Synthetic Aperture Radar Interferometry

(InSAR): Principles and Applications

SAR Interferometry (InSAR):
Synthetic Aperture Radar (SAR) Interferometry (InSAR) means Measurement of
certain parameters viz., topography, surface change and surface displacement from the
interference of phase of two or more SAR acquisitions over the same area. SAR systems
record both amplitude and phase of the backscattered echoes. It has been observed that
the phase spectrum of an image is more characteristic to the image than its magnitude
spectrum (Ghiglia and Pritt, 1998). However, the phase of each pixel of a focussed SAR
image is the sum of three distinct contributions

the two -ways travel path (sensor-target-sensor, which is hundreds of kilometres in

case of spaceborne SAR system) divided by the used radar wavelength (a few
centimetres), corresponds to millions of cycles,

the interaction between the incident electromagnetic waves and the scatterers
within the ground resolution cell,

the phase shift induced by the processing system used to focus the image.

Therefore, the phase of a single SAR image is of no practical use (Rocca et al.,
1997). On the contrary, if two SAR images from slightly different viewing angles are
considered (for interference), their phase difference can be exploited efficiently to
generate digital elevation models (DEMs), to monitor terrain changes and to measure
minute surface displacements. It seems that interferometric technique was first reported
by Rodgers and Ingalls (1969) while mapping the planet Venus using an earth-based
range Doppler radar system. However, terrestrial use of SAR interferometry was first
proposed by Graham (1974), later extended by Zebker and Goldstein (1986) and was
further improved by Li and Goldstein (1990).
In SAR interferometry technique, the phases of two SAR images of the same area
are made to interfere to generate an interferogram having a number of interference

fringes. Each interference fringe corresponds to a phase difference of 2 between the two
component SAR images. The interferogram is the complex product of two component
SAR images. These two SAR images can be taken simultaneously by an aircraft/satellite
with two antennae separated by a distance known as baseline. This is referred to as
Single Pass or Single Track Interferometry. On the other hand, if the two SAR images
are acquired in two different passes of the same satellite, this is referred to as Repeat
Pass Interferometry. The two SAR images acquired by two different satellites revolving
along more or less the same orbit with certain time difference (as in case of ERS-1 and
ERS-2) may also be used for generating an interferogram.

SAR Interferometric Satellites

At present, no satellite is available for acquiring data for single -pass
interferometry. Single-pass interferometry is however possible presently in airborne and
space-borne SAR systems. In airborne SAR interferometry, orbital correction is a
difficult task. In Shuttle Radar Topographic Mission (SRTM) launched by USA , for a
period of 11-days during February 11-22, 2000, the space-borne SAR system has
acquired single-pass InSAR data over the entire globe. Digital elevation model (DEM)
generated from the acquired single-pass InSAR data is freely available. For repeat-pass
interferometry, satellite data are available from ERS-1, ERS-2, JERS-1 (archived data
only), RADARSAT-1 and ENVISAT (presently functional) satellites .

Interferometric SAR Data Structure

In contrast to standard data where only amplitude is available for each pixel,
interferometric data are supplied in complex format (e.g., single look complex or SLC
format), to enable extraction of amplitude as well as phase information as described
below.

Each pixel in SLC format is represented by a+ib.

Now, Amplitude = a 2 + b 2 , Phase = tan 1

b
a

For each pixel, a and b together occupy 8 bytes (4 bytes each) in case of
complex float format or 4 bytes (2 bytes each) in case of complex integer format which is
2

usually supplied by NRSA Data Ce ntre, India. On the other hand, in case of standard
data, each pixel may be represented by a single character i.e., just one byte for an
averaged multi look data. This shows the enormity in the size of interferometric data in
comparison to standard data products.

Mathematical Background
Consider an image as a set of values A(x,y) where, X-coordinate is in the
direction of platform motion (Azimuth) and Y-coordinate is in the direction of
illumination (Range). The height of the point P above the reference surface (h=0,
Figure 2) is represented by Z-coordinate.
In Figure 2, the imaging geometry of SAR Interferometry is illustrated. A1 and
A2 represent two antennae viewing the same surface simultaneously or a single antenna
viewing the same surface on two s eparate passes.

