallow has the potential to
decrease h by several hundred meters. Similarly, Betts et al. (2013b)
found that there was a systematic reduction in hLCL in the Canadian
Prairies of approximately 200 m from 1953 to 2011, which is similar
to the calculated mean change in h of 210 m, highlighting the
uncertain nature of the net effect on convective development. h
and hLCL have more crossings in the fallow simulation yet fallow
reduction is thought to result in an increase in cloud development
and convective initiation. Our results suggest that studies of the
effects of fallow on near-surface temperature and humidity
following Betts et al. (2013b) should be extended to the regions
of fallow reduction in the U.S. to ascertain how land use changes
across the entire northern North American Great Plains has
impacted regional climate.
The details of the differences – and similarities – among
modeled h and hLCL as they impact cloud formation processes and
convective precipitation over wheat growing areas has yet to be
ascertained. Given the reduction in the practice of fallow over the
past four decades on the order of tens of millions of hectares
(Fig. 1), the role of cropping systems and planting decisions needs
to be further investigated to quantify the proposed role of
agricultural management in the observed summertime cooling
trend and increase in precipitation across parts of the northern
North American Great Plains (Betts et al., 2013a, 2013b; Mahmood
et al., 2014). Future research should study the impacts of
management patterns in space and time on h and hLCL dynamics
– and measurements of h should be made or estimated from
radiosonde data (e.g. Seidel et al., 2010) – to understand the
consequences of land management across the northern North
American Great Plains on atmospheric dynamics of importance to
convective precipitation.
5. Conclusions
Greater CO2 uptake by the winter wheat field (ca. !200 g C m!2)
than the spring wheat (ca. !100 g C m!2) field during the April–
September study period were largely due to the effects of
ecosystem management: Carbon losses in April and May from
the spring wheat field combined with a larger respiratory source
during the main growth period – consistent with the respiration of
residues from the previous winter wheat crop – explain most of the
difference in Fc among the wheat fields. CO2 losses from the wheat
fields before plant emergence and after harvest occur at a similar
rate as CO2 losses from the fallow field, which represented a total
C!!CO2 loss to the atmosphere of over 100 g C m!2 during the
April–September study period. Total April–September ET was
similar between the winter wheat and spring wheat fields despite
differences in prevailing climatic conditions and crop emergence
and harvest dates, suggesting that ecosystem water use efficiency
was greater in the winter wheat field with greater Fc. Cropping
systems management also had important implications for H and
thereby ABL dynamics, the maximum daily value of which was ca.
200 m less in spring wheat versus fallow simulations, which is on
the order of the reduction of hLCL demonstrated by Betts et al.
(2013b) from the period 1951–1991 to 1992–2011. Future research
should quantify the amount of ET resulting from evaporation and
transpiration to understand sources of water to the atmosphere
and its interaction with plant water use and growth, and
investigate the implications of widespread changes in agricultural
management in the northern North American Great Plains –
including alternate cropping sequences – on regional climatology.
Acknowledgements
This work was supported by the Montana Wheat and Barley
Committee, the Montana University System Water Center, the
USDA National Institute of Food and Agriculture Hatch project
228396, and NSF DEB 1552976 ‘The role of ecosystem management
on boundary layer development and precipitation in the Northern
Plains’. We would like to thank Adam Sigler, Simon Fordyce, Aiden
Johnson, Robby Robertson, and Dr. Lin Hua for field assistance and
coordination and David Walker for advice on agricultural statistics.
We would also like to thank the landowner Bob Otten and the farm
operator Brandon Morris for permission, coordination, and
invaluable insight. Stephanie Ewing, Perry Miller, and Rob Payn
provided valuable comments on the manuscript and Clain Jones
and Adam Sigler provided valuable information on soil character-
istics.
References
Anthoni, P.M., Freibauer, A., Kolle, O., Schulze, E.-D., 2004a. Winter wheat carbon
exchange in Thuringia, Germany. Agric. For. Meteorol. 121, 55–67.
Anthoni, P.M., Knohl, A., Rebmann, C., Freibauer, A., Mund, M., Ziegler, W., Kolle, O.,
Schulze, E.-D., 2004b. Forest and agricultural land-use-dependent CO2
exchange in Thuringia, Germany. Glob. Change Biol. 10, 2005–2019.
