A Field Procedure to Derive Heat, Water Vapor and Carbon Dioxide Exchange Rates from Digital Images of Vegetative Canopies

Abstract

This analysis of canopy function utilized principles of biophysics (radiation capture and flux gradient theory) and biochemistry (radiation-use efficiency) to infer physiological indicators of canopy productivity and water-use efficiency from digital images of vegetative canopies. Digital images used to construct canopy vegetative indices were scaled to represent absorbed photosynthetically active radiation (APAR), which was derived from down-welling irradiance and fabric with known reflective properties. Net CO 2assimilation (A) was calculated from APAR and expected light utilization coefficients. Corresponding digital images of canopy radiometric temperature were utilized to calculate sensible (H) and latent heat (λT) components of the leaf energy balance and solved for effective leaf conductance (gtWV). Molecular scale indicators of water productivity, including intrinsic transpiration efficiency, can be derived from A, λT and leaf temperature. Corn leaf A, λT, temperature and conductance, determined by gas exchange techniques, were compared with corresponding quantities calculated from digital images of vegetative index and thermal irradiance, as outlined above. Assimilation calculated from digital images was 38% greater than that determined by gas exchange. The mean apparent leaf-air thermal gradient calculated from image analysis was 3.7 times greater than that observed by gas exchange - likely propagating bias in λT and g tWV. However, the relation of leaf conductance to leaf-air thermal gradient was consistent for both methods and bias was substantially reduced when assuming a 1 °C bias in thermal imagery. Field results support further development of this image analysis procedure, based on potential for large throughput support of genetic mapping and classical crop breeding programs related to crop water productivity.

Full text

The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily reflect the
official position of the American Society of Agricultural and Biological Engineers (ASABE), and its printing and distribution does not
constitute an endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review process by
ASABE editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should state that it is
from an ASABE meeting paper. EXAMPLE: Author's Last Name, Initials. 2011. Title of Presentation. ASABE Paper No. 11----. St. Joseph,
Mich.: ASABE. For information about securing permission to reprint or reproduce a technical presentation, please contact ASABE at
rutter@asabe.org or 269-932-7004 (2950 Niles Road, St. Joseph, MI 49085-9659 USA).
An ASABE Meeting Presentation
Paper Number: 1111667
A Field Procedure to Derive Heat, Water Vapor and
Carbon Dioxide Exchange Rates from Digital Images of
Vegetative Canopies
Robert M. Aiken, Associate Professor, Ph.D.
Kansas State University. Northwest Research--Extension Center, 105 Experiment Farm Road,
Colby, KS 67701. raiken@ksu.edu
Patrick I. Coyne, Professor Emeritus, Ph.D.
Kansas State University. Agricultural Research Center-Hays, 1232 240th Avenue, Hays, Kansas
67601. Coyne@ksu.edu
Abdrabbo A. Aboukheira, Associate Research Scientist, Ph.D.
Agricultural Water and Energy Management and Irrigation Technology, Columbia Water Center, The
Earth Institute, Columbia University in The City of New York. 500 W 120 St. 1842 S W Mudd, MC
4711 New York, NY 10027. aas2243@columbia.edu
Written for presentation at the
2011 ASABE Annual International Meeting
Sponsored by ASABE
Gault House
Louisville, Kentucky
August 7 – 10, 2011
Abstract. This analysis of canopy function utilized principles of biophysics (radiation capture and flux
gradient theory) and biochemistry (radiation-use efficiency) to infer physiological indicators of canopy
productivity and water-use efficiency from digital images of vegetative canopies. Digital images used
to construct canopy vegetative indices were scaled to represent absorbed photosynthetically active
radiation (APAR), which was derived from down-welling irradiance and fabric with known reflective
properties. Net CO2 assimilation (A) was calculated from APAR and expected light utilization
coefficients. Corresponding digital images of canopy radiometric temperature were utilized to
calculate sensible (H) and latent heat (λT) components of the leaf energy balance and solved for
effective leaf conductance (gtWV). Molecular scale indicators of water productivity, including intrinsic
transpiration efficiency, can be derived from A, λT and leaf temperature. Corn leaf A, λT, temperature
and conductance, determined by gas exchange techniques, were compared with corresponding
quantities calculated from digital images of vegetative index and thermal irradiance, as outlined
above. Assimilation calculated from digital images was 38% greater than that determined by gas
exchange. The mean apparent leaf-air thermal gradient calculated from image analysis was 3.7