X (Azimuth)
A2
90-+

A1 90-+

h
P

z(y)
Y (Range)
Figure 2: Basic geometry of SAR intrferometry. A1 and A2 represent two antennae
viewing the same surface

The elevation of the terrain surface at point P say z(y) is given by

z ( y ) = h .Cos

(1)

Where, h = height of the platform/satellite

= look angle of the radar
= slant range distance of antenna A1
Considering A2A1P of A1PA2, we have the Cosine relationsh ip as given below :
Cos(90-+) =

B 2 + 2 ( + ) 2
2B

...(2)

Where, = difference in slant range distance between the two antennae A1 and A2
= angle made by the baseline vector w.r.t. horizontal line
= slant range distance of antenna A1
Or, 2BCos(90-+) = B2 + 2 - (+)2
Or, 2BCos(90+(-)) = B2 + 2 -2 - 2* - 2
Or, -2BSin(-) = B 2 - 2* - 2
Or, (2 - 2Bsin(-)) = B2 - 2
Or, =

B 2 2
2( BSin ( ))

...(3)

Substituting the value o f in Eqn.2, we get terrain height at point 'P'

z(y) = h -

B 2 2
.Cos
2( BSin( ))

...(4)

The measured quantity is the phase difference between the two antennae (),
which is directly proportional to the path difference ( for single-pass interferometry
and 2 for repeat-pass interferometry) with a constant of proportionality 2/ .
For repeat-pass interferometry-

2
4
* 2 =
*

Or, =

(5)

Substituting the value of in Eqn.4 in terms of phase difference () as obtained from

Eqn.5, we obtain
z( y ) = h

B2 ( ) 2
.Cos
2( BSin( ))

(6)

Differential SAR Interferometry

Consider the radar antenna positions during passes 1 and 2, A1 and A2, are
viewing the same surfa ce and separated by baseline B, tilted with respect to horizontal by
an angle (Figure 3). A2 shows satellite antenna position over another pass (pass 3)
imaging the same area with baseline B with respect to A1 and tilt angle with respect to
horizontal. It is assumed that there has been some surface displacement at point P
between passes 2 and 3. Using D-InSAR techniques, it is possible to estimate change in
position of the point P between passes 2 and 3.
In Figure 3, , (+) and ( + ) are the path lengths for three different passes of
satellite.
For passes 1 and 2, the phase difference is given by :
=

4
*

(7)

Where, is the difference in slant range distance between passes 1 and 2.

B
90-+

A2

A2

90- +

A1

h
P
z(y)

Y(Range)

Figure 3: Geometry of Differential SAR Interferometry (D-InSAR). A2 and A2 represent

three antennae viewing the same surface. Data acquired from A1 and A2 are
for measuring the topographic phase and that from A2 and A2 (before and
after the event) for measuring ground deformation.
Considering A2A1P of A1PA2, we have the Cosine relationship as illustrated earlier.

Cos(90-+) =

B 2 + 2 ( + ) 2
2B

Or, 2BCos(90-+) = B2 + 2 - (+)2

Or, 2BCos(90-(-)) = B2 + 2 -2 - 2* - 2
Neglecting 2 (being very small) and rearranging the equation, we have Or, 2* = 2BSin(-) + B2
Or, BSin(-) +

B2
2

(8)

For space-borne geometry, the second term on the right hand side of Eqn.8 can be
ignored using parallel ray approximation.
Hence,
BSin(-) B

(9)

Where, B is the component of baseline parallel to the look direction.

Substituting the value of as obtained from Eqn.9 in the Eqn. 7, we obtain
=

4
4
* BSin ( ) =
* B

(10)

Similarly, for passes 1 and 3, the phase d ifference is given by :

=

4
4

* B Sin ( ) =
*B

(11)

Where, B is the component of baseline B parallel to the look direction.

Dividing Eqn.10 by Eqn. 11, we have

B

(12)

The Eqn.12 is independent of topography since radar wavelength is constant. It is

assumed that there is some surface displacement (which may be due to earthquake,
landslide, ground subsidence or volcanic swelling) between passes 2 and 3 in a coherent
manner. As a consequence, in addit ion to topography, there is an additional phase

contribution due to change in radar time for surface displacement along radar look
direction.

So, phase difference for passes 1 and 3 may be modified as given below.
=

4
* ( B + )

(13)

Where, surface displacement along look direction, adds to the topographic

phase term B .
Before differentiating the ph ase values of two interferograms generated from
passes 1 and 2, and 1 and 3 respectively, scaling of interferometric phases (normalization
of interferograms) with respect to their baselines is required. The differential phase,
which is responsible for surface displacement may now be calculated using the following
relation.
=

B
B

(14)

B
4

4
* (B + ) *
* B

Or, =

4
*

(15)

Hence, the component of surface displacement along the radar look direction (line-ofsight), may be calculated using the above equation.

Procedure
The first and the vital step of InSAR procedure is selection of appropriate
interferometric image pair for generating interferogram for a particular application.

