IEEE Geoscience and Remote Sensing Magazine - March 2020 - 75

In addition, the performance of SAR images with different wavelengths in DInSAR-based mining displacement detection has also been discussed [50], [57]-[59]. The results
suggest that SAR data sets with longer wavelengths (e.g.,
L band) have a greater ability to measure mining-induced
displacement than do those with shorter wavelengths (e.g.,
X and C bands), especially for those areas with large displacement gradients due to their better performance in
reducing the impact of interferometric decorrelation and
phase discontinuity.
To reveal the temporal evolution of mining deformation,
TS-InSAR techniques have been exploited in mining areas.
For example, Colesanti et al. [60] detected the TS displacement associated with iron mining in Lorraine, France, from
57 ERS-1/2 SAR images spanning 1995 to 2000 using the
PS-InSAR method. Following this, some successful cases
of PS-InSAR-based TS mining displacement measurements
were reported (e.g., [61]-[66]). In fact, compared with the
strategy of the one single-reference image of PS-InSAR, the
SBAS-InSAR method shows a better performance in retrieving TS mining displacement, owing to the fact that its strategy of multiple reference images with small baselines can
minimize the effect of spatiotemporal decorrelation. This
has been proven in many works, such as [40] and [67]-[71].
Furthermore, tests on measuring TS mining displacement using other TS-InSAR techniques such as intermittent
SBAS [72], [73], SqueeSAR [74], [75], and the temporarily
coherent point InSAR method [76] were also conducted
in recent years. Note that recent attempts were made to
retrieve TS mining displacement using large gradients by
combining DInSAR observations with OT measurements
[77], [78], range split-spectrum interferometry measurements [79], and 3D laser scanner observations [80].

W
d LOS
cos i sin i sin a h - sin i cos a h
E, (6)
<
F>N H = ;
0
d Azi
cos a h
sin a h
E

where a h and i are the flight and incidence angles of
the SAR sensor, respectively. This implies that LOS and
azimuth displacement measurements can perfectly reveal deformation patterns only when the ground surface
points move completely along the LOS or azimuth directions. However, such a scenario is rather rare in practice.
For instance, Figure 1 shows a typical deformation basin

At
y

An
e

lle

gl

ra

An

Pa
l

en
ar

l
cu

di

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

	

rp

MARCH 2020

LIMITATIONS OF SAR/InSAR 1D/2D DISPLACEMENT
MEASUREMENTS
As stated previously, InSAR/SAR methods can offer
1D/2D surface deformation only along the LOS and/
or azimuth directions, i.e., d LOS and d Azi, from a single
SAR/InSAR pair using the DInSAR, MAI, or OT techniques. The InSAR-/SAR-derived 1D/2D displacements
are actually the projections of the real surface 3D deformation components in the vertical, east, and north
(denoted by 6W, E, N@T) onto the LOS and azimuth directions, respectively [15], [86]:

Pe

2D LOS AND AZIMUTH MINING DISPLACEMENT
MEASUREMENTS
Compared with the extensive tests of InSAR-based, 1D LOS
displacement measurements in mining areas, works focusing on simultaneously retrieving 2D LOS azimuth mining
displacements are rare, to the best of our knowledge. In
2013, Zhao et al. [81] generated the first large-scale mining
deformation map (with a maximum subsidence of more
than 4 m) along the LOS direction from Advanced Land Observing Satellite (ALOS) phased-array type L-band SAR (PALSAR)-1 images using the OT technique and further obtained
the accumulated LOS deformation in the Bulianta and
Shangwan coal mining areas of China. However, mininginduced azimuth displacement maps could not be presented
in [81], possibly due to the coarse azimuth resolution of selected SAR images and their smaller horizontal movement
compared to vertical subsidence in these mining areas.
In 2015, Chen et al. [82] generated 2D LOS and azimuth mining deformation maps over the Daliuta coal
mining area of China from high-resolution (roughly 1 m)
TerraSAR-X images and demonstrated the relative error of
the derived results vis a vis the maximum deformation to

be approximately 4.2% by comparing the results with GPS
measurements. Following this, other cases and improved
OT methods for OT-based 2D mining deformation measurement (e.g., [83]-[85]) have been reported. Aside from
the OT techniques, combining DInSAR with MAI has been
proven to be a feasible strategy to retrieve 2D mining deformation along both the LOS and azimuth directions [46].

sin

ng

ini

M

Ba

n

ide

s
ub

ce

S

FIGURE 1. A 3D view of the SAR-/InSAR-based, mining-induced
deformation measurement. The black arrows denote the real
deformation vectors of the ground surface points. The red arrows
represent the SAR/InSAR-measured LOS deformation vectors at
three points, where the surface moves perpendicular to (i.e., zero
vector), parallel to, and with an angle to the LOS direction.

75
IEEE Geoscience and Remote Sensing Magazine - March 2020

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