The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily reflect the
official position of the American Society of Agricultural and Biological Engineers (ASABE), and its printing and distribution does not
constitute an endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review process by
ASABE editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should state that it is
from an ASABE meeting paper. EXAMPLE: Author's Last Name, Initials. 2011. Title of Presentation. ASABE Paper No. 11----. St. Joseph,
Mich.: ASABE. For information about securing permission to reprint or reproduce a technical presentation, please contact ASABE at
rutter@asabe.org or 269-932-7004 (2950 Niles Road, St. Joseph, MI 49085-9659 USA).
times greater than that observed by gas exchange--likely propagating bias in λT and gtWV. However,
the relation of leaf conductance to leaf-air thermal gradient was consistent for both methods and bias
was substantially reduced when assuming a 1 oC bias in thermal imagery. Field results support
further development of this image analysis procedure, based on potential for large throughput
support of genetic mapping and classical crop breeding programs related to crop water productivity.
Keywords. carbon dioxide, energy balance, photosynthesis, reflectance spectra, remote sensing,
transpiration, water-use efficiency

2
Introduction
Increased crop water productivity can lead to agricultural gains. Water productivity, also known
as water-use efficiency, can be evaluated in terms of economic yield relative to unit water use.
Vegetative canopies mediate the atmospheric exchanges of water, energy and carbon dioxide
(CO2). At the molecular level, water productivity can be evaluated as the ratio of CO2 uptake by
leaf carbon-fixing enzymes to the rate of water vapor diffusion across the leaf boundary layer
(Nobel, 1983; Campbell and Norman, 1998). Canopy reflectance of visible, near-infrared, and
thermal radiation can help quantify these exchange processes, and thereby support crop
improvement and decisions associated with crop, range and ecosystem management.
Fundamental gains in water productivity may derive from germplasm which maintain biomass
productivity while increasing the driving gradient for CO2 uptake (e.g., decreased CO2 partial
pressures in sub-stomatal cavities, Ci, Xin et al., 2009). Biological traits that alter assimilatory
processes and stomatal regulation are likely to involve multiple genes (Passioura, 1977;
Condon et al., 2004; Sinclair et al., 2005; Reynolds and Tuberosa, 2008; Sinclair et al., 2008);
and are, therefore, considered as complex or quantitative traits. Advances in genetic mapping of
complex traits and/or selection for these traits in classical hybrid development programs require
large throughput evaluation systems (Xin et al., 2009). Remote sensing techniques, that yield
information about water use and assimilation, provide tools that can be used support gains in
crop water productivity. Our research objective was to develop remote sensing techniques with
potential for large throughput field evaluations of water and CO2 exchange rates for vegetative
canopies.
Safety Emphasis
Image and weather data acquisition required deployment of battery powered instrumentation on
a towable lift system with boom extended 4m above the crop canopy. Operator safety required
attention to boom construction, cable deployment and operational procedures. Reducing focal
length of digital camera lenses can reduce boom height required for image acquisition.
Theory
CO2 Assimilation and Photosynthetically Active Radiation
Net assimilation (A,) can be calculated by Fick's law of diffusion, assuming 1) one-dimensional
geometry for an amphistomatous leaf, 2) a driving CO2 gradient from the well-mixed canopy air
layer (Ca) across the leaf boundary layer to sub-stomatal cavities of upper and lower leaf
surfaces (Ci) and 3) a semi-empirical transfer coefficient formulated as a conductance (gtCO2),
representing leaf, stomatal geometry, and boundary layer effects (Nobel, 1983, pp.439, 443).
()
ia
CO
tCCgA −•= 2 (1)
where
A= net assimilation of CO2 (µmol CO2 m-2 s-1)
gtCO2 = leaf conductance of CO2, (m s-1)
Ca = CO2 concentration in the well-mixed canopy air layer (µmol m-3)
Ci = CO2 concentration in leaf sub-stomatal cavities (µmol m-3)

