Full text of DOI:10.1080/01431160117383

This article was downloaded by: [McGill University Library]
On: 03 December 2014, At: 06:41
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK
International Journal of
Remote Sensing
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/tres20
Study of crop growth
parameters using Airborne
Imaging Spectrometer data
N. K. Patel , C. Patnaik , S. Dutta , A. M. Shekh &
A. J. Dave
Published online: 25 Nov 2010.
To cite this article: N. K. Patel , C. Patnaik , S. Dutta , A. M. Shekh & A. J. Dave
(2001) Study of crop growth parameters using Airborne Imaging Spectrometer data,
International Journal of Remote Sensing, 22:12, 2401-2411
To link to this article: http://dx.doi.org/10.1080/01431160117383
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all
the information (the “Content”) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of
the authors, and are not the views of or endorsed by Taylor & Francis.
The accuracy of the Content should not be relied upon and should be
independently verified with primary sources of information. Taylor and
Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in
relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,

reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-and-
conditions

int. j. remote sensing, 2001, vol. 22, no. 12, 2401–2411
Study of crop growth parameters using Airborne Imaging
Spectrometer data
N. K. PATEL, C. PATNAIK, S. DUTTA
Agricultural Resources Group, Space Applications Centre (ISRO), Ahmedabad
380 015, India
A. M. SHEKH and A. J. DAVE
Gujarat Agricultural University, Anand 388 110, India
(Received 7 December 1998; in Ž nal form 11 April 2000)
Abstract. High-spectral-resolution data of the Airborne Imaging Spectrometer
(AIS) developed by the Indian Space Research Organisation were analysed to
check their potential for extraction of information concerning crop growth para-
meters. Re ectance spectra of wheat plots at diVerent growth stages were generated
to study the position of red edge. The shift of in ection wavelength towards
longer wavelength was observed with increase in Leaf Area Index (LAI) and
chlorophyll content. For the range of LAI from 0.08 to 3.16 and chlorophyll
2
content from 2 to 39 (gmÕ ), the shift of in ection wavelength was observed in
the range of 713–723nm. A linear relationship was observed between the in ection
wavelength and LAI as well as chlorophyll content.
1. Introduction
The imaging spectrometer provides re ectance spectra through a sequence of
contiguous bands. These data have a capability to detect sharp absorption features
manifested due to composition and certain physical, chemical and biophysical condi-
tion of materials (Rast 1991). Within the optical region of the electromagnetic spectrum,
natural surfaces and plants are characterized through absorption features due to water,
chlorophyll and minerals. The most dominating factor for study of vegetation is
chlorophyll content, which causes spectral distortion. The studies have been concen-
trated on the transition of the re ectance between the red and near-infrared regions
of the spectrum. In this domain the re ectance presents a sharp increase between 670
and 760 nm, which is called ‘red edge’. This red edge presents an in ection point, the
wavelength of which depends on the plant status (Collins 1978, Baret et al. 1987).
Guyot et al. (1992) has shown that the red edge has an in ection point associated
with a particular wavelength speciŽ c to each plant spectra dependent on phenology
and health status. Green leaves with high chlorophyll concentration show this in ection
point wavelength progressively shifting towards longer wavelengths in progression to
increase in Leaf Area Index (LAI). The position of the red edge is largely in uenced
by LAI. It is independent of soil background and solar zenith angle, and the atmosphere
has a minor eVect (Clevers and Buker 1991).
The potential of imaging spectrometer data for various thematic applications is
summarized by Srivastava et al. (1998). In vegetation studies, the red edge of the
International Journal of Remote Sensing
ISSN 0143-1161 print/ISSN 1366-5901 online © 2001 Taylor & Francis Ltd
http://www.tandf.co.uk/journals

