IEEE Geoscience and Remote Sensing Magazine - March 2020 - 131

HYDRODYNAMIC MODELS
2D flood-inundation modeling has been widely calibrated
using point-based in situ gauges and profile-based satellite radar-altimetry measurements. The unique ability of
InSAR to capture wetlands' 2D 2h/2t suggests a new type
of measurement that can be used for flood-inundationmodel calibration.
The authors of [75] focused on the feasibility of using SAR
interferometry to support 2D hydrodynamic-model calibration in the Louisiana wetlands. The LISFLOOD-ACC hydrodynamic model [76] has been set up in the central Atchafalaya Basin Floodway System in Louisiana during typical spring
floods with high flows. For Manning's roughness, Jung et al.
[75] used two different calibration schemes. First, the model
was calibrated using water elevations at a single in situ gauge,
which represented a more traditional approach. Due to the
gauge location in the channel, the calibration showed more
sensitivity to the channel roughness relative to wetland
roughness. Second, the model was calibrated in terms of
2h/2t calculated from an ALOS PALSAR interferogram generated from 16 April to 1 June 2008. The calibration showed
more dependency on wetland roughness. The best-fit model
showed a mean 2h/2t error of 5.7 cm/46 d (Figure 10). Overall, the modeling results suggested that the 2D 2h/2t from
InSAR in the Atchafalaya basin offered an improved understanding and modeling of wetland hydrodynamics.

have less spatial coverage than the image-based InSAR method. Thus, many wetland water bodies are not covered by
either dataset. L-band SAR-backscattering coefficients have
been used to obtain the offset at a fish farm, based on the
fact that the SAR-backscattering coefficient changes with the
water level [40]. L-band SAR-backscattering coefficients have
been found to correlate with water-level changes in wetlands
and can be used to estimate them [41], [77]. Therefore, it is
possible to use this backscattering variation to estimate the
offsets for InSAR processing in wetlands [41], [42], [77]. However, the challenge lies in the complex relationship between
the backscattering variations and water-level changes due to
the spatial heterogeneous vegetation types and density.

FUTURE TRENDS

FILTERING
To remove the speckle and random noise, many traditional
InSAR filtering algorithms use the window-bus technique
[78]. However, the wetland-area resolution becomes compromised. Since the aforementioned MB InSAR processes
data on a 3D domain (with the third domain serving as
the observation with different baseline lengths), the mean
in the baseline domain can be used instead of the spatial
mean, and, therefore, the nonlocal noise-filtering methodology [79] can be applied. For example, Ferraioli et al. [80]
applied the MB InSAR framework to denoise the interferometric phase gradient nonlocally. Applying the nonlocal
noise-filtering methodology to a wetland study area could
be an interesting future research project.

OFFSET ESTIMATION THROUGH SAR INTENSITY
The offset-estimation method that uses in situ gauges and
radar altimeters has a spatial-coverage limitation. In other
words, in situ gauge measurement is point-based, while radar-altimeter measurement is profile-based, and they both

NEW SATELLITE MISSIONS
The low temporal resolution and temporal decorrelation effects are expected to be solved to a great extent through current and future satellite missions. For example, the C-band
SAR-system Sentinel-1 has a revisiting period of 6 d, which

Riv
er
Flo
w

∆h

Lf
t

2
dmax

1
dmax

1
halt

1
dVS

hbathymetry

2
halt

hbathymetry

∆h
t2
dVS

d t2 = 0
Distal Floodplain

d t2 = d t1 - ∆h
Proximal Floodplain
River

Floodplain
(a)

Upland

River

Floodplain
(b)

Upland

FIGURE 9. The schematic plot of the wetlands geomorphology and water surfaces at (a) the high-water-level season (t 1) and (b) with the

water level decreased by Th (t 2). Yellow triangles represent the maximum water-depth locations. The purple area in (b) represents the
region with water at t 1 and without water at t 2 (from [13]).
MARCH 2020

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

131

IEEE Geoscience and Remote Sensing Magazine - March 2020

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