Remote Sensing Letters

ISSN: 2150-704X (Print) 2150-7058 (Online) Journal homepage: http://www.tandfonline.com/loi/trsl20

Estimating leaf area index of maize using airborne

full-waveform lidar data

Sheng Nie, Cheng Wang, Pinliang Dong & Xiaohuan Xi

To cite this article: Sheng Nie, Cheng Wang, Pinliang Dong & Xiaohuan Xi (2016) Estimating
leaf area index of maize using airborne full-waveform lidar data, Remote Sensing Letters, 7:2,
111-120, DOI: 10.1080/2150704X.2015.1111536

To link to this article: http://dx.doi.org/10.1080/2150704X.2015.1111536

Published online: 09 Nov 2015.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=trsl20

Download by: [Wuhan University] Date: 10 November 2015, At: 06:47

 REMOTE SENSING LETTERS, 2016
VOL. 7, NO. 2, 111120
http://dx.doi.org/10.1080/2150704X.2015.1111536

Estimating leaf area index of maize using airborne

full-waveform lidar data
Sheng Niea,b, Cheng Wanga, Pinliang Dongc and Xiaohuan Xia
a
Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, Chinese Academy
of Sciences, Beijing, China; bUniversity of Chinese Academy of Sciences, Beijing, China; cDepartment of
Geography, University of North Texas, Denton, TX, USA

ABSTRACT ARTICLE HISTORY

The leaf area index (LAI) is a key input parameter in ecosystem Received 12 May 2015
Downloaded by [Wuhan University] at 06:47 10 November 2015

models and plays a vital role in gasvegetation exchange pro- Accepted 1 October 2015
cesses. Several studies have recently been conducted to estimate
the LAI of low-stature vegetation using airborne discrete-return
light detection and ranging (lidar) data. However, few studies have
been carried out to estimate the LAI of low-stature vegetation
using airborne full-waveform lidar data. The objective of this
research is to explore the potential of airborne full-waveform
lidar for LAI estimation of maize. First, waveform processing was
conducted for better extraction of waveform-derived metrics for
LAI estimation. A method of faint returns retrieval was also pro-
posed to obtain ground returns. Second, the LAIs of maize were
estimated based on the BeerLambert law. Finally, the LAI esti-
mates were validated using eld-measured LAIs in Huailai, Hebei
Province of China. Results indicated that maize LAI could be
successfully retrieved with high accuracy (R2 = 0.724,
RMSE = 0.449) using full-waveform lidar data by the method
proposed in this study.

1. Introduction
Leaf area index (LAI) is dened as the total one-sided leaf area per unit ground area
(Watson 1947) and is a key biophysical parameter relating to gasvegetation exchange
between vegetation and the atmosphere (Capodici, DUrso, and Maltese 2013; Zhao and
Popescu 2009). It is therefore necessary to accurately estimate vegetation LAI. Traditional
methods of determining the LAI, such as direct eld measurement, can accurately
retrieve vegetation LAI (Riao et al. 2004). However, it is labour-intensive and time-
consuming, making it only practical in small areas.
Remote sensing data, on the other hand, can provide eective solutions to estimate
vegetation LAI at regional or global scale (Capodici, DUrso, and Maltese 2013; Zheng
and Moskal 2009). Among the existing remote sensing techniques, light detection and
ranging (lidar) is an active remote sensing laser-mapping technology capable of
capturing both horizontal and vertical vegetation structure information with high
accuracy (Lefsky et al. 2002; Mallet and Bretar 2009). In addition, it can overcome

CONTACT Cheng Wang wangcheng@radi.ac.cn

2015 Taylor & Francis
 112 S. NIE ET AL.

the limitation of optical sensors for LAI estimations such as the saturation problems
(Armston et al. 2013; Lin and Zhang 2014; Zheng and Moskal 2009). It has therefore
been a very useful tool for accurately estimating vegetation LAI. Many studies have
demonstrated the capability of airborne discrete lidar data to retrieve forest LAI using
the method based on the BeerLambert law (Morsdorf et al. 2006; Solberg 2010).
However, airborne discrete lidar data cannot accurately estimate vegetation LAI in
short, dense vegetation areas because the data do not contain sucient ground
returns due to the poor penetration ability of discrete lidar in high-LAI regions (Cui
et al. 2011; Luo et al. 2014, 2015). In recent years, airborne full-waveform lidar, an
emerging and rapidly advancing lidar sensor, has provided new opportunities for
estimating vegetation parameters. Unlike conventional small-footprint discrete-return
lidar data that record only the most signicant echoes, this new generation of sensors
records the entire backscattered energy, which greatly improves the penetration ability
in high-LAI regions (Mallet and Bretar 2009; Pirotti 2011; Wang et al. 2011). Full-
Downloaded by [Wuhan University] at 06:47 10 November 2015