Aubinet, M., Grelle, A., Ibrom, A., Rannik, U., Moncrieff, J., Foken, T., Kowalski, A.S.,
Martin, P.H., Berbigier, P., Bernhofer, C., Clement, R., Elbers, J., Granier, A.,
Grunwald, T., Morgenstern, K., Pilegaard, K., Rebmann, C., Snijders, W., Valentini,
R., Vesala, T., 2000. Estimates of the annual net carbon and water exchange of
forests: the EUROFLUX methodology. Adv. Ecol. Res. 30, 113–175.
Béziat, P., Ceschia, E., Dedieu, G., 2009. Carbon balance of a three crop succession
over two cropland sites in South West France. Agric. For. Meteorol. 149, 1628–
1645.
Baldocchi, D.D., 2008. Breathing of the terrestrial biosphere: lessons learned from a
global network of carbon dioxide flux measurements systems. Turner Rev. Aust.
J. Bot. 56, 1–26.
Betts, A.K., Desjardins, R., Worth, D., 2013a. Cloud radiative forcing of the diurnal
cycle climate of the Canadian Prairies. J. Geophys. Res. Atmos. 118, 8935–8953.
Betts, A.K., Desjardins, R., Worth, D., Cerkowniak, D., 2013b. Impact of land use
change on the diurnal cycle climate of the Canadian Prairies. J. Geophys. Res.
Atmos. 118, 11–996.
Billesbach, D.P., Berry, J.A., Seibt, U., Maseyk, K., Torn, M.S., Fischer, M.L., Abu-Naser,
M., Campbell, J.E., 2014. Growing season eddy covariance measurements of
carbonyl sulfide and CO2 fluxes: COS and CO2 relationships in Southern Great
Plains winter wheat. Agric. For. Meteorol. 184, 48–55.
Bonan, G.B., 2008. Forests and climate change: forcings feedbacks, and the climate
benefits of forests. Science (80-.) 320, 1444–1449.
Burgess, M.H., Miller, P.R., Jones, C.A., 2012. Pulse crops improve energy intensity
and productivity of cereal production in Montana, USA. J. Sustain. Agric. 36,
699–718.
Buyanovsky, G.A., Wagner, G.H., 1998. Changing role of cultivated land in the global
carbon cycle. Biol. Fertil. Soils 27, 242–245.
Chi, J., Waldo, S., Pressley, S., O’Keeffe, P., Huggins, D., Stöckle, C., Pan, W.L., Brooks, E.,
Lamb, B., 2016. Assessing carbon and water dynamics of no-till and
conventional tillage cropping systems in the inland Pacific Northwest US using
the eddy covariance method. Agric. For. Meteorol. 218–219, 37–49.
Cihacek, L.J., Ulmer, M.G., 1995. Estimated soil organic carbon losses from long-term
crop-fallow in the Northern Great Plains of the USA. Soil Manag. Greenh. Effect
85–92.
Dhuyvetter, K.C., Thompson, C.R., Norwood, C.A., Halvorson, A.D., 1996. Economics
of dryland cropping systems in the Great Plains: a review. J. Prod. Agric. 9, 216–
222.
Dufranne, D., Moureaux, C., Vancutsem, F., Bodson, B., Aubinet, M., 2011.
Comparison of carbon fluxes, growth and productivity of a winter wheat crop in
three contrasting growing seasons. Agric. Ecosyst. Environ. 141, 133–142.
Yearbook, FAO Statistical. “World food and agriculture.” Food and Agriculture
Organization of the United Nations, Rome (2013).
Falge, E., Baldocchi, D., Olson, R., Anthoni, P., Aubinet, M., Bernhofer, C., Burba, G.,
Ceulemans, R., Clement, R., Dolman, H., Granier, A., Gross, P., Grunwald, T.,
Hollinger, D., Jensen, N.O., Katul, G., Keronen, P., Kowalski, A., Lai, C.T., Law, B.E.,
Meyers, T., Moncrieff, J., Moors, E., Munger, J.W., Pilegaard, K., Rannik, U.,
Rebmann, C., Suyker, A., Tenhunen, J., Tu, K., Verma, S., Vesala, T., Wilson, K.,
Wofsy, S., 2001a. Gap filling strategies for defensible annual sums of net
ecosystem exchange. Agric. For. Meteorol. 107, 43–69.
