IEEE Geoscience and Remote Sensing Magazine - March 2020 - 32

optimization to estimate the time-series InSAR deformation parameter. To speed up execution time, some
fast algorithms have been proposed [50], [82], [83].
Some methods consider installing artificial reflectors
in the study area to assist with obtaining deformation
information [84]-[88]. On one hand, the use of artificial reflectors can mitigate the lack of PSs and DSs,
which increases the spatial sampling density. Conversely, the experimental results can be verified on the artificial reflectors, and the estimated results are compared
with the real results.
It is also important to analyze the decorrelation noise
and atmospheric delay of time-series InSAR and the performance of the algorithm. In [89] and [90], the authors
developed a covariance model over temporal decorrelation and spatial decorrelation for interferometric phase
noise that can be applied to evaluate the performance of
time-series InSAR techniques. Guarnieri and Tebaldini
[91] derived hybrid Cramer-Rao bounds for surface deformation field estimators in InSAR, which provide a quick
performance assessment of an InSAR system as a function
of its configuration (wavelength, resolution, and SNR), the
intrinsic scene decorrelation, and APS variance. In [92] and
[93], the authors introduced an error-estimation method in
time-series InSAR techniques, which obtained more robust
parameter estimation.
APPLICATIONS
For the past two decades, along with the progress of timeseries InSAR techniques, a wide range of corresponding
appli-c ations have been developed. The applications of
time-series InSAR techniques include urban subsidence, infrastructure deformation, volcanoes, earthquakes, mining
areas, landslides, and so on. As mentioned previously, we
introduced the following advantages of time-series InSAR
techniques in deformation retrieval: high accuracy, large
scale, and low cost. However, data requirements, method
adoption, and acquisition results vary considerably from
one mission to another.
First, in different applications, the observation targets
are not the same, so the requirements for the resolution,
band, and extent of the SAR image are quite different. For
example, the monitoring of large-scale subsidence and
landslides requires larger images. But the resolution of the
corresponding image does not need to be high, usually 10 m
to tens of meters. For the monitoring of infrastructures such
as bridges, buildings, and railways, high-resolution images
are needed; otherwise, the target will not be recognized effectively. For various ground scenes, SAR has different acquisition capabilities in different bands. For observing cities, it
is better to have short-wavelength data (e.g., X band), while
in mountainous areas, long-wavelength data are more appropriate (e.g. L band). Second, due to the differences in
data, the effect varies from different processing technologies. For example, we should consider the use of PSs and/
or DSs, the choice of deformation models, the influence of
32

atmospheric effects, and so on. Third, because the time-series InSAR technique can retrieve average linear deformation rates, elevation corrections, time-series deformations,
periodic deformations, thermal expansion components,
atmospheric delay signals, and so on, we need to confirm
which signals of interest are extracted and how to best present them.
All of these situations require specific considerations.
In the following sections, three specific applications
that illustrate the differences among various applications
are discussed.
CITIES
With the acceleration of urbanization, the urban population
has gradually increased, and there are more and more infrastructures, such as buildings, roads, and bridges. As a result,
monitoring cities is crucial. Urban areas contain a large number of artificial buildings, and there are many stable scatterers, which provide high SNR signals for the application of
time-series InSAR.
In cities, subsidence is an important geological phenomenon. The main reasons for subsidence occurrence are the
extraction of groundwater, geological activities, and other
human and natural factors. Urban surface subsidence mostly resembles a funnel, which brings great harm. When the
subsidence exceeds a certain threshold, it leads to building
collapse and home destruction, endangering people's lives
and property. To monitor surface subsidence, more attention is paid to average subsidence rate. Its monitoring range
tends to be large, but it is not overly concerned with the
details of characterization; therefore, medium-resolution
images with wide coverage are used.
There have been numerous related applications and analysis cases. For instance, in [94], the authors used PSI to analyze the subsidence caused by groundwater exploitation in
Mexico City and compared the results with GPS data. Pepe
et al. [95] retrieved deformation results in Naples, Italy, using the SBAS method. Stramondo et al. [96] used IPTA to
analyze 10-year deformation results in Rome, Italy, from
1995 to 2005. The authors in [97] used SBAS to analyze
the subsidence results in Guangzhou, China, from 2006 to
2012 and compared the data with various bands and leveling results. From this study, it was determined that the
overexploitation of groundwater is the main reason for
subsidence. Qu et al. [98] obtained deformation results
in Xi'an City, China, and used them to estimate building
height information. Wu et al. [99] used the SBAS method
to derive the surface deformation of Taiyuan City, China,
which was mainly caused by groundwater exploitation and
mining. In [100], the authors analyzed the deformation results caused by groundwater exploitation in Beijing, using
PSI and SBAS, respectively, which are in good agreement
with the leveling data.
In this section, a specific case study is presented. Fig--
ure 14(a) shows the deformation map of some parts of
Beijing, including the Fangshan, Mentougou, Fengtai,
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