Data Selection
Depending on the nature of application, the following factors should be duly
considered for appropriate data selection.
(i) Baseline

The concept of baseline is very important in data selection. Baseline may be (a) spatial
as well as (b) temporal.
(a) Spatial Baseline - In repeat-pass interfe rometry, spatial baseline varies from place to
place and from one pass to another. The recommended spatial baseline in case of ERS
and ENVISAT data pairs varies from 50m. to 300m. for terrain topography and surface
change studies. In case of surface displa cement studies, near-zero ( 5m.) spatial baseline
of the data pairs does not require topographic phase removal in a topographically uneven
terrain.
(b) Temporal Baseline - For some applications viz., topographic mapping, ocean surface
mapping, the time interval between the two SAR acquisitions i.e., temporal baseline often
causes the second image to lose coherence * relative to the first one, a phenomenon called
temporal decorrelation. The phase of repeat -pass interferogram over ocean will be a pure
noise process if the temporal baseline is longer than a few seconds. Surface changes in
shape or dielectric constant such as freezing, thawing, vegetation movements due to wind
or surface changes due to precipitation, harvesting of fields or other cultural activ ities
seriously degrade the quality of the acquired interferogram. However, temporal baseline
is very useful for other applications , where dynamic phenomena are being studied. The
difference in the phases of two scenes, one acquired before the phenomenon a nd another
after, is utilized to determine the changes on the surface due to the said phenomenon.

(Ii) Atmospheric Effects

Atmospheric factors viz., cloud, haze, and particularly their differential effects on two
SAR scenes may sometimes be of significant concern in selecting the image pair.
Atmospheric effects and their correction need further discussion.

(Iii) Orbital Deviation

In repeat-pass interferometry and differential interferometry, the orbit of data
acquisitions should be similar in nature and necessary corrections should be performed
based on the available orbital parameters. This is the reason why either of ascending or
descending pair is recommended for interferogram generation.
In differential interferometry, two pairs of interferometric SAR images are generally
required one for obtaining the topographic phase contribution and the other consisting
9

of two images, one before and another after the said dynamic phenomenon being studied.
One image of the second pair may be common in the first pair, which is then called the
master image. For the first image pair, the spatial baseline should be 50m. to 300m and
temporal baseline should ideally be zero, which for repeat-pass interferometry is small
enough to preserve coherence between the two images. In the second pair, if spatial
baseline is zero, surface displacement can be estimated exclusively from the second pair
without the aid of the first pair. Because, in this case, there is no topographic phase
contribution to the resulting interfero metric phase.

Data Processing
Starting from the processing of SLC data to generation of differential interferogram
consists of two major stages.
A. Generation of two interferograms from two pairs of SAR SLC images, where, one
may be common to both the pairs the master image.
B. Extraction of phase component for surface movements from the resultant phase in the
second interferogram by subtracting the phase due to topography obtained from the
first interferogram.

Generation of interferogram from two SAR SLC images

For generating interferogram, two SLC images are to be registered very accurately in
both amplitude and phase (Rao and Rao, 1999). Registration of two images should be
done at an accuracy of 1/100th of a pixel which may be accomplished using one of the
following techniques (i)

Cross correlation of pixel amplitudes,

(ii) Maximum value of coherence (Prati et al., 1994),

(iii) Maximum signal (amplitude) to noise ratio (SNR) in fringe spectrum (Gabriel and
Goldstein, 1988) and
(iv) Minimization of average fluctuation of the phase difference (Lin et al., 1992).

Subsequent to image registration, the phases of two SLC images are subtracted
from one another and interferogram is generated. This is done by multiplying one input
image with complex conjugate of the other input image.
10

Generation of DEM
Generation of DEM consists of four steps - flattening, phase unwrapping, absolute
phase determination and phase to height conversion.

Flattening
To obtain the phase difference solely due to topographic unevenness, the phase
difference which may be resulted in an ideally flat terrain should be removed from the
resultant interferometric phase. For this purpose, an interferogram is simulated for the
given interferometric geometry for an ideally flat terrain . The phase of this simulated
interferogram is subtracted from the interferometric phase of the given data pair.
The interferometric geometry is used to evaluate the range ramp by using the
following relation.
=

(4 / ) * ( * B )

(16)

Phase ramps in azimuth may sometimes also be present due to change in parallel
component of the baseline along it.

Phase unwrapping
As the height of the terrain increases, the phase also increases steadily. Since phase
values are a periodic function of 2, they automatically get wrapped after reaching 2
(Figure 4). The interferometric phase therefore needs to be unwrapped . Phase
unwrapping may be performed by
(i) path following algorithm or (ii) least square algorithm. Least square technique is
further subdivided in to (a)unweighted robust technique, (b) weighted least square
technique and (c) Picard iteration technique. A detailed account of various phase
unwrapping techniques may be obtained from Ghiglia and Pritt (1998).