3
Alternatively, Krall and Edwards (1991) showed that A can be calculated from utilization of
absorbed photosynthetically active radiation (APAR), considering the quantum yield of
photosystem II (PSII) and the ratio of quantum yield for CO2 assimilation (CO2) to PSII1.
PSII
CO
PSII APARA Φ
Φ
••Φ= 2 (2)
where
A= net assimilation of CO2 (µmol CO2 m-2 s-1)
PSII =quantum yield of photosystem II (µmol electrons m-2 s-1 µmol-1 absorbed photons utilized
by PSII m2 s1)
APAR = absorbed photosynthetically active radiation (µmol photons absorbed by leaf m-2 s-1)
CO2 = quantum yield for CO2 assimilation (µmol CO2 m-2 s-1 µmol-1 photons APAR m2 s1)
Rochette et al (1996) demonstrated the relation of radiation use to water-use efficiency in
maize. Earl and Davis (2003) showed that PSII could be used to quantify maize biomass
productivity, in relation to APAR, under field conditions. Absorbed PAR can be calculated from
down-welling irradiance, reflectance (ρPAR) and transmittance (τPAR).
()
AR
PPARPAR
RAPAR
τρ
−−•↓= 1 (3)
where
APAR = absorbed photosynthetically active radiation (µmol photons absorbed by leaf m-2 s-1)
RPAR↓= down-welling photosynthetically active radiation (µmol photons m-2 s-1)
ρPAR = leaf reflectance for photosynthetically active radiation
τPAR = leaf transmittance for photosynthetically active radiation
Leaf energy balance
Solution of Equation 1 requires information about gtCO2 which can be derived by inverse solution
of an energy balance equation. One formulation of the leaf energy balance partitions leaf net
radiation (Rn) to sensible (H) and latent (λT) heat flux components.
THRn
λ
+= (4)
where
Rn = net radiation for the leaf (J m-2 s-1)
H = sensible heat flux (J m-2 s-1)
λT = latent heat flux associated with transpiration (J m-2 s-1)
Net radiation at the canopy level can be calculated as the sum of net shortwave (Rns, J m-2 s-1)
and longwave (Rnl, J m-2 s-1) radiation. Assuming shortwave radiation in the near-infrared region
is either transmitted or reflected by leaves (Campbell and Norman, 1998), Rns can be
1 Eight electrons per molecule of CO2 are required for assimilation by C4 plants, based on stoichiometry
(Krall and Edwards, 1991). In practice, Krall and Edwards (1991) found 10-12 electrons derived from PSII
were required for each CO2 molecule fixed. Therefore assimilation rate can be calculated from APAR and
quantum yields of PSII and CO2 assimilation.

4
approximated by APAR, using the conversion of 4.6 µmol PAR J-1 (Dye, 2004). Net longwave
radiation can be calculated from the leaf-air temperature gradient2 (LiCor p. 17-3)
()()
alanl TTTR −•+••••−≈ 3
27342
σε
(5)
where
Rnl = net long wave radiation for the leaf (J m-2 s-1)
ε = leaf thermal emissivity
σ = Stefan-Boltzmann constant, (5.67 x 10-8 J m-2 s-1 oK-1)
Ta = air temperature in the well-mixed canopy layer (oC)
Tl = leaf temperature (oC)
Sensible heat flux, for both sides of a leaf blade, can be calculated from a heat transfer
coefficient and the temperature gradient between the leaf and the well-mixed ambient canopy
layer (Nobel, p. 363).
(
)
alc TThH −•= 2 (6)
where
H= sensible heat flux for the leaf (J m-2 s-1)
hc = heat transfer coefficient (J m-2 s-1 oC-1)
Tl = leaf temperature (oC)
Ta = air temperature in the well-mixed canopy layer (oC)
The heat transfer coefficient can be calculated from the ratio of the thermal conductivity of air to
the leaf boundary layer thickness (Nobel, p. 363).
b
air
c
K
h
δ
= (7)
where
hc = heat transfer coefficient (J m-2 s-1 oC-1)
Kair = thermal conductivity of air (J m-1 s-1 oC-1)
δb = boundary layer thickness (m).
A semi-empirical relationship for the depth of laminar flow in the boundary layer for a flat surface
is dependent on a characteristic length and mean wind velocity adjacent to the boundary layer
(Nobel, p. 358).
v
l
b•= 004.0
δ
(8)
where
δb = boundary layer thickness (m)
l = characteristic leaf dimension, perpendicular to air flow, e.g. leaf width (m)
2 Net long wave thermal radiation for upper and lower leaf surfaces can be calculated as the difference of
incoming (2εσ(Ta + 273)4) and out-going (2εσ(Tl + 273)4) components. The out-going component can be
approximated by noting (Tl + 273)4 ≈ (Ta + 273)4+4(Ta+273)3*(Tl-Ta).