2402
N. K. Patel et al.
re ectance spectrum and especially the wavelength position of the red edge have been
investigated in detail. A large number of studies have been carried out with Ž eld
spectrometers (Boochs et al. 1990, Miller et al. 1991, Plummer et al. 1991). Their
results show that the position and shape of the red edge contain information about
the biomass, chlorophyll content and physiological stress of vegetation. Collins (1978)
showed that the wavelength position of the red edge is a good parameter to describe
the growth of vegetation. Clevers and Buker (1991) modelled that the in ection
wavelength is strongly in uenced by the LAI. Results of the study conducted during
Multi-sensor Airborne Campaign (MAC) Europe ‘91 have shown a strong correlation
between the wavelength position of the red edge and the vegetation height as well as
biomass of corn (Bach et al. 1995). In a recent study conducted on wheat crop, Singh
et al. (1997) detected the temporal shift of the red edge using IRS-P3 Modular Opto-
electronic Sensor (MOS)-B data. The expected relation between the growth and
development of wheat and the red edge of the wheat spectra may be analysed by
correlations between the red edge parameters and the vegetation growth parameters.
In order to investigate the potential applications of high-spectral-resolution
remote sensing data for assessment of crop growth condition, a study was conducted
using data acquired during a test  ight of the Airborne Imaging Spectrometer (AIS)
developed by the Indian Space Research Organisation (ISRO). The main objective
of this study was to check whether this type of data is capable of detecting variation
in crop growth parameters such as LAI, chlorophyll content and biomass. An attempt
was made to study the position of the red edge for wheat crop at diVerent growth
conditions and derive a relationship between crop growth parameters and in ection
wavelength computed from red-edge parameters.
2. Study area
A wheat-growing area of Dholka Taluka (administrative block) of the Ahmedabad
district, Gujarat state, was selected for this study. It is an area irrigated by the
Fatehwadi canal. Sowing and harvesting of wheat crop is done in the middle of
October and the middle of March, respectively. An area of approximately 10 2 km
×
was selected for the AIS  ight. The location of the study area is shown in Ž gure 1.
Four runs of  ight in a north-to-south direction were planned. A reconnaissance
survey was conducted before the  ight. Selection of the study area was done using
a false colour composite (FCC) generated by merging IRS 1C LISS III and panchro-
matic data. Wheat plots having diVerent growth conditions were selected for detailed
study of crop growth parameters.
3. Data used
High-spectral-resolution data were acquired on 24 and 26 February 1997 between
1100 and 1130 hrs from 3 km  ight height using the AIS. The broad speciŽ cations
of the AIS (Anon. 1993) are given in table 1. Ground data collection was carried out
synchronous to the AIS  ight. Plots were identiŽ ed using the cadastral map of
Visalpur village. A portable Ground Truth Radiometer (GTR) developed by the
Space Applications Centre of the Indian Space Research Organisation (ISRO) was
used to collect spectral data of selected plots (Ž gure 2). The Ž eld of view of radiometer
was 15 and the central wavelengths of 11 bands were 490, 565, 660, 670, 710, 745,
°
785, 880, 960 and 1025 nm. The bandwidth of each band was 10 nm. Measurement
of re ected radiance was made holding the radiometer vertically approximately 1.5 m
above a plot, and measurement of irradiance was made using barium sulphate coated

Crop growth parameters using AIS
2403
Figure 1. Location of the test site.
reference plate. Percent re ectance in each band was calculated by taking ratio of
radiance to irradiance. For measurements of LAI, chlorophyll content and biomass,
samples of 10 randomly selected plants were removed from each plot. Fresh weight
of these plants was taken and the number of plants per 1m was counted at three
2
places within each plot to Ž nd out the plant population. Leaf area was measured
using leaf area metre LI COR model LI 3000. Computation of LAI and biomass
was done using measurements of sample plants and plant population.
The chlorophyll content was determined by the method of Hiscox and Israelstam
(1979). Fresh material (50 mg) chopped into small pieces was placed into test tubes
containing 10 ml of dimethyl sulfoxide (DMSO), extra pure AR (Sisco Research
Lab.). The tubes containing leaf tissues and DMSO were kept in an oven adjusted