waveform lidar sensors can therefore become a more appropriate choice for accurately
estimating the LAI of short, dense vegetation.
The objective of this study is to examine the capability of airborne full-waveform lidar
data to estimate LAI of maize. To full this goal, three main steps were carried out: (1)
processing of the full-waveform lidar data, including a method for retrieving faint
ground returns; (2) estimating the LAI of maize using the method based on the Beer
Lambert law; (3) validating the accuracy of LAI estimations by comparing full-waveform
lidar-derived LAI with eld measurements.

2. Materials
2.1. Study area
The study area is located in Huailai, Hebei Province of China (Figure 1). The terrain in this
experiment site is relatively at, and the average elevation is about 480 m above sea
level. The dominant crop in the study area is maize.

2.2. Airborne lidar data

Small footprint full-waveform airborne lidar data were acquired in July 2014 using a Leica
ALS70-HA system (Leica Geosystems Ltd., Aarau, Switzerland), with a pulse rate of 50 KHz
and a scanning angle of 12 from nadir. The system was operated at 1600 m above ground
level, with the beam divergence xed with 0.22 mrad, leading to a footprint diameter of
approximately 0.35 m. The average pulse density was approximately 8 return echoes /m2.
To generate a digital terrain model (DTM) from lidar data for better calculating
waveform metrics, the discrete point clouds data were extracted from full-waveform
signals using a waveform decomposition approach in this study. This approach is based
on tting sums of Gaussian functions to the lidar waveform data (Wagner et al. 2006).
After retrieving the point clouds, the processing of point clouds was also conducted to
obtain the accurate DTM. The frequency histogram method was rst carried out to
remove outliers. Then lidar point clouds were classied as canopy and ground returns
using the TerraScan software (TerraSolid Ltd., Helsinki, Finland; Axelsson 2000), and a
 REMOTE SENSING LETTERS 113
Downloaded by [Wuhan University] at 06:47 10 November 2015

Figure 1. The location of eld samples shown on the airborne lidar-derived DTM.

DTM was nally created using the ground returns within each grid cell of 0.5 m 0.5 m.
In addition, the height accuracy of DTM was assessed using GPS points as being
0.05 0.12 m. Using this DTM, we calculated the relative height of full-waveform lidar
data above ground to eliminate the eect of ground.

2.3. Field measurement

Forty samples of maize were collected from 7 to 12 July 2014 in test site. The size of each
plot is 6.0 m 6.0 m. The geographic coordinates of each plot were determined using a
real-time kinematic GPS (RTK GPS). In addition, the eld LAIs were measured using a
Plant Canopy Analyser (LAI-2200, Li-COR, Inc., Lincoln, NE, USA) based on gap fraction
measurements. Eective LAI values were calculated using the vendor-provided FV-2000
software (LI-COR 2010).

3. Methods
3.1. Waveform processing
In this study, waveform processing was performed to accurately retrieve waveform
parameters for eectively estimating LAI of maize. The procedure was performed
in three stages: waveform preprocessing, waveform normalization, and waveform
stacking.
 114 S. NIE ET AL.

3.1.1. Waveform preprocessing

To ensure the reliability of the retrieved waveform parameters, airborne full-
waveform data must be rst preprocessed. Waveform preprocessing mainly involves
the retrieval of real waveform signal and waveform smoothing. The full-waveform
lidar system continuously records the entire backscattered energy. The raw waveform
contains useless signal not including any information of the Earths surface. A step
was therefore necessary to extract the actual waveform signal from the raw wave-
form. To retrieve the actual waveform signal, the mean and standard deviation of
background noise were rst estimated from the raw waveform based on the fre-
quency histogram (Sun et al. 2008), and the mean background noise was subtracted
from the raw waveform to remove the background noise (Duong et al. 2008). Then
the waveform start and waveform end were identied according to the threshold,
which was set as three times standard deviation of background noise, and the real
signal waveform was nally extracted as the part of waveform between waveform
Downloaded by [Wuhan University] at 06:47 10 November 2015

start and waveform end (Pirotti et al. 2014). In addition, the actual waveform signal
contains noise due to the limitations of sensor capacity and interactions between the
emitted pulses with the Earths surface. To reduce the noise and obtain the
smoothed waveform, waveform smoothing was also performed using a Gaussian
lter in this study (Mallet and Bretar 2009).