11

B. Unwrapped actual phase

A. Wrapped phase

Figure 4: Concept of phase unwrapping showing wrapped (A) and unwrapped (B)
phases
The unwrapped phase may be described as shown in the following relation.

u p = w p + 2 * n

(17)

Where, up is unwrapped phase, wp is wrapped phase and n is a positive integer

(including zero for points at bottom reference surface).
Determination of absolute phase
After phase unwrapping is performed successfully, it is necessary to determine the
multiples of 2 to be added to the measured phase to obtain the absolute phase value.
This is also known as phase calibration. The absolute phase value at each point may be
des cribed as given below.
abs = up + offset

(18)

The simple method to determine offset is to find one target in the interferogram
where both position and height information are known. Using this information, the
integer number of phase cycles over the entire interferogram can be calculated.
It is however desirable to obtain this value from the data directly without
foreknowledge of the surface using some automatic algorithms. Different algorithms
have been in use to determine this residual cycle offset (Madsen, 1995).

Phase to height conversion

This may be carried out using one of the following models.

Normal baseline model,

12

Integrated incidence angle model and

Baseline rotation model.

In normal baseline model, the change in heig ht (h) at any point of the terrain is related
to the change in phase () as described below.
h =

Sin
*
4 * B

(19)

Where, is slant range distance of the first image, B is baseline component

normal to the look direction, is radar wavelength and is incident angle.
The change in height (h) between two adjacent points of the terrain for 2 radian
phase difference () is called altitude of ambiguity.

The step-by-step procedure in SAR interferometry starting from SLC data to

DEM generation has been illustrated in a flow diagram (Figure 5).

Extraction of phase component due to surface movements by D-InSAR technique

The differential phase () which is responsible for surface displacement may be
calculated after normalizing the phase value of the second interferogram with respect to
the baseline of the first interferogram as described earlier.
= ' - (B||' /B||)*
Where, ' and are phase values of two interferograms, and B|| ' and B|| are parallel
components of the baselines (parallel to the line of sight) of the corresponding
interferograms.
Subsequently, the component of surface displacement along the line-of-sight () may be
obtained using the following mathematical relation as described earlier.
=

4
*

13

SAR Raw image

SAR
SLC Processor

Ephemeris data

Ephemeris data

Baseline
Estimation

Sub-pixel
Registration

SAR Raw image

SAR
SLC Processor

Sub-pixel
Registration

Interferogram Generation

Orbital Flattening

Phase Unwrapping

Absolute Phase Determination

Phase to
Height
Conversion

DEM Generation

Geocoding of DEM

Figure 5: Flow diagram describing step -by-step procedure for DEM generation by SAR
interferometry technique.

14

Applications:
SAR Interferometry (InSAR) is a powerful application technique for generating highresolution digital elevation model (DEM) of the Earth's surface directly from SAR image
data. InSAR was first introduced for topographic mapping by Zebker and Goldstein in
1986.
The phase information from each pixel of a SAR image back scattered from the
target surface is utilized for calculating height information. By utilizing the geometry
provided by two marginally displaced, coherent observations of the surface, phase
difference between the two observations, which is the difference in path length from a
given pixel to each antenna of the SAR Interferometer, can be related directly to the
altitude of the antenna above the ground on a pixel-by-pixel basis to generate a DEM of
the surface.
Phase difference due to topographic relief is measured in modulo of 2. The
correct integer number of phase cycles must be added to each phase measurement to
obtain absolute height at each pixel. This is done by performing appropriate phase
unwrapping algorithm.

The vertical precision of the DEM generated by SAR Interferometry is

approximately 1/20th - 1/40th of the Altitude of Ambiguity. It depends on the phase
measurement uncertainty as described below:

h =

h
Sin
= a
2B
2

Where, is phase measurement uncertainty which in case of ERS data is 10 -20 .

It is clear from the above equation that normal spatial baseline of the interferometric data
pair plays an important role in the vertical precision of the DEM. So, baseline of the data
pair should be large enough to obtain significant vertical precision as long as there is not
much decorrelation due to larger spatial baseline. The geometric criterion for
interferogram restricts the maximum spatial baseline of two images to typically less than
1 km for ERS satellites. In practical terms, it is observed that the spatial baseline between
two images should be less than 300m for good geometric correlation.
15