2404
N. K. Patel et al.
Table 1. Salient features of the AIS.
×
660, 2m 2m from altitude of 3 km
14.5, 770m from 3 km
(384)
1. Instantaneous geometric Ž eld of view (mrad)
2. Swath width (degrees)
(no. of pixels)
3. Spectral range (nm)
450–880 nm
10
12
/
4. Encoding bits pixel
5. Formatted bits pixel
/
6. Band to band registration
Inherent
Raw
143
Spectral
143
3
Spatial
17*
3 to 24
7. Number of spectral bands
8. Spectral bandwidth (nm)
9. Spatial resolution
3
Across track (mrad)
Along track (mrad)
10. Field of view (degrees)
0.66
0.66
14.5
5.28
0.66
14.5
0.66
0.66
14.5
*Within the speciŽ ed interval.
Table 2. Crop growth parameters measured for selected Ž elds.
Fresh biomass Dry matter Chlorophyll Plant height
2
(kg mÕ )
2
(kg mÕ )
2
(g mÕ )
Plot no.
Growth stage
LAI
(cm)
W1024
W1023
W1020
W1011
W100
W1008
W1175
W1087
W1165
W1081
W1146
W1539
W1213
W1520
W1525
W1515
W1514
Flowering
Flowering
Maturity
Flowering
Maturity
Milking
2.72
2.14
3.14
1.27
0.68
2.59
1.16
1.23
3.63
1.65
2.79
2.39
1.66
0.47
1.03
1.99
2.26
0.59
0.56
1.45
0.38
0.63
1.00
0.23
0.53
1.45
0.37
0.83
0.69
0.66
0.11
0.54
0.93
0.88
3.16
1.75
0.46
0.71
0.30
1.22
1.90
0.83
1.08
1.51
2.31
0.59
0.35
0.43
0.08
1.17
0.71
39.91
25.77
10.36
13.63
2.23
16.35
15.94
12.55
27.22
18.05
23.51
13.33
16.06
5.66
76.2
58.3
78.6
68.3
56.2
83.9
43.0
68.3
83.9
86.6
90.1
88.5
72.1
51.1
65.2
86.1
91.8
Booting
Soft dough
Soft dough
Flowering
Milking
Milking
Soft dough
Flowering
Maturity
Milking
7.01
22.44
22.01
Soft dough
at 65 C temperature for about 3 h. During the incubation period, a gentle shaking
°
was given to all the tubes. Chlorophyll was extracted into  uid, transferred to
graduated tubes and made up to a total volume of 10 ml by adding extra DMSO.
A 3-ml sample of chlorophyll extract was transferred to a cuvette, and absorbance
was measured as the optical density values (OD) at 663 and 645 nm, using a Backman
Spectrometer. Chlorophyll a, b and total chlorophyll (a b) content (expressed as
+
mg g from leaf weight) was calculated by the following equations:
Õ1
chlorophyll a 12.7 E
=
2.63 E
(1)
(2)
(3)
-
663
645
645
663
chlorophyll b 22.9 E
=
4.68 E
-
total chlorophyll 22.2 E
=
8.02 E
+
645
663
where E and E are the absorbance values (OD) at 663 and 645 nm, respectively.
663 645

Crop growth parameters using AIS
2405
Figure 2. Location of the selected Ž elds in Visalpur village.
4. Data analysis
The major steps involved in the data analysis are shown in Ž gure 3. Data acquired
in raw mode on 24 and 26 February 1997 were used. Data of Visalpur village were
extracted and converted into a sub image of 143 bands. The FCC prepared using
one each in green, red and near-infrared bands was displayed and used for identiŽ ca-
tion of selected plots. Signatures of selected plots were generated in all bands and
mean values of digital counts obtained. Radiometric calibration constants of gain
and oVset for each band provided along with the raw data were used to convert
digital counts into radiance.
Wheat plots at diVerent growth stages and bare soil (fallow plot) were selected
to study spectral characteristics. Figures 4 and 5 show spectral radiance response of

2406
N. K. Patel et al.
Figure 3. Major steps of the data analysis.
Wavelength (nm)
Figure 4. Spectral radiance of soil.

Crop growth parameters using AIS
2407
Wavelength (nm)
Figure 5. Spectral radiance of wheat crop.
soil and wheat, respectively. Wheat shows low radiance in the red region and higher
radiance in the near-infrared region compared to soil. The absorption dip observed
at 760 nm is mainly due to absorption by atmospheric oxygen (Bach et al. 1995 ).
Spectral re ectance of soil (fallow plot) measured by radiometer was assumed to be
same at the altitude of the AIS. Considering soil re ectance as reference, irradiance
was calculated and further used to calculate percent re ectance from selected plots.
Absorption by photosynthetic active biomass of crop canopy in the red region near
670 nm and high re ectance in the near-infrared region above 780 nm is observed
in Ž gure 5.
The position of the red edge is determined by in ection wavelength, which is
deŽ ned as the wavelength where the rate of increase of re ectance is the maximum.
The characteristics of position and shape of the red edge in the visible and near-
infrared are good indicators of plant parameters. Miller et al. (1991) evaluated an
inverted Gaussian model for the vegetation red-edge re ectance. Fitting of inverse
Gaussian function to the spectral data in this region leads to four parameters, which
represents the red-edge characteristics (Bach and Mauser 1991). Spectral re ectance
curves for vegetation exhibit a consistent shape in the red-edge region characterized
by a relatively broad  at minima in the 670 nm region, followed by a sharp increase
in re ectance beginning at about 685 nm and an asymptotic re ectance plateau
reached at wavelengths beyond 780 nm. The inverted Gaussian model, which
represents the red edge by the re ectance, is
(l l)
- 2
-
0
R(l)
R
(R R ) exp
(4)
=
-
-
s
s
0
A
infrared region and s is the Gaussian function deviation parameters. A Ž fth parameter
2s
B
2
where R is the maximum or shoulder spectral re ectance, R is the minimum spectral
re ectance corresponding to the chlorophyll absorption well, l is the central min-
imum or peak absorption wavelength, l is the wavelength in the red and near-
s
0
0
is l , the wavelength of in ection of the red re ectance edge slope, deŽ ned by the
p

