IEEE Geoscience and Remote Sensing Magazine - March 2020 - 122

	

m
Tt = - 4r { def , (3)

where Tt is the surface deformation and m is the SAR system's wavelength.
BACKSCATTERING IN WETLANDS
The radar signal's interaction with wetland vegetation is a
complex process that depends on the radar's wavelength
and polarization mode, the vegetation type, and the water
level. Based on the physics of radar scattering [22], the scattering mechanism across the vegetation includes surface
scattering, volume scattering, and double-bounce scattering, as illustrated in Figure 2. It is generally agreed that a
SAR system with a longer wavelength has a stronger ability
to penetrate vegetation canopies [22], [23]. Thus, the scattering mechanism across the vegetation depends on the
SAR system's wavelength and the vegetation type. For example, the dominant scattering mechanism for farmland
at the C band (wavelength: +5.6 cm) is volume scattering, while the at the L (wavelength: + 23.6 cm) and P band
(wavelength: +68 cm) it is surface scattering because there
is more penetration [22].
Most spaceborne SAR systems operate in L-, C-, or
X-band wavelengths. The L-band SAR signal with a wavelength of 23.6 cm can penetrate forest canopy and herbaceous vegetation [24], [25]. The dominant scattering of
L-band SAR across wetlands varies with water levels. During dry seasons, the volume scattering from the canopy
becomes the dominant mechanism. With the presence of
water beneath the vegetation, double-bounce backscattering occurs. However, if the water level is high enough to
submerge the herbaceous wetlands, specular scattering
becomes dominant. Since the water level is usually lower

than the forest top, double-bounce backscattering becomes
prevalent in flooded swamp forests.
On the other hand, while C-band SAR systems with a
wavelength of 5.6 cm penetrate herbaceous vegetation, the
microwave energy from C-band SAR is believed to be too
weak to penetrate dense canopy and yield double-bounce
backscattering between tree trunks and water surfaces [25].
Hong and Wdowinski [26] examined all of the quadrantpolarimetric interferograms from Radarsat-2 across the Everglades herbaceous wetlands and suggested that the double
bounce is the dominant backscattering for all four polarization modes. The polarization modes include horizontal
transmit and horizontal receive (HH), horizontal transmit
and vertical receive (HV), vertical transmit and horizontal
receive (VH), and vertical transmit and vertical receive (VV).
However, the C-band SAR signal was able to maintain double-bounce backscattering during the leaf-off season when
there was a medium-low canopy closure [27].
The authors of [28] examined the wetland feasibility of
multipolarized X-band SAR images with a wavelength of
3.1 cm. The X-band signal interacts mostly with the lower
trunks of the vegetation, and multipolarized (that is, HH,
HV, VH, and VV) interferograms showed the water-levelchange signals. Hong et al. [28], [29] suggested that, in the Everglades, a combination of double- and multiple-bounce scattering formed the dominant X-band SAR-image mechanism.
COHERENCE ACROSS WETLANDS
Coherence is an important measure of InSAR accuracy.
InSAR coherence is calculated using
	

c total =

S 1 S )2
. (4)
S 1 S )1 S 2 S 2)

Water

Surface Backscattering

Volume Backscattering

Double-Bounce
Backscattering

Enhanced Double-Bounce
Backscattering

(a)

Water

Surface Backscattering

Volume Backscattering
(b)

Double-Bounce
Backscattering

Specular Backscattering

FIGURE 2. The SAR backscattering across the (a) forest and (b) herbaceous wetlands that results from canopy-volume scattering, doublebounce backscattering, and specular scattering (modified from [42]).

122

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

MARCH 2020
IEEE Geoscience and Remote Sensing Magazine - March 2020

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2020