Falge, E., Baldocchi, D., Olson, R., Anthoni, P., Aubinet, M., Bernhofer, C., Burba, G.,
Ceulemans, R., Clement, R., Dolman, H., Granier, A., Gross, P., Grunwald, T.,
Hollinger, D., Jensen, N.O., Katul, G., Keronen, P., Kowalski, A., Lai, C.T., Law, B.E.,
Meyers, T., Moncrieff, J., Moors, E., Munger, J.W., Pilegaard, K., Rannik, U.,
Rebmann, C., Suyker, A., Tenhunen, J., Tu, K., Verma, S., Vesala, T., Wilson, K.,
Wofsy, S., 2001b. Gap filling strategies for long term energy flux data sets. Agric.
For. Meteorol. 107, 71–77.
Franssen, H.J.H., Stöckli, R., Lehner, I., Rotenberg, E., Seneviratne, S.I., 2010. Energy
balance closure of eddy-covariance data: a multisite analysis for European
FLUXNET stations. Agric. For. Meteorol. 150, 1553–1567. doi:http://dx.doi.org/
10.1016/j.agrformet.2010.08.005.
Gameda, S., Qian, B., Campbell, C.A., Desjardins, R.L., 2007. Climatic trends
associated with summerfallow in the Canadian Prairies. Agric. For. Meteorol.
142, 170–185.
Goulden, M.L., Daube, B.C., Fan, S.-M., Sutton, D.J., Bazzaz, A., Munger, J.W., Wofsy, S.
C., 1997. Physiological responses of a black spruce forest to weather. J. Geophys.
Res. 102, 28987–28996.
Gu, L.H., Falge, E.M., Boden, T., Baldocchi, D.D., Black, T.A., Saleska, S.R., Suni, T.,
Verma, S.B., Vesala, T., Wofsy, S.C., Xu, L.K., 2005. Objective threshold
determination for nighttime eddy flux filtering. Agric. For. Meteorol. 128, 179–
197.
Högberg, P., Nordgren, A., Buchmann, N., Taylor, A.F.S., Ekblad, A., Högberg, M.N.,
Nyberg, G., Ottoson-Löfvenius, M., Read, D.J., 2001. Large-scale forest girdling
shows that current photosynthesis drives soil respiration. Nature 411, 789–792.
Hawkesford, M.J., Araus, J., Park, R., Calderini, D., Miralles, D., Shen, T., Zhang, J.,
Parry, M.A.J., 2013. Prospects of doubling global wheat yields. Food Energy
Secur. 2, 34–48.
IPCC, 2007. Climate Change 2007: the Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA.
Ingwersen, J., Steffens, K., Högy, P., Warrach-Sagi, K., Zhunusbayeva, D., Poltoradnev,
M., Gäbler, R., Wizemann, H.-D., Fangmeier, A., Wulfmeyer, V., Streck, T., 2011.
Comparison of Noah simulations with eddy covariance and soil water
measurements at a winter wheat stand. Agric. For. Meteorol. 151, 345–355.
Jacobson, M.D., 2010. Inferring snow water equivalent for a snow-covered ground
reflector using GPS multipath signals. Remote Sens. 2, 2426. doi:http://dx.doi.
org/10.3390/rs2102426.
Juang, J.-Y., Katul, G.G., Porporato, A., Stoy, P.C., Siqueira, M.S., Detto, M., Kim, H.-S.,
Oren, R., 2007a. Eco-hydrological controls on summertime convective rainfall
triggers. Glob. Change Biol. 887–896.
Juang, J.-Y., Porporato, A., Stoy, P.C., Siqueira, M.B.S., Oishi, A.C., Detto, M., Kim, H.-S.,
Katul, G.G., 2007b. Hydrologic and atmospheric controls on initiation of
convective precipitation events. Water Resour. Res. 43.