5
v= mean wind velocity adjacent to the boundary layer (m s-1)
Latent heat flux, associated with transpiration, can be calculated from leaf conductance of water
vapor transfer, gtWV, and the driving gradient of water vapor partial pressures between leaf sub-
stomatal cavities (es(Tl)) and air in the well-mixed canopy layer (ea, Nobel p. 410).
()()
vals
WV
tHeTegT •−•=
λ
(9)
where
λT = latent heat flux, associated with transpiration (J m-2 s-1)
gtWV = leaf conductance of water vapor (m s-1)
es(Tl) = saturated vapor pressure at leaf temperature (mol m-3)
ea = vapor pressure of air in well-mixed canopy layer (mol m-3)
Hv = latent heat of vaporization for water (J mol-1)
Latent heat of vaporization for water (Hv, J mol-1) converts transpiration flux (mol m-2 s-1) to
energy equivalence (J m-2 s-1). When Rn and H are defined, λT can be calculated from equation
4; then the effective leaf conductance (glwv) can be calculated from equation 9 and knowledge of
the vapor pressure driving gradient. Leaf conductances for water vapor and CO2 can be related
by the ratio of the molecular weights of water and CO2.
Intrinsic transpiration efficiency
Intrinsic transpiration efficiency (nTE, Pa, Tanner and Sinclair, 1983; Condon et al., 2004; Xin et
al., 2009) and related quantities can be obtained by dividing equation 1 by equation 9.
()()
2CO
t
wv
t
als g
g
T
A
eTenTE ••−= (10)
where
nTE = intrinsic transpiration efficiency (Pa)
es(Tl) = saturated vapor pressure at leaf temperature (kPa)
ea = vapor pressure of air in well-mixed canopy layer (kPa)
A= net assimilation of CO2 (µmol CO2 m-2 s-1)
T = leaf transpiration (mmol m-2 s-1)
glwv = leaf conductance for water vapor (m s-1)
gtCO2 = leaf conductance of CO2, (m s-1)
Ca = CO2 concentration in the well-mixed canopy air layer (µmol m-3)
Ci = CO2 concentration in leaf sub-stomatal cavities (µmol m-3)