In case of repeat -pass SAR Interferometry, temporal correlation of the

interferometric data pair is another important parameter to consider in obtaining a useful
interferogram from which topographic elevation can be extracted. To avoid temporal
decorrelation, simultaneous-baseline or single-pass Interferometry has been attempted
with an 11-day space mission called Shuttle Radar Topographic Mission (SRTM)
principally by National Imagery and Mapping Agency (NIMA) and the National
Aeronautics and Space Administration (NASA) during February 11-22, 2000. In this
space mission, single -pass interferometric data with 60 m. spatial baseline has been
acquired for ~80% of the land surface from 60 N to 58 S latitude.
Airborne systems can provide the highest resolution DEMs, but spaceborne SAR
provides greater data coverage, especially in some remote regions. The ERS-1 & 2 SAR
has a steep (~ 23 degree) incidence angle that causes layover and prevents DEM
extraction in high-relief areas, but in flatter areas it can produce DEMs with grid ~ 20 m
and noise levels roughly 5-10 m of relative height errors (depending on normal baseline).
The SIR-C SAR with higher bandwidth and adjustable incidence angle (up to 52 degrees)
can produce higher resolution DEMs (~ 10 m grid) in both low- and high-relief areas.
Noise levels for one-day repeat-pass SIR-C DEMs correspond to 2-3 m of relative height
error for C-band (wavelength 5.6 cm) and 8-12 m for L-band (wavelength 24 cm) in an
area with greater than 2000 m of relief. A comparative evaluation of DEM accuracy
obtained from available InSAR data is tabulated below (Table 1).

16

Table 1: Comparative evaluation of DEM accuracy obtained from the available InSAR
data.
Satellite/
Spaceshuttle
& Sensor
ERS 1 & 2
JERS-1
Radarsat-1
(Standard)
Radarsat-1
(Fine)
SIR-C

Horizontal
Resolution
(m)
20
30
30

SRTM
Airborne

Horizontal Relative Vertical

Accuracy
Accuracy (m)
(m)
20-40
5-10
30-60
10-15
30-60
10-30

Absolute Vertical
Accuracy m)
5-20
10-20
30-50

10

10-20

1-5

5-20

10

10-20

30 (C-band)
10m (X-band)
5

20

2-3 (C-band)
8-12 (L-band)
> 10
>6
1-3

5-10 (?) (C-band)

10-20 (?) (L-band)
16

4-8

2-6

The fringe density in the interferogram for the same terrain depends on a number
of factors namely, altitude of ambiguity (i.e., normal baseline of the data pair for the
same satellite mission and beam mode) and radar wavelength. With increasing normal
baseline and decreasing radar wavelength, fringe density in the interferogram increases
proportionally (Figure 6).

Figure 6a: Digital Elevation Model of Mt. Fuji, Japan derived from JERS-1 data.
(Courtesy: NASDA/EORC)

17

Figure 6b & 6c: SIR-C L- and C-band interferograms respectively of Ft. Irwin, California.
(Courtesy: NASDA/EORC)
The precise DEM generated from SAR Interferometry can be utilized for diverse
geoscientific applications viz., geomorphic mapping, mapping tectonic and structural
features expressed through relief variation, mapping resistant rock units and intrusive
bodies, and monitoring environmental impacts.
Geomorphological mapping can be easier when radar images are used with
precise digital elevation model, obtained from interferometric data pair. Radar images
being

side-looking

and

therefore

sensitive

to

terrain

geometry,

highlight

geomorphological features and dissection of the terrain better than that in optical images.
In the following fused L- and C-band radar FCC overlain on precise DEM derived
from SIR-C C-band data describes the following geomorphological features: (1)The
dissected plateau in the eastern part, (2) The dissected valley slopes in the central part
highlighting two litho units in the eastern valley slope identified from the break-in-slope.
18

The upper part is showin g medium-to-high slope suggesting some poorly -cemented hard
materials and the lower part is showing subdued character representing less resistant
thickly-bedded sedimentary rocks.

Figure 7: Perspective view of Karakax Valley, Northern Tibet derived from

SIR-C Interferometric DEM and SIR-C imagery. (L-band : Red,
L+C: Green, and C-band: Blue).
In the following colour-coded and hill-shaded topographic representation of precise
DEM derived from SRTM data of Haro and Kas Hills, India, geological and
geomorphological features can be easily interpreted. Geomorphologically, the area shows
a E-W trending curvi-linear strike ridge and a spindle -shaped denudo-structural h ill in the
north-western part. Sedimentary beds and flat-iron features are clearly discernible. From
flat-iron features, the spindle -shaped denudo-structural hill can be interpreted as the
remnant part of a double -plunging fold interspersed by a number of ~N-S transverse
faults causing lateral displacements of the rock formation. The wavy nature of the strike ridge made of resistant rock and doubly -plunging nature of the folds suggest 2-phases of

19

deformation in the terrain: an E-W upright folding followed by an N-S upright folding.
Small-scale folds in the strike -ridge strengthens this proposition.

Figure 8: SRTM Color-coded and Hill-shaded Topography of Haro and Kas

Hills, India
Using the coherence information derived from the complex correlation of two coregistered SAR images, attempts have been made by several groups of workers for
landuse/landcover classification and change detection studies (Borgeaud and Wegmller,
1996; Wegmller and Werner, 1996). The synergetic approach, using amplitude and
coherence information together, has also been found to have tremendous potential in
landuse-landcover classification (Chatterjee et al., 2004), in deciphering geological
information and in coastal mapping (Rudant et al., 1999; Trebossen, 1999; Chatterjee et
al., 2004).