Kaimal, J.C., Finnegan, J.J., 1994. Atmospheric Boundary Layer Flows: Their Structure
and Measurements. Oxford, New York.
Lloyd, J., Taylor, J.A., 1994. On the temperature dependence of soil respiration. Funct.
Ecol. 8, 315–323.
Long, J.A., Lawrence, R.L., Greenwood, M.C., Marshall, L., Miller, P.R., 2013. Object-
oriented crop classification using multitemporal ETM+ SLC-off imagery and
random forest. GISci. Remote Sens. 50, 418–436.
Long, J.A., Lawrence, R.L., Miller, P.R., Marshall, L.A., Greenwood, M.C., 2014.
Adoption of cropping sequences in northeast Montana: a spatio-temporal
analysis. Agric. Ecosyst. Environ. 197, 77–87.
Lubowski, R.N., Vesterby, M., Bucholtz, S., Baez, A., Roberts, M.J., 2006. Major Uses of
Land in the United States, 2002. Department of Agriculture, Economic Research
Service, United States.
Luyssaert, S., Jammet, M., Stoy, P.C., Estel, S., Pongratz, J., Ceschia, E., Churkina, G.,
Don, A., Erb, K., Ferlicoq, M., Gielen, B., Grunwald, T., Houghton, R.A., Klumpp, K.,
Knohl, A., Kolb, T., Kuemmerle, T., Laurila, T., Lohila, A., Loustau, D., McGrath, M.
J., Meyfroidt, P., Moors, E.J., Naudts, K., Novick, K., Otto, J., Pilegaard, K., Pio, C.A.,
Rambal, S., Rebmann, C., Ryder, J., Suyker, A.E., Varlagin, A., Wattenbach, M.,
Dolman, A.J., 2014. Land management and land-cover change have impacts of
similar magnitude on surface temperature. Nat. Clim. Change 4, 389–393.
Mahmood, R., Pielke, R.A., Hubbard, K.G., Niyogi, D., Dirmeyer, P.A., McAlpine, C.,
Carleton, A.M., Hale, R., Gameda, S., Beltrán-Przekurat, A., 2014. Land cover
changes and their biogeophysical effects on climate. Int. J. Climatol. 34, 929–
953.
Mauder, M., Foken, T., 2004. Documentation and Instruction Manual of the Eddy
Covariance Software Package TK2. Univ., Abt. Mikrometeorologie.
Merrill, S.D., Black, A.L., Fryrear, D.W., Saleh, A., Zobeck, T.M., Halvorson, A.D.,
Tanaka, D.L., 1999. Soil wind erosion hazard of spring wheat?fallow as affected
by long-term climate and tillage. Soil Sci. Soc. Am. J. 63, 1768–1777.
Miller, P.R., McConkey, B.G., Clayton, G.W., Brandt, S.A., Staricka, J.A., Johnston, A.M.,
Lafond, G.P., Schatz, B.G., Baltensperger, D.D., Neill, K.E., 2002. Pulse crop
adaptation in the northern Great Plains. Agron. J. 94, 261–272.
Miller, P.R., Gan, Y., McConkey, B.G., McDonald, C.L., 2003. Pulse crops for the
northern Great Plains. Agron. J. 95, 980–986.
Moncrieff, J.B., Malhi, Y., Leuning, R., 1996. The propagation of errors in long-term
measurements of land-atmosphere fluxes of carbon and water. Glob. Change
Biol. 2, 231–240.
Motulsky, H.J., Ransnas, L.A., 1987. Fitting curves to data using nonlinear regression:
a practical and nonmathematical review. FASEB J. 1, 365–374.
Moureaux, C., Debacq, A., Hoyaux, J., Suleau, M., Tourneur, D., Vancutsem, F., Bodson,
B., Aubinet, M., 2008. Carbon balance assessment of a Belgian winter wheat crop
(Triticum aestivum L.). Glob. Change Biol. 14, 1353–1366.