6
Procedures
Information about leaf nTE (eq. 10) required solutions for A (eq. 1) and λT (eq. 9). The field
procedure under development utilized digital images of vegetative index3 and thermal irradiance
to quantify APAR (eq. 3) and canopy temperature (Tl). A solution to the leaf energy balance for
fully-illuminated horizontal leaf segments was derived from information extracted from these
digital images as described below.
Field measurements were completed on irrigated corn during mid-vegetative growth. The digital
images analyzed are presented in figure 1. Ambient conditions were sampled within the canopy
(temperature, humidity, wind-speed) using an aspirated and shielded temperature-humidity
sensor (Vaisalla HMP45C) and a cup anemometer (Met One 014A) placed at the height of
upper canopy elements; above canopy irradiance was measured with a horizontal pyranometer
(solar, LiCor LI200X) and quantum sensor (PAR, LiCor LI190SB).
Absorbed photosynthetically active radiation and assimilation
Calculating APAR (eq. 3) from information extracted from the vegetative index image (Tetracam
ADC) for ten leaf segments required information about PAR transmittance (τPAR assumed to be
0.1) and PAR reflectance (ρPAR). The latter was derived from an image-specific calibration
relationship calculated as a weighted average of leaf reflectance in the green and red bands of
the visible spectra. The fraction of PAR energy accounted for by green (0.194) and red (0.134)
bands served as weighting factors. This calculation of ρPAR required calibration of image
luminance (pixel values in red and green wavebands) in relation to known standards (fig. 2).
Ten fabric samples (see fig. 1), ranging from white to black, served as calibration standards.
Reflectance of each fabric section was previously determined by a hyperspectral radiometer
(GER 1500) with fiber fore-optic. Expected assimilation of CO2 under light-limiting or enzyme-
limiting conditions was calculated from principles of leaf photo-biochemistry (eq. 2, see also von
Caemmerer and Furbank, 1999).
Leaf energy balance
Apparent radiometric leaf temperature was extracted for ten leaf segments from thermal images
(ICI 7320), adjusted for air temperature, using manufacturer's software; these leaf segments
corresponded to those selected from the vegetative index image (see above). Apparent bias in
surface temperature was evaluated by comparison of fabric temperature (measured by
thermocouples in contact with the lower surface of the fabric) with temperature indicated by
digital thermal imagery. Net long-wave radiation was calculated from the apparent canopy-air
thermal gradient (eq. 5). Net absorbed radiation was calculated as the sum of APAR and Rnl,
assuming near-infrared radiation was either reflected or transmitted by canopy elements
(Campbell and Norman, 1998). Saturated vapor pressure was calculated for canopy
temperature by Tetens' relationship (Tetens, 1930). Vapor pressure was converted to
concentration (molar volume) by the Ideal Gas Law, recognizing that n/V = P/RT. Sensible heat
flux across the leaf boundary layer was calculated from the apparent thermal gradient and
thermal conductivity of the air boundary layer (eq. 6 - 8). Latent heat flux was calculated as a
residual of the leaf energy balance (eq. 4). Apparent canopy conductance was calculated from
apparent latent heat flux and vapor pressure gradient (eq. 9).
3 Vegetative indices, used in remote sensing, distinguish active vegetative canopy from other surfaces,
based on the strong absorption of visible light (particularly red band, 630 - 690 nm) and enhanced
emittance in the near-infrared band (e.g. 750 - 900 nm) of photosynthetic pigments.

7
Independent gas exchange measurements
An independent evaluation of quantities calculated in the image analysis procedure required
knowledge of A, λT and related boundary conditions. Leaf A and λT were determined using a
portable gas exchange system (LiCor 6400). Leaf measurements, conducted in survey mode,
utilized incident light and a 2 L buffer volume for the instrument's airstream intake;
measurements stabilized within 30 s and represented upper- canopy elements.
Results
Images selected for analysis (fig. 1) corresponded with solar irradiance of 928 J m-2 s-1, PAR
irradiance of 1898 µmol quanta m-2 s-1, canopy air temperature and humidity of 28.68 oC and
0.373, respectively, and wind speed at canopy height of 0.73 m s-1. The canopy sample
corresponded to a 0.5 x 1.0 m area enclosed by a metal frame. Fabric samples, used to
calibrate image reflectance and thermal image bias, were attached to lower and left sections of
the frame. Information regarding pixel values (vegetative index image) and radiometric
temperature were extracted using manufacturers' software.
Figure 1. Digital images used to construct vegetative indices (top) and represent radiometric
temperature (bottom) were subjected to biophysical image analysis.
Fabric reflectance in red and green bands was strongly related to corresponding pixel
luminance values (fig. 2). Apparent bias in thermal image increased with fabric absorbance
(data not shown) and was used to adjust radiometric temperature of leaf segments. Weighted
average apparent leaf reflectance (ρPAR) ranged from 0.01 to 0.10.

8
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0 50 100 150 200 250 300
VI image pixel values
Weighted Fabric Reflectance
Red Green
Red Calibration Green Calibration
Figure 2. Calibration information regarding reflectance was extracted from vegetative index (VI)
image pixel values corresponding to fabric samples with known reflective properties.
Reflectance in the visible light spectrum (400 - 700 nm, photosynthetically active radiation, PAR)
was calculated as a weighted average (weighted by red and green band fraction of PAR) of red
and green reflectance.
Mean APAR and corresponding A, calculated from image analysis, exceeded that observed by
gas exchange by 9 and 38%, respectively (table 1, fig. 3).
0
10
20
30
40
50
60
1200 1300 1400 1500 1600 1700 1800
APAR (µmol quanta m
-2
s
-1
)
A (µmol m
-2
s
-1
)
Gas Exchange
Image Analys is
Figure 3. Net assimilation of CO2 (A) by corn, determined by a portable gas exchange system
(LiCor 6400), is shown in relation to absorbed photosynthetically active radiation (APAR). Leaf
measurements, conducted in survey mode with incident light and a 2 L buffer volume to the
airstream, stabilized within 30 s and represent upper-canopy elements. Expected assimilation
was calculated from apparent APAR derived from digital images of vegetative index using
equation 2.