20

20km

0
Scale

LEGEND
Water and Flat Area
(e.g., Airport Runway)

Grassland or Open Field

Dense Vegetation or Forest

Moderate to Sparse Vegetation

Moderate to Sparse Settlement

City Centre and Dense Settlement

Figure 9: Unsupervised classification of landuse/landcover types in Paris City, France

and surrounding areas from the fused product of ERS SAR amplitude and
coherence of a tandem pair and topography of the terrain.

The precise DEM generated from SAR interfe rometric data pair along with
amplitude and coherence information of the interferometric data pair can be used for
monitoring environmental hazards.
21

Using interferometric software SARDA, developed by Space Application

Centre, Ahmedabad, India, high res olution DEM and coherence image have been
generated from ERS SAR tandem pair of Jharia Coalfield, Jharkhand, India with spatial
baseline ~120 m. and temporal baseline 1-day. The DEM has a vertical accuracy better
than 5m while compared statistically w.r.t. map-based ground control points (GCPs).
From the DEM, opencast mining blocks (including abandoned, closed and non operational categories) having lower elevation and mining -related dumps having higher
elevation compared to the surrounding areas could be delineated (Chatterjee et al., 2003).
The abandoned/closed/non-operational opencast mines could not be differentiated from
the existing opencast mining areas in optical remote sensing data which could be
accomplished from the coherence information of the tandem pair (Chatterjee et al., 2003).

8615E

8625E

2350N

2340N

Figure 10: Interferogram generated from ERS SAR tandem pair of Jharia
Coalfield, Jharkhand, India

22

Figure 11: Co lour-coded DEM of Jharia Coalfield, Jharkhand, India derived from ERS
SAR tandem pair

D-InSAR technique has proved its potential in various surface deformation studies
carried out by the previous workers.
Earthquake/crustal deformation studies
It can provide high-resolution coseismic and post-seismic deformation maps as
demonstrated by Massonnet et al (1993, 1994, 1994), Zebker et al (1994), Massonnet and
Feigl (1995), Peltzer and Rosen (1995) in case of Landers, Northridge and Eureka Valley
earthquakes, California. Various studies hav e also been carried out for estimating crustal
deformation during a period of observation (Vadon and Sigmundsson, 1997; Yonezawa
and Takeuchi, 1998).

23

Figure 12: Differential interferogram showing co -seismic displacements of Landers

Earthquake (June 28, 1992; Mw =7.3), California, USA and the locii of the
rupture planes. The earthquake produced a surface break of ~70 km long with
up to 6.2 m of right-lateral offset.

Volcano monitoring studies

Volcanoes usually experience significant pre -eruptive surface deformation or bulging.
The scale of this surface deformation may range from cm to mm over tens of square
kilometers and over periods of weeks to years. Precise monitoring of surface deformation
24

may lead to accurate predictions of volcanic eruption. It seems, till date, no study on
prediction of volcanic eruption by D-InSAR techniques has been attempted. However,
various studies carried out by the earlier workers have successfully estimated the
deflation associated with Mount Etna volcanic eruption (Massonnet et al., 1995; Briole et
al., 1997; Lanari et al., 1998; Williams and Wadge, 1998) or Long Valley Caldera
(Thacher and Massonnet, 1997) and the subsidence associated with Krafla volcano
(Sigmundsson et al., 1997).

Figure 13a: DEM of Mount Etna, Sicily, generated from ERS -1 and ERS-2 interferometric image
pairs

Figure 13b: Differential SAR interferograms showing deformation fringes at different

stages of recent eruption and post-eruptive subsidence fringes during
May, 1992 October, 1993.

25

Landslide monitoring studies

D-InSAR technique has the capacity of providing continuous displacement fields of a
landslide. It allows determination of displacement vectors in different parts of the
landslide and thereby to delineate different units of the landslide based on varying rates
of displacement. The Saint-Etienne-de-Tinee landslide in the southern French Alps was
studied by different groups of workers (Fruneau et al., 1996; Rocca et al., 1997) and
successfully establish the displacement models for different parts of the landslide.
It is observed in the differential interferograms (shown above) that the number of

Figure 14: Geocoded differential SAR interferograms showing line -of-sight

displacement fringes of Saint -Etienne-de-Tine landslide, South France.

(a) t=3 days, B = 43 m. (b) t = 6 days, B = -298 m. (c) t = 9 days, B = -4 m. (d) t

= 9 days, B = 248 m. (e) t =11days, B = 291 m . (f) t=14days, B = -301 m.

fringes increases linearly with the elapsed time between the various image acquisitions,
while their overall geometry remains the same. This suggests that the observed landslide
motion is stationary over the period surveyed. On all six interferograms, NW -SE trending
fringes demonstrate downhill movement characterized by a gradient of displacement

26

from the top to the bottom of the landslide, the motion decreasing towards the bottom.
The fringe intervals are not constant over the landslide, suggesting both downhill and
lateral variations of the displacement gradient.