Oren, R., Hsieh, C.I., Stoy, P.C., Albertson, J.D., McCarthy, H.R., Harrell, P., Katul, G.G.,
2006. Estimating the uncertainty in annual net ecosystem carbon exchange:
spatial variation in turbulent fluxes and sampling errors in eddy-covariance
measurements. Glob. Change Biol. 12, 883–896.
Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W.,
Longdoz, B., Rambal, S., Valentini, R., Vesala, T., Yakir, D., 2006. Towards a
standardized processing of Net Ecosystem Exchange measured with eddy
covariance technique: algorithms and uncertainty estimation. Biogeosciences 3,
571–583.
Papale, D., 2012. Data gap filling. Eddy Covariance. Springer, pp. 159–172.
Peichl, M., Sonnentag, O., Wohlfahrt, G., Flanagan, L.B., Baldocchi, D.D., Kiely, G.,
Galvagno, M., Gianelle, D., Marcolla, B., Pio, C., Migliavacca, M., Jones, M.B.,
Saunders, M., 2013. Convergence of potential net ecosystem production among
contrasting C3 grasslands. Ecol. Lett. 16, 502–512.
Priestley, C.H.B., Taylor, R.J., 1972. On the assessment of surface heat flux and
evaporation using large-scale parameters. Mon. Weather Rev. 100, 81–92.
Reichstein, M., Falge, E., Baldocchi, D., Papale, D., Aubinet, M., Berbigier, P.,
Bernhofer, C., Buchmann, N., Gilmanov, T.G., Granier, A., Grünwald, T.,
Havránková, K., Ilvesniemi, H., Janous, D., Knohl, A., Laurila, T., Lohila, A.,
Loustau, D., Matteucci, G., Meyers, T., Miglietta, F., Ourcival, J.-M., Pumpanen, J.,
Rambal, S., Rotenberg, E., Sanz, M., Tenhunen, J., Seufert, G., Vaccari, F., Vesala, T.,
Yakier, D., Valentini, R., 2005. On the separation of net ecosystem exchange into
assimilation and ecosystem respiration: review and improved algorithm. Glob.
Change Biol. 11, 1424–1439.
Reichstein, M., Stoy, P.C., Desai, A.R., Lasslop, G., Richardson, A.D., 2012. Partitioning
of net fluxes. Eddy Covariance. Springer, pp. 263–289.
Richardson, A.D., Hollinger, D.Y., Burba, G.G., Davis, K.J., Flanagan, L.B., Katul, G.G.,
Munger, J.W., Ricciuto, D.M., Stoy, P.C., Suyker, A.E., Verma, S.B., Wofsy, S.C.,
2006. A multi-site analysis of random error in tower-based measurements of
carbon and energy fluxes. Agric. For. Meteorol. 136, 1–18.
Richardson, A.D., Hollinger, D.Y., Aber, J.D., Ollinger, S.V., Braswell, B.H., 2007.
Environmental variation is directly responsible for short- but not long-term
variation in forest-atmosphere carbon exchange. Glob. Change Biol. 13, 788–
803. doi:http://dx.doi.org/10.1111/j.1365-2486.2007.01330.x.
Richardson, A.D., Mahecha, M.D., Falge, E., Kattge, J., Moffat, A.M., Papale, D.,
Reichstein, M., Stauch, V.J., Braswell, B.H., Churkina, G., Kruijt, B., Hollinger, D.Y.,
2008. Statistical properties of random CO2 flux measurement uncertainty
inferred from model residuals. Agric. For. Meteorol. 148, 38–50.
Ryan, M.G., Law, B.E., 2005. Interpreting, measuring, and modeling soil respiration.
Biogeochemistry 73, 3–27.
Scanlon, T.M., Kustas, W.P., 2010. Partitioning carbon dioxide and water vapor fluxes
using correlation analysis. Agric. For. Meteorol. 150, 89–99.
Scanlon, T.M., Sahu, P., 2008. On the correlation structure of water vapor and carbon
dioxide in the atmospheric surface layer: a basis for flux partitioning. Water
Resour. Res. 44, W10418.
Schmidt, M., Reichenau, T.G., Fiener, P., Schneider, K., 2012. The carbon budget of a
winter wheat field: an eddy covariance analysis of seasonal and inter-annual
variability. Agric. For. Meteorol. 165, 114–126.