9
Table 1. Comparison of corn canopy energy, CO2 and water vapor exchange components
calculated from biophysical analysis of digital images and as measured by a portable field gas
exchange system.
a Absorbed photosynthetically active radiation
b Net CO2 assimilation
c Temperature gradient between leaf (Tl) and air in well-mixed canopy layer (Ta)
d Transpiration
e Leaf conductance of water vapor from sub-stomatal cavity across the leaf boundary layer
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
01234
T
l
-T
a
(
o
C)
g
t
(m s
-1
)
Gas Exc hange
Image Analysis
Figure 4. Apparent leaf conductance (gtWV) of corn, determined by gas exchange techniques
(LiCor 6400) is shown in relation to the thermal gradient across the leaf boundary layer.
Corresponding leaf conductance, calculated from the residual of an energy balance equation for
a horizontal leaf, was derived from digital images of vegetative index and thermal irradiance.
Mean apparent latent heat of transpiration, calculated from image analysis, corresponded to
68% of APAR (converted to energy equivalence, fig. 5); while latent heat calculated from gas
exchange observations exceeded APAR by 12%.
APARa
(µmol m-2 s-1)
Ab
(µmol m-2 s-1)
Tl - Tac
(oC)
Td
(mmol m-2 s-1)
gtWVe
(m s-1)
Image
Analysis
n=10
Mean
1600
55.0
1.75
5.42
0.0030
St. Err. 20.0 0.16 0.26 0.45 0.00048
Gas
Exchange
n-22
Mean
1464
39.9
0.47
8.12
0.0080
St. Err. 16.6 1.17 0.10 0.20 0.00041

10
100
150
200
250
300
350
400
450
250 300 350 400
APAR (J m
-2
s
-1
)
λT (J m
-2
s
-1
)
Gas Exc hange
Image Analysis
one to one
Figure 5. Transpiration (λT) of corn leaves, determined by gas exchange techniques (LiCor
6400) is shown in relation to absorbed photosynthetically active radiation (APAR). Units are
converted to energy equivalents using latent heat of vaporization (43,800 J/mol) for T and 4.6
umol quanta/J PAR for APAR. Apparent λT, calculated as a residual of an energy balance for
horizontal leaf segments, was derived from digital images of vegetative index and thermal
irradiance.
Intrinsic transpiration efficiency is expected to be linearly related to the ratio of A to gtCO2 (here
gtCO2 is converted to units of umol m-2 s-1, corresponding to that of A, Nobel pp. 401-403). This
relation holds for nTE and A gtCO2 -1 computed from both gas exchange and image analysis (fig.
6). However, results from image analysis ranged from similar to three times that obtained by the
gas exchange method.
0
10
20
30
40
50
60
70
0 200 400 600 800 1000 1200
A g
t-1
nTE (Pa)
Gas Exchange
Image Analy sis
Figure 6. Intrinsic transpiration efficiency (nTE) of corn leaves, determined by gas exchange
techniques (LiCor 6400) is shown in relation to the ratio of A gtCO2 -1. Apparent nTE, calculated
from A (Equation 2), and λT (Equation 9), was derived from digital images of vegetative index
and thermal irradiance.