3.1.2. Waveform normalization

The preprocessed waveform was normalized to enable a comparison or accumulation of
waveforms captured within the same plot. Due to dierent emitted pulses and dierent
atmospheric conditions, the returned waveforms may be dierent, even if the Earths
surface is homogeneous, and the preprocessed waveform must therefore be normalized
to remove those eects (Duong et al. 2008). The normalization is a procedure in which
the received energy at moment i, Vi , was divided by the total energy VT , as shown in
Equation (1):

Vi XN
VN i with VT Vi (1)
VT i1

where N is the number of waveform bins. In this research, N equals 128.

3.1.3. Waveform stacking

In previous studies, each small-footprint waveform was usually decomposed into
several points using a custom Gaussian decomposition method (Wagner et al. 2006).
However, this method has a limitation that it cannot detect the faint returns which are
the result of reduced waveform due to obscurant scattering, attenuation, and absorp-
tion. The faint returns are the ground returns in an area of short, dense vegetation,
which is important for estimating LAI of vegetation using the method based on the
BeerLambert law. In order to obtain the faint ground returns, waveform stacking was
conducted in this research. The lidar waveform stacking is an eective technique for
faint ground return detection and enhancement, which has been demonstrated by the
previous studies (Allouis et al. 2013; Magruder, Neuenschwander, and Marmillion 2010).
The waveform stacking was performed in there stages in this paper. First, since LAS
 REMOTE SENSING LETTERS 115

Specication Version 1.3 (LAS 1.3) data records the elevation of the beginning of each
waveform and the vertical height value of each waveform bin, the relative elevation of
each waveform bin was therefore calculated using DTM. In addition, the vertical height
value of each waveform bin is dierent for dierent waveforms, and a new waveform
was then generated from each normalized waveform with 15 cm vertical height for
each waveform bin. Finally, all those new waveforms located within the same plot
were directly summed up to obtain a cumulative waveform. By this way, the random
noise can be reduced and the faint returns enhanced. The ground returns can be
therefore easily obtained from the cumulative waveform through setting the optical
height threshold.

3.2. LAI estimation

Downloaded by [Wuhan University] at 06:47 10 November 2015

The previous studies demonstrated that LAI can be deduced by the BeerLambert law.
By the transformation equation of the BeerLambert law, LAI can be calculated based on
the gap fraction I=I0 once k is known, as shown in Equation (2) (Li et al. 2015; Solberg
2010):
1
LAI  lnI=I0 (2)
k
where I is the below canopy light intensity, I0 is the above canopy light intensity, and k is
the extinction coecient (Richardson, Monika Moskal, and Kim 2009).
Following Luo et al. (2013), the gap fraction could be replaced by the ratio Er of
ground return energy to total return energy of a lidar waveform. But due to the
reectance dierence between the canopy and the ground, the gap fraction cannot
be replaced by Er directly. To reduce the eect of this reectance dierence, we
corrected Er from the cumulative waveform for better estimating LAI of maize according
to Equation (3):
Rg =g nRg
Er (3)
Rc =c Rg =g Rc nRg

where g the average ground reectance, c the vegetation reectance, Rg is the ground
return energy of the cumulative waveform, Rc is the canopy return energy of the
cumulative waveform, and n is the adjusting factor of reectance.
According to Luo et al. (2014), the adjusting factor of reectance was set to 4 for our
study site. Following Zhao and Popescu (2009), we examined a sequence of height
thresholds from 0.05 to 0.4 m incremented by 0.05 m. We found that 0.1 m is the
optimal height threshold to separate ground and canopy in our study. After obtaining
the optimal height threshold, the ground return energy Rg and the canopy return
energy Rc can be calculated by summing up the energy of the ground returns and the
canopy returns, respectively. Once Rc , Rg , c , and g are known, Er can be calculated
according to Equation (3), LAI can be therefore estimated using Equation (4) once k is
known:
1
LAI  ln Er (4)
k
 116 S. NIE ET AL.