Seidel, D.J., Ao, C.O., Li, K., 2010. Estimating climatological planetary boundary layer
heights from radiosonde observations: comparison of methods and uncertainty
analysis. J. Geophys. Res. Atmos. 115, D16113.
Stoy, P.C., Katul, G.G., Siqueira, M.B.S., Juang, J.-Y., McCarthy, H.R., Oishi, A.C.,
Uebelherr, J.M., Kim, H.-S., Oren, R., 2006a. Separating the effects of climate and
vegetation on evapotranspiration along a successional chronosequence in the
southeastern U. S. Glob. Change Biol. 12, 2115–2135.
Stoy, P.C., Katul, G.G., Siqueira, M.B.S., Juang, J.-Y., Novick, K.A., Uebelherr, J.M., Oren,
R., 2006b. An evaluation of models for partitioning eddy covariance-measured
net ecosystem exchange into photosynthesis and respiration. Agric. For.
Meteorol. 141, 2–18.
Stoy, P.C., Palmroth, S., Oishi, A.C., Siqueira, M.B.S., Juang, J.-Y., Novick, K.A., Ward, E.J.,
Katul, G.G., Oren, R., 2007. Are ecosystem carbon inputs and outputs coupled at
short time scales? A case study from adjacent pine and hardwood forests using
impulse-response analysis. Plant Cell Environ. 30, 700–710.
Stoy, P., Mauder, M., Foken, T., Marcolla, B., Boegh, E., Ibrom, A., Arain, M.A., 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, 137–152.
Stull, R.B., 2012. An Introduction to Boundary Layer Meteorology. Springer Science &
Business Media.
Sulman, B.N., Roman, D.T., Scanlon, T.M., Wang, L., Novick, K.A., 2016. Comparing
methods for partitioning a decade of carbon dioxide and water vapor fluxes in a
temperate forest. Agric. For. Meteorol. 226–227, 229–245.
Sus, O., Williams, M., Bernhofer, C., Béziat, P., Buchmann, N., Ceschia, E., Doherty, R.,
Eugster, W., Grünwald, T., Kutsch, W., 2010. A linked carbon cycle and crop
developmental model: description and evaluation against measurements of
carbon fluxes and carbon stocks at several European agricultural sites. Agric.
Ecosyst. Environ. 139, 402–418.
Taylor, C.M., Parker, D.J., Harris, P.P., 2007. An observational case study of mesoscale
atmospheric circulations induced by soil moisture. Geophys. Res. Lett. 34.
Waldo, S., Chi, J., Pressley, S.N., O’Keeffe, P., Pan, W.L., Brooks, E.S., Huggins, D.R.,
Stöckle, C.O., Lamb, B.K., 2016. Assessing carbon dynamics at high and low
rainfall agricultural sites in the inland Pacific Northwest US using the eddy
covariance method. Agric. For. Meteorol. 218–219, 25–36.
Wehr, R., Munger, J.W., McManus, J.B., Nelson, D.D., Zahniser, M.S., Davidson, E.A.,
Wofsy, S.C., Saleska, S.R., 2016. Seasonality of temperate forest photosynthesis
and daytime respiration. Nature 534, 680–683.
Welch, C.M., Stoy, P.C., Rains, F.A., Johnson, A.V., McGlynn, B.L., 2016. The impacts of
mountain pine beetle disturbance on the energy balance of snow during the
melt period. Hydrol. Process. 30, 588–602.
West, T.O., Marland, G., 2002. A synthesis of carbon sequestration, carbon emissions,
and net carbon flux in agriculture: comparing tillage practices in the United
States. Agric. Ecosyst. Environ. 91, 217–232.
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.
Wischmeier, W.H., 1959. A rainfall erosion index for a universal soil-loss equation.
Soil Sci. Soc. Am. J. 23, 246–249.
Zhang, Y., Liu, C., Yu, Q., Shen, Y., Kendy, E., Kondoh, A., Tang, C., Sun, H., 2004. Energy
fluxes and the Priestley–Taylor parameter over winter wheat and maize in the
North China Plain. Hydrol. Process. 18, 2235–2246.