Glacier dynamics/Ice sheet motion studies

Glacier dynamics is one of the most important observable factors in global climate
studies. Using D-InSAR technique, a number of studies have been carried out on
Antarctic (Goldstein et al., 1993) and Greenland (Dammert and Hagberg, 1994; Joughin
et al., 1995; Rignot et al., 1995). Kwok and Fahnestock (1996) have presented an

Figure 15: Radar interferogram of an area that includes a portion of the Rutford
Ice Stream, Antarctica.
elaborated account on SAR interferometry techniques for studying ice sheet motion. The
parameters that glaciologists are interested in and that SAR is capable of providing are

27

snow facies, seasonal melt, icebergs, surface morphology, ice velocity, and surface
topography.

Land subsidence studies

Land subsidence may occur due to various anthropogenic factors such as underground
mining of minerals, excessive extraction of ground water or construction of underground
queries and dewatering of aquifers for establishing underground transport facilities
(Metro Railways).
Various studies have been attempted for detection and mapping of urban subsidence
and for identification of the possible controls (Perski, 1998; Strozzi and Wegmller,
1999; Amelung et al, 1999). In areas of slow deformation and under standard
atmospheric condition (particularly, where the atmosphere is heterogeneous), it is a
challenging task to differentiate the areas of ground subsidence from atmospheric
artifacts.

Figure 16: Differential SAR interferogram showing patches of ground subsidence

caused by underground coal mining over 35 -days period during
04.10.92-08.11.92 in Upper Silesia, Poland.

28

Figure 17a: IHS colour composition of ERS SAR amplitude (I), D-InSAR
interferometric phase (H) and correlation rate between the
interferograms showing subsidence fringes in Paris City, France during
06.10.93 - 21.07.95 during Metro Rail installation (RER -E).

2 km
Scale

Figure 17b: IHS colour composition of IGN topographic map (I), D-InSAR
interferometric phase (H) and correlation rate between the
interferograms showing subsidence fringes in Paris City, France during
19.05.93 - 07.04.95 during Metro Rail installation (RER -E).
29

BIBLIOGRAPHY & REFERENCES

Amelung, F., Galloway, D.L., Bell, J.W., Zebker, H.A. and Laczniak, R.J., 1999. Sensing
the ups and downs of Las Vegas: InSAR reveals structural control of land subsidence
and aquifer-system deformation, Geology, 27(6), 483-486.
Borgeaud, M. and Wegmller, U., 1996. On the use of ERS SAR interferometry for the
retrieval of geo- and bio-physical information, Fringe 96, Zrich, Switzerland,
available at http://www.geo.unizh.ch/rsl/fringe96/papers/borgeaud -wegmuller/
Briole, P., Massonnet, D. and Delacourt, C., 1997. Post-eruptive deformation associated
with the 1986-1987 and 1989 lava flows of Etna detected by radar interferometry,
Geophys. Res. Lett., 24, 37-40.
Chatterjee, R.S., Sarkar, T., Roy, P.S., Rudant, J.P., Frison, P.L., Fruneau, B., Trebossen,
H., Mitra, D., Gaurav, P. and Banerjee, S., 2004. Potential of ERS and Envisat SAR
data for environmental monitoring of Kolkata (Calcutta) city and coastal region of
West Bengal, India, Presented in ENVISAT Symposium (organized by European
Space Agency), Salzburg, Austria, September 06-10, 2004.
Chatterjee, R.S., Mathur, P., Jha, V.K. and Roy, P.S., 2003. SAR interferometric phase
and coherence information as supplements to high resolution optical remote sensor
data for environmental impact assessment in coalmining areas, Presented in the
Symposium titled Advances in Microwave Remote Sensing Applications, CSRE, IIT
Bombay, January 21-23, 2003.
Dammert, P.B.G. and Hagberg, J.O., 1994. SAR interferometry for measuring fast ice
displacement and ice ridge height, in Baltic Experiment for ERS-1 (BEERS), edited by
L. Ulander, Res. Rep., 51, 123-136, Winter Navig. Res. Board, Gteborg, Sweden.
Fruneau, B., Achache, J. and Delacourt, C., 1996, Observation and modelling of the
Saint-Etienne-de-Tine landslide using SAR interferometry, Tectonophyiscs, 265,
p. 181-190.
Gabriel, A. K. and Goldstein, R. M., 1988. Crossed orbit interferometry: theory and
experimental results from SIR-B, International Journal of Remote Sensing, 9(5),
pp.857-872.
Ghiglia, D. C. and Pritt, M. D., 1998. Two dimensional phase unwrapping theory,
algorithms and software, John Wiley and sons, New York.
30