To obtain an optimal LAI estimation model, the linear regression analysis of the
logarithm of Er against the eld-measured LAI was carried out. In addition, some
eld-measured LAIs, not used in building the LAI estimation model, were left for the
model validation. In our study, both the coecient of determination (R2) and the
root mean square error (RMSE) were calculated for the assessment of LAI estimation
accuracy.

4. Results and discussion

To estimate the vegetation LAI, the extinction coecient k must be rst obtained
according to the BeerLambert law. In this study, a simple linear regression between
eld-measured LAI and the logarithm of Er was conducted to obtain the best
extinction coecient k. Among all the 40 eld-measured LAIs, 25 samples were
Downloaded by [Wuhan University] at 06:47 10 November 2015

used for the linear regression analysis, while the other 15 samples were used for
model validation. The regression model between the full-waveform lidar-derived
variable and the eld-measured LAI was shown in Equation (5). The scatter plot of
the eld-measured LAI versus  lnEr and the corresponding regression line were
illustrated in Figure 2.
y 0:740x 0:597 (5)

where y is LAI and x is  ln Er .

The linear regression result showed a strong relationship between the eld-measured
LAI and  lnEr (R2 = 0.734, RMSE = 0.439).
To assess the reliability of the model, the accuracy of the LAIs predicted using the
full-waveform lidar was evaluated using the other 15 eld-measured LAIs. Figure 3 is a
scatter plot of the full-waveform lidar-predicted LAI against the eld-measured LAI.
The results in Figure 3 indicated that the accuracy of the full waveform lidar-predicted
LAI is high (R2 = 0.724, RMSE = 0.449).

5
R2 = 0.734
RMSE = 0.439
4
Field-measured LAI

0
0 1 2 3 4 5
lnEr

Figure 2. The scatter plot of the eld-measured LAI versus  lnEr and the corresponding regression
model (n = 25).
 REMOTE SENSING LETTERS 117

4
R2 = 0.724
RMSE = 0.449
3

Predicted LAI
2

0
Downloaded by [Wuhan University] at 06:47 10 November 2015

0 1 2 3 4
Field-measured LAI

Figure 3. The scatter plot of the full-waveform lidar-predicted LAI against the eld-measured LAI
(n = 15, the solid line is the 1:1 line).

Many previous studies have demonstrated that airborne discrete lidar data cannot
accurately estimate the LAI of short, dense vegetation because discrete lidar does not
yield sucient ground returns due to its poor penetration ability in high-LAI regions
(Luo et al. 2015; Richardson, Monika Moskal, and Kim 2009). However, the results in
this study indicated that the LAI of short, dense vegetation can be estimated based on
the BeerLambert law using airborne full-waveform lidar data. A possible explanation
is that the full-waveform lidar system records the entire backscattered energy, which
greatly improved the penetration ability in high-LAI regions (Richardson, Monika
Moskal, and Kim 2009). In addition, a method of faint returns retrieval was also
proposed to better obtain ground returns in this study, which also greatly contributed
to the LAI estimation of vegetation. In addition, unlike the previous studies, this paper
utilized the lidar intensity instead of lidar counts to estimate the LAI of maize, which
was also helpful for the improvement of LAI estimation accuracy, as demonstrated by
Luo et al. (2014).
Since the scanning angles in this area were less than 12, the inuence of scanning
angles on lidar intensity was not taken into account in this study. Future studies
should evaluate the inuence of scanning angles on lidar intensity values. In addition,
this paper only conducted the LAI estimation of maize using airborne full-waveform
lidar data, and the ability of airborne full-waveform lidar data has not yet demon-
strated for other low-height vegetation types with dierent canopy structure.
Moreover, the method of faint returns retrieval proposed in this paper may not be
suitable for all vegetation types. Therefore, future studies should be carried out to
explore the potential of full-waveform lidar data for estimating parameters of other
short vegetation types, and further studies on faint returns retrieval for low-height
vegetation are also needed.
 118 S. NIE ET AL.