2408
N. K. Patel et al.
wavelength of the maximum in the Ž rst derivative of the Gaussian function (Miller
et al. 1990):
l
l
s
+
(5)
=
p
0
The parameter l provides another measure of the position of the vegetation red
p
re ectance edge. The standard numerical procedure is employed to produce a best
Ž t to re ectance data according to least-square criterion. The details of the procedure
are described by Bonham-Carter (1988). For inverted Gaussian model Ž tting,
re ectance data from 650 to 780 nm were used, except data from 750 to 770 nm
wavelengths due to absorption dip by atmospheric oxygen at 760 nm. This model
was used for 17 plots of wheat with diVerent growth conditions.
5. Results and discussion
The red edge is a phenomenon caused by the combination of the chlorophyll
absorption in the red region and the scattering by the plant cells in the near-infrared
region of the electromagnetic spectrum. The wavelength position of the red edge is
known to be parameter which is sensitive to plant development (Guyot et al. 1992 ).
The spectral re ectance curves of three wheat plots at diVerent growth stages viz.
 owering, soft dough and maturity stages are shown in Ž gure 6. The diVerence in
crop growth is re ected in spectral responses around the chlorophyll absorption in
the red and near-infrared regions. Separation between spectral response of wheat
plots at diVerent growth stages increased in the region of high re ectance from 750
to 880 nm. In the near-infrared region, maximum re ectance was recorded from plot
W1024, which was at the  owering stage, and more chlorophyll content and minimum
re ectance was recorded from plot W1020, which was at the maturity stage. This
trend was reversed at the shorter wavelengths, where the wheat plot at  owering
stage exhibited enhanced absorption. Plot W1087 shows the intermediate values.
This data indicate the change of re ectance values with crop growth.
The position of the red edge is determined by in ection wavelength, which is
Wavelength (nm)
Figure 6. Spectral re ectance of wheat at diVerent growth stages.

Crop growth parameters using AIS
2409
deŽ ned as the wavelength where the rate of increase of re ectance is the maximum.
It is determined through Gaussian curve Ž tting to the re ectance data which can
provide an eVective quantitative representation of the shape and position of vegeta-
tion red-edge re ectance in terms of physical signiŽ cance (Miller et al. 1991). The
in ection wavelength of selected wheat plots was calculated using inverted Gaussian
Ž t. Figure 7 shows the graph for three wheat plots at diVerent LAIs. Modelling of
the red edge shows longest in ection wavelength of plot number W1024, with a LAI
of 3.16. The reduction in LAI and chlorophyll in the senescence phase is re ected in
the graph line of plots W1087 and W1020, with LAI of 0.83 and 0.46, respectively.
The shift of in ection wavelength is marked in Ž gure 7. The chlorophyll content of
plot W1024 was higher compared to the other two plots. It indicated that the
position of the in ection wavelength is shifting towards longer wavelength with
increase in LAI.
An attempt was made to derive a relationship between in ection wavelength and
crop growth parameters. The scatter plot of in ection wavelength with LAI and
chlorophyll content is shown in Ž gures 8 and 9, respectively. Regression analysis
was carried out and a linear relationship was observed between in ection wavelength
and LAI and chlorophyll content. Wheat crop was in the post heading stage at the
time of the AIS test  ight in the month of February but, along with green leaves,
dry leaves and other components, also contributed to spectral response, resulting in
low values of correlation coeYcients. This preliminary experiment has indicated the
feasibility to study crop growth parameters using AIS data. Detailed investigation
may be carried out with detailed temporal data incorporating atmospheric correction.
6. Conclusion
A shift of the in ection wavelength of spectra of wheat from 713 to 723 nm is
evident for wheat plots at diVerent growth stages. This shift was determined by
extracting the in ection wavelength of the red edge of the re ectance spectra.
A linear relation of in ection wavelength with LAI and chlorophyll content of
wheat indicates use of high-spectral-resolution data for assessment of crop growth
Wavelength (nm)
Figure 7. Inverted Gaussian Ž t of wheat spectra.

