PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
January 2014
33
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
January 2014
33
Abstract
Fully polarimetric L-band Synthetic Aperture Radar (
SAR
)
backscatter was collected using
NASA
’s Unmanned Aerial
Vehicle (
UAV
)
SAR
and regressed with in situ measure-
ments of basal area (
BA
) and above ground biomass (
AGB
)
of mature loblolly pine stands in North Carolina. Results
found
HH
polarization consistently displayed the lowest
correlations where
HV
and
VV
exhibited the highest cor-
relations in all groups for both
BA
and
AGB
. When planta-
tion stands were analyzed separately (plantation versus
natural), correlation improved significantly for both
BA
(R
2
= 0.65,
HV
) and
AGB
(R
2
= 0.66,
VV
). Similarly, results
improved when natural stands were analyzed separately
resulting in the highest correlation for
AGB
(R
2
= 0.63,
HV
and
VV
). Data decomposition using the Freeman 3-compo-
nent model indicated that the relative low correlations were
due to the saturation of the L-band backscatter across the
majority of the study area.
Introduction
The derivation of forest biophysical factors using remote sens-
ing approaches has the potential to minimize requirements
for labor intensive, ground-based estimates and facilitate the
generation of data sets across extended geographic areas. The
objective of this study was to assess the relationship of high
resolution fully polarimetric L-band Synthetic Aperture Radar
(
SAR
) backscatter collected using
NASA
’s Unmanned Aerial
Vehicle (
UAV
)
SAR
with
in situ
measurements of basal area
(
BA
) and above ground biomass (
AGB
) of mature loblolly pine
stands using simple linear regression.
UAV
platforms with optical or lidar sensors have been
especially useful in addressing optimal spatial and spectral
resolutions and repeat cycles in the acquisition of environ-
mental data (Berni
et al.
, 2009). For example, the assessment
of forest biophysical parameters utilizing
UAV
platforms have
been demonstrated for estimating canopy structure by extract-
ing three-dimensional canopy surfaces from high resolution
two-dimensional imagery (Dandois and Ellis, 2010). Leaf area
index (
LAI
), chlorophyll content, water stress detection, and
canopy temperature have been estimated indirectly using a
John S. Iiames and Ross S. Lunetta are with the United States
Environmental Protection Agency, 109 T.W. Alexander Dr.,
MD E243-05, Research Triangle Park, North Carolina 27711
(
).
William L. Marks and Siamak Khorram are with North
Carolina State University, Department of Forestry and
Environmental Resources, Box 7106, Raleigh, NC 27695.
Thomas H. Mace is with the U.S. National Aeronautics and
Space Administration, Office of the Associate Director for
Operations, Dryden Flight Research Center, Edwards, CA
93523.
Photogrammetric Engineering & Remote Sensing
Vol. 80, No. 1, January 2014, pp. 33–42.
0099-1112/14/8001–33/$3.00/0
© 2014 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.80.1.33
helicopter
UAV
mounted with a thermal and hyperspectral
optical sensor (Zarco-Tejada
et al.
, 2009). The benefits of
employing a
UAV
SAR
system for the study of forest biomass
include: (a) counteracting the difficulties of wind gusts and
turbulence in flying the same pass on multiple occasions, and
(b) maintaining antennae angle despite differing yaw angles
created by cross-wind issues (Rosen
et al.
, 2006). The ability
to maintain aerial track and antennae pointing capabilities
warrant the use of these controlled flight management systems
(Rosen
et al.
, 2006).
Factors affecting forest
AGB
estimation from radar
backscatter include forest structure (size and age class,
stand density, branch angular patterns), site characteristics
(slope, aspect, soil moisture content), dialectric constant
(plant water content and specific gravity), radar measure-
ment geometry patterns (incidence angle and spatial resolu-
tion), and radar bandwidth (Robinson
et al.
, 2013). Specific
to loblolly pine forest structure, leaf biomass accumulation
also varies inter-annually by producing two to three nee-
dles flushes throughout the growing season, with maximum
leaf biomass (i.e.,
LAI
) occurring in August (Dewey
et al.
,
2006). Site characteristics such as fertility and drought
also affect loblolly pine canopy architecture indicated by
significant variation in indeterminate growth (multiple
flushes) and high plasticity (i.e., developmental patterns)
in foliage accretion and abscission (Sampson
et al.
, 2003;
Iiames
et al.
, 2008).
Past
SAR
studies have used L-band radar preferentially
over shorter wavelength (i.e., C-band or X-band) systems to
maximize above ground biomass (
AGB
) sensitivity (Imhoff
et al.
, 1998). A study by Wu and Sader (1987) examined cor-
relations on 18 forested plots with fully polarimetric
NASA
AIRSAR
(L-band) backscatter to derive vegetative parameter
data for natural pine, bottomland hardwood, and swamp for-
est with deciduous understory across the southern US Gulf
Plain physiographic region. Wu and Sader (1987) documented
the potential application of
SAR
to quantify vegetation charac-
teristics including total-tree biomass, basal area (
BA
), and tree
height. Past research has focused on establishing correlations
between tree height,
BA
, and
AGB
using backscatter responses
from single polarizations and polarization ratios. Results from
linear regression analysis have determined that the best cor-
relations were achieved with the
VH
channel, with all three
polarizations ranging from (R
2
= 0.50 to 0.82) for
BA
, (R
2
= 0.45
to 0.52) for tree height, and (R
2
= 0.74 to 0.89) for
AGB
(Wu and
Sader, 1987). Additionally, the
HV
channel dominated canopy
Basal Area and Biomass Estimates of Loblolly
Pine Stands Using L-band UAVSAR
William L. Marks, John S. Iiames, Ross S. Lunetta, Siamak Khorram, and Thomas H. Mace