5. Conclusions
In this study, a method for retrieving the LAI of short, dense vegetation using airborne
full-waveform lidar data was proposed and applied to the LAI estimation of maize. To
estimate LAI of maize, three steps were performed. First, the processing of point clouds
and waveform were conducted to generate DTM and accurately retrieve waveform
parameters, respectively. In addition, a method of faint returns retrieval was also
proposed to obtain ground returns. Second, the LAIs of maize were estimated using
the method based on the BeerLambert law. Finally, the LAI estimations were validated
by eld-measured LAIs. Results showed that the LAI of maize was successfully estimated
using airborne full-waveform data with an RMSE of 0.449.
This research can be useful for improving LAI estimation accuracy of short vegeta-
tion. Further study is needed to develop the method of ground return retrieval,
which is very critical for the LAI estimation of short vegetation. In addition, to better
Downloaded by [Wuhan University] at 06:47 10 November 2015

estimate the LAI of short vegetation, future studies should also place an emphasis on
intensity correction algorithms to decrease the inuence of scanning angles on lidar
intensity.

Acknowledgements
The authors would like to thank two anonymous reviewers for the valuable comments and
suggestions.

Disclosure statement
No potential conict of interest was reported by the authors.

Funding
This work was supported by Projects of International Cooperation [2013DFG21640]; the Bureau of
International Cooperation, Chinese Academy of Sciences [241311KYSB20130001]; the National
Natural Science Foundation of China [41171279], [41271428].

References
Allouis, T., S. Durrieu, C. Vega, and P. Couteron. 2013. Stem Volume and Above-Ground Biomass
Estimation of Individual Pine Trees from Lidar Data: Contribution of Full-Waveform Signals. IEEE
Journal of Selected Topics in Applied Earth Observations and Remote Sensing 6 (2): 924934.
doi:10.1109/jstars.2012.2211863.
Armston, J., M. Disney, P. Lewis, P. Scarth, S. Phinn, R. Lucas, P. Bunting, and N. Goodwin. 2013.
Direct Retrieval of Canopy Gap Probability Using Airborne Waveform Lidar. Remote Sensing of
Environment 134: 2438. doi:10.1016/j.rse.2013.02.021.
Axelsson, P. 2000. DEM Generation from Laser Scanner Data Using Adaptive TIN Models.
International Archives of Photogrammetry and Remote Sensing 33 (B4/1; PART 4): 111118.
Capodici, F., G. DUrso, and A. Maltese. 2013. Investigating the Relationship between X-Band SAR
Data from COSMOS-SkyMed Satellite and NDVI for LAI Detection. Remote Sensing 5 (3): 1389
1404. doi:10.3390/rs5031389.
 REMOTE SENSING LETTERS 119

Cui, Y., K. Zhao, W. Fan, and X. Xu. 2011. Retrieving Crop Fractional Cover and LAI Based on
Airborne Lidar Data. Journal of Remote Sensing 15 (6): 12761288.
Duong, V. H., R. Lindenbergh, N. Pfeifer, and G. Vosselman. 2008. Single and Two Epoch Analysis
of ICESat Full Waveform Data over Forested Areas. International Journal of Remote Sensing 29
(5): 14531473. doi:10.1080/01431160701736372.
Lefsky, M. A., W. B. Cohen, G. G. Parker, and D. J. Harding. 2002. Lidar Remote Sensing for
Ecosystem Studies Lidar, an Emerging Remote Sensing Technology that Directly Measures the
Three-Dimensional Distribution of Plant Canopies, Can Accurately Estimate Vegetation
Structural Attributes and Should be of Particular Interest to Forest, Landscape, and Global
Ecologists. BioScience 52 (1): 1930.
Li, W., Z. Niu, N. Huang, C. Wang, S. Gao, and C. Wu. 2015. Airborne Lidar Technique for Estimating
Biomass Components of Maize: A Case Study in Zhangye City, Northwest China. Ecological
Indicators 57: 486496. doi:10.1016/j.ecolind.2015.04.016.
Lin, X., and J. Zhang. 2014. Segmentation-Based Filtering of Airborne LiDAR Point Clouds by
Progressive Densication of Terrain Segments. Remote Sensing 6 (2): 12941326. doi:10.3390/
rs6021294.
Downloaded by [Wuhan University] at 06:47 10 November 2015