11
Discussion
Independent measure of leaf exchange of heat, water vapor and CO2 indicate systematic bias in
corresponding calculations developed for analysis of thermal and vegetative index images.
Positive bias in the Tl-Ta thermal gradient corresponded with negative bias in λT (relative to
APAR) and gtWV. These biases are likely associated as λT and gtWV are calculated as a residual
of the leaf energy balance--assuming an additional 1 oC bias in thermal imagery resulted in a
smaller thermal gradient, smaller sensible heat flux, with corresponding increase in the
calculated λT and gtWV, and decreased nTE that were generally similar in magnitude to that
observed from gas exchange (results not shown).
The positive bias in APAR likely results from negative bias in leaf reflectance, non-uniform
radiance distribution among canopy elements, and non-horizontal leaf orientation for selected
leaf segments. These issues are more complex (Campbell and Norman) than the simple one-
dimensional conceptual model employed here. Trade-offs of complexity and predictive accuracy
require careful consideration. The positive bias in A likely results from assumptions regarding
maximum enzyme-limited assimilation capacity (e.g., Rubisco activity), that may not apply under
these experimental conditions.
Evidence of systematic bias, rather than random error, indicates predictive error is subject to
diagnosis, testing and correction. Near-canopy imaging and corresponding gas exchange
measurements are appropriate for this initial phase of procedural development. Simply adjusting
canopy thermal imagery by 1 oC substantially reduced bias in evaluated parameters, relative to
that observed by gas exchange--demonstrating feasibility for this image analysis procedure.
Considering the potential for high throughput applications, additional development is warranted.
Conclusion
Reasonable values for APAR and gtWV can be constructed from biophysical analysis of digital
thermal and vegetative index images. Systematic biases in all parameters were identified and
are subject to diagnosis and correction. Independent field evidence demonstrates feasibility of a
biophysical image analysis procedure and warrants further development for high throughput
applications related to crop water productivity.
Acknowledgements
This work was supported by the Kansas Agricultural Experiment Station and the Ogallala
Aquifer Program, a consortium between USDA-ARS, Kansas State University, Texas A&M
University, Texas Tech University, and West Texas A&M University.
References
Campbell, G. S. and J. M. Norman. 1998. An Introduction to Environmental Biophysics. 2nd ed.
New York. Springer.
Condon, A. G., R. A. Richards, G. J. Rebetzke and G. D. Farquhar. 2004. Breeding for high
water-use efficiency. J. Exp. Bot. 55(407):2447-2460.
Dye, D. G., 2004: Spectral composition and quanta-to-energy ratio of diffuse photosynthetically
active radiation under diverse cloud conditions. J. Geophys. Res., 109, D10203,
doi:10.1029/2003JD004251.
Earl, H. J. and R. F. Davis. 2003. Effect of drought stress on leaf and whole canopy radiation
use efficiency and yield of maize. Agron. J. 95:688-696.

12
Krall, J. P. and G. E. Edwards. 1991. Environmental effects on the relationship between the
quantum yields of carbon assimilation and in vivo PSII electron transport in maize. Aust.
J. Plant Physiol. 18:267-278.
Nobel, P. S. 1983. Biophysical Plant Physiology and Ecology. New York, W. H. Freeman and
Co.
Passioura, J. B. 1977. Grain yield, harvest index and water use of wheat. J. Aust. Inst. Agric.
Sci. 43:117-120.
Reynolds, M. and R. Tuberosa. 2008. Translational research impacting on crop productivity in
drought-prone environments. Curr. Opin. Plant. Biol. 11(2):171-179.
Rochette, P., R. L. Desjardins, E. Pattey and R. Lessard. 1996. Instantaneous measurement of
radiation and water use efficiencies of a maize crop. Agron. J. 88:627-635.
Sinclair, T. R., G. L. Hammer and E. J. van Oosterom. 2005. Potential yield and water-use
efficiency benefits in sorghum from limited transpiration rate. Functional Plant Biology
32(10):945-952.
Sinclair, T. R., M. A. Zwieniecki and N. M. Holbrook. 2008. Low leaf hydraulic conductance
associated with drought tolerance in soybean. Phys. Plant. 132:446-451.
Tanner, C. B. and T. R. Sinclair. 1983. Efficient water use in crop production: Research or re-
search? In Limitations to efficient water use in crop production, 1-27. Madison,
WI.:American Society of Agronomy.
Tetens, O. 1930. Uber einige meteorologische Begriffe. Z. Geophys. 6:297-309.
Using the LI-6400 Portable Photosynthesis System. 1999. 17: Using an Energy Balance.
Lincoln, NE: LI-COR.
von Caemmerer, S. and R. T. Furbank. 1999. Modeling C4 Photosynthesis. In C4 Plant Biology,
173-211. San Diego, CA.: Academic Press.
Xin, Z., R. M. Aiken and J. Burke. 2009. Genetic diversity of transpiration efficiency in sorghum.
Field Crops Research 11(1-2):74-80.