2410
N. K. Patel et al.
In¯ ection wavelength (nm)
Figure 8. Relationship between in ection wavelength of wheat spectra and LAI.
In¯ ection wavelength (nm)
Figure 9. Relationship between in ection wavelength of wheat spectra and chlorophyll
content.
conditions and identiŽ cation of stressed crop. This study has provided an opportunity
for understanding data acquired by an imaging spectrometer and the utility of this
data in the study of crop growth parameters, and it has highlighted the need for
further research into understanding the spectral response of crops grown under
diVerent growth conditions.
Acknowledgments
The authors are grateful to Dr R. R. Navalgund (Deputy Director, Remote
Sensing Applications Area) and Shri J. S. Parihar (Group Director, Agricultural
Resources Group) for encouragement and guidance. They are also thankful to Shri
A. S. Kirankumar and Shri V. H. Patel for providing AIS data. Support provided
by Shri N. P. Pandya for  ight planning and software provided by Shri R. P. Singh
are gratefully acknowledged.

Crop growth parameters using AIS
2411
References
Anon., 1993, Design and development of Airborne Imaging Spectrometer. SAC EOSDG
/
/
/
/
AIS 09 93, Space Applications Centre, Ahmedabad, India.
Bach, H., Demircan, A., and Mauser, W., 1995, The use of AVIRIS data for the determination
of agricultural plant development and water content. Proceedings of MAC Europe
’91—Final Results. ESA WPP-88 (Paris: ESA), Lenggries, Germany.
Bach, H., and Mauser, W., 1991, Extraction of agricultural parameters from imaging spectro-
metry data and simulated MERIS data through red edge analyses. European Imaging
Spectroscopy Aircraft Campaign. ESA SP-360 (Noordwijk: ESA).
Baret, F., Champion, I., Guyot, G., and Podaire, A., 1987, Monitoring wheat canopies with
a high spectral resolution radiometer. Remote Sensing of Environment, 22, 367–378.
Bonham-Carter, G. F., 1988, The numerical procedures and computer program for Ž tting
an inverted Gaussian model to vegetation re ectance data. Geoscience, 14, 339–356.
Boochs, F., Kufer, G., Docker, K., and Kuhbauch, W., 1990, Shape of the red edge as
vitality indicator for plants. International Journal of Remote Sensing, 11, 1741–1753.
Clevers, J. G. P. W., and Buker, C., 1991, Feasibility of the red edge index for the detection
of nitrogen deŽ ciency. Proceedings of the Fifth International Colloquium, Physical
Measurements and Signatures in Remote Sensing, 14–18 January, Courchevel,
pp. 165–168.
Collins, W., 1978, Remote Sensing of crop type and maturity. Photogrammetric Engineering
and Remote Sensing, 44, 43–55.
Guyot, G., Baret F., and Jacuemoud, S., 1992, Imaging spectroscopy for vegetation studies.
In Fundamentals and Prospective Applications, edited by F. Toselli and J. Bodechtel
(Kluwer Academic), pp. 145–165.
Hiscox, J. D., and Israelstam, C. F., 1979, A method for the extraction of chlorophyll from
leaf tissue without maceration, Caadian. Journal of Botany, 57, 1332–1334.
Miller, J. R., Hare, E. W., and Wu, J., 1990, Quantitative characterization of vegetation
red edge re ectance: an inverted-Gaussian re ectance model. International Journal of
Remote Sensing, 2, 1755–1773.
Miller, J. R., Jiyou Wu, Boyer, M. G., Belanger, M., and Hare, E. W., 1991, Seasonal
patterns in leaf re ectance red edge characteristics. International Journal of Remote
Sensing, 12, 1509–1523.
Plummer, S. E., Wilson, A. K., and Jones, A. K., 1991, On the relationship between high
spectral resolution canopy re ectance data and plant biochemistry. EARSeL Advances
in Remote Sensing, 1, 27–33.
Rast, M., 1991, Imaging spectroscopy and its aplication in spaceborne systems. ESA SP-1144,
December 1991.
Singh, R. P., Oza, S. R., and Dadhwal, V. K., 1997, Detection of red edge shift from
temporal growth of wheat crop using high resolution imaging MOS-B spectrometer
data. NNRMS Bulletin, B-21, 49–53.
Srivastava, S. K., Nageswara, Rao, P. P., and Jayaraman, V., 1998, Towards space-borne
terrestrial imaging spectrometry. ISRO-NNRMS-SR-41-98, Indian Space Research
Organisation, Bangalore.