Luo, S., C. Wang, G. Li, and X. Xi. 2013. Retrieving Leaf Area Index Using ICESat/Glas Full-Waveform
Data. Remote Sensing Letters 4 (8): 745753. doi:10.1080/2150704x.2013.790573.
Luo, S., C. Wang, F. Pan, X. Xi, G. Li, S. Nie, and S. Xia. 2015. Estimation of Wetland Vegetation
Height and Leaf Area Index Using Airborne Laser Scanning Data. Ecological Indicators 48: 550
559. doi:10.1016/j.ecolind.2014.09.024.
Luo, S., C. Wang, X. Xi, and F. Pan. 2014. Estimating FPAR of Maize Canopy Using Airborne
Discrete-Return Lidar Data. Optics Express 22 (5): 5106. doi:10.1364/oe.22.005106.
Magruder, L. A., A. L. Neuenschwander, and S. P. Marmillion. 2010. Lidar Waveform Stacking
Techniques for Faint Ground Return Extraction. Journal of Applied Remote Sensing 4: 043501.
doi:10.1117/1.3299657.
Mallet, C., and F. Bretar. 2009. Full-Waveform Topographic Lidar: State-of-the-Art. ISPRS Journal of
Photogrammetry and Remote Sensing 64 (1): 116. doi:10.1016/j.isprsjprs.2008.09.007.
Morsdorf, F., B. Koetz, E. Meier, K. I. Itten, and B. Allgoewer. 2006. Estimation of LAI and Fractional
Cover from Small Footprint Airborne Laser Scanning Data Based on Gap Fraction. Remote
Sensing of Environment 104 (1): 5061. doi:10.1016/j.rse.2006.04.019.
Pirotti, F. 2011. Analysis of Full-Waveform Lidar Data for Forestry Applications: A Review of
Investigations and Methods. Iforest - Biogeosciences and Forestry 4 (3): 100106. doi:10.3832/
ifor0562-004.
Pirotti, F., G. Laurin, A. Vettore, A. Masiero, and R. Valentini. 2014. Small Footprint Full-Waveform
Metrics Contribution to the Prediction of Biomass in Tropical Forests. Remote Sensing 6 (10):
95769599. doi:10.3390/rs6109576.
Riao, D., F. Valladares, S. Conds, and E. Chuvieco. 2004. Estimation of Leaf Area Index and
Covered Ground from Airborne Laser Scanner (Lidar) in Two Contrasting Forests. Agricultural
and Forest Meteorology 124 (34): 269275. doi:10.1016/j.agrformet.2004.02.005.
Richardson, J. J., L. Monika Moskal, and S.-H. Kim. 2009. Modeling Approaches to Estimate
Eective Leaf Area Index from Aerial Discrete-Return Lidar. Agricultural And Forest
Meteorology 149 (67): 11521160. doi:10.1016/j.agrformet.2009.02.007.
Solberg, S. 2010. Mapping Gap Fraction, LAI and Defoliation Using Various ALS Penetration
Variables. International Journal of Remote Sensing 31 (5): 12271244. doi:10.1080/
01431160903380672.
Sun, G., K. Ranson, D. Kimes, J. Blair, and K. Kovacs. 2008. Forest Vertical Structure from GLAS: An
Evaluation Using LVIS and SRTM Data. Remote Sensing of Environment 112 (1): 107117.
doi:10.1016/j.rse.2006.09.036.
Wagner, W., A. Ullrich, V. Ducic, T. Melzer, and N. Studnicka. 2006. Gaussian Decomposition and
Calibration of a Novel Small-Footprint Full-Waveform Digitising Airborne Laser Scanner. ISPRS
Journal of Photogrammetry and Remote Sensing 60 (2): 100112. doi:10.1016/j.
isprsjprs.2005.12.001.
 120 S. NIE ET AL.

Wang, X., X. Cheng, P. Gong, H. Huang, Z. Li, and X. Li. 2011. Earth Science Applications of ICESat/
GLAS: A Review. International Journal of Remote Sensing 32 (23): 88378864. doi:10.1080/
01431161.2010.547533.
Watson, D. J. 1947. Comparative Physiological Studies on the Growth of Field Crops. I (http://aob.
oxfordjournals.org/content/11/1/41.extract). Variation in Net Assimilation Rate and Leaf Area between
Species and Varieties, and within and between Years. Annals of Botany 11 (41): 4176.
Zhao, K., and S. Popescu. 2009. Lidar-Based Mapping of Leaf Area Index and its use for Validating
GLOBCARBON Satellite LAI Product in a Temperate Forest of the Southern USA. Remote Sensing
of Environment 113 (8): 16281645. doi:10.1016/j.rse.2009.03.006.
Zheng, G., and L. M. Moskal. 2009. Retrieving Leaf Area Index (LAI) Using Remote Sensing:
Theories, Methods and Sensors. Sensors (Basel) 9 (4): 27192745. doi:10.3390/s90402719.
Downloaded by [Wuhan University] at 06:47 10 November 2015