Goldstein, R.M., Engelhardt, H., Kamb, B. and Frolich, R.M., 1998. Satellite radar
interferometry for monitoring ice sheet motion: Application to an Antarctic ice stream,
Science, 262, 1525-1530.
Graham, L. C., 1974. Synthetic interferometer radar for topographic mapping,
Proceedings of IEEE, 62(6), pp.763-768.
Joughin, I.R., Winebrenner, D.P. and Fahsestock, M.A., 1995. Observations of ice-sheet
motion in Greenland using satellite radar interferometry, Geophys. Res. Lett., 22,
571-574.
Kwok, R. and Fahnestock, M.A., 1996. Ice sheet motion and topography from radar
interferometry, IEEE Trans. Geosci. Remote Sens., 34, 189-200.
Lanari, R., Lundgren, P. and Sansosti, E., 1998. Dynamic deformation of Etna volcano
observed by satellite radar interferometry, Geophys. Res. Lett., 25, 1541-1544.
Li, F. K. and Goldstein, R. M., 1990. Studies of multibaseline spaceborne interferometric
synthetic aperture radars, IEEE Transactions on Geosciences and Remote Sensing,
28(1), pp.88-98.
Lin, Q., Vesecky, J. F. and Zebker, H. A., 1992. New approaches in interferometric SAR
data processing, IEEE Transactions on Geosciences and Remote Sensing, 30(3),
pp.560-567.
Madsen, S. N., 1995. On absolute phase determination techniques in SAR interferometry,
SPIE Conference on radar sensor technology, SPIE-2487, April, pp.393-401.
Massonnet, D. and Feigl, K.L., 1995. Satellite radar interferometric map of the coseismic
deformation field of the M=6.1 Eureka Valley, California, earthquake of May 17,
1993, Geophys. Res. Lett., 22, 1541-1544.
Massonnet, D., 1999. Landers earthquakes using SAR interferometry, Lecture delivered
in International Seminar on Recent Advances in Differential SAR Interferometry and
Applications organized by CSRE, IIT Bombay.
Massonnet, D., Briole, P. and Arnaud, A., 1995. Deflation of Mount Etna monitored by
spaceborne radar interferometry, Nature, 375, 567 -570.
Massonnet, D., Feigi, K., Rossi, M. and Adragna, F., 1994. Radar interferometric
mapping of deformation in the year after the Landers earthquake, Nature, 369,
pp.227-230.

31

Massonnet, D., Feigl, K.L., Vadon, H. and Rossi, M., 1994. Coseismic deformation field
of the M=6.7 Northridge, California, earthquake of Januay 17, 1994, recorded by two
radar satellites using interferometry, Geophys. Res. Lett., 23, 969-972.
Massonnet, D., Rossi, M., Carmona, C., Adaragna, F., Peltzer, G., Feigl, K. and Rabaute,
T., 1993. The displacement field of the Landers earthquake mapped by radar
interferometry, Nature, 364, 138-142.
Peltzer, G. and Rosen, P., 1995. Surface displacement of the 17 May 1993 Eureka Valley,
California, earthquake observed by SAR interferometry, Science, 268, 1333-1336.
Perski, Z., 1998. Applicability of ERS-1 and ERS-2 InSAR for land subsidence
monitoring in the Silesian coal mining region, Poland, Intl. Arch. of Photogrammetry
and Rem. Sens., 32(7), 555-558.
Prati et al., 1994. Report on ERS-1 SAR interferometric techniques and applications,
ESA contract N.3-7439/92/GHE-I, Department of Electronics, Politechnic of Milano,
Italy.
Rao, K. S. and Rao, Y. S. (1999) Introduction to SAR interferometry and applications,
Lecture delivered in International Seminar on Recent Advances in Differential SAR
Interferometry and Applications organized by CSRE, IIT Bombay.
Rignot, E., Jezek, K.C. and Sohn, H.G., 1995. Ice flow dynamics of the Greenland ice
sheet from SAR interferometry, Geophys. Res. Lett. 22, 575-578.
Rocca, F., Prati, C. and Ferretti, A. (1997) An overview of ERS-SAR interferometry, 3rd
ERS Symposium, ESA, SP-414, 1, p.xxvii-xxxvi..
Rodgers, A. E. E. and Ingalls, R. P., 1969. Venus: Mapping the surface reflectivity by
radar interferometry, Science, 165, pp.797-799.
Sigmundsson, F., Vadon, H. and Massonnet, D., 1997. Readjustment of the Krafla
spreading center segment to crustal rifting measured by satellite radar interferometry,
Geophys. Res. Lett., 24, 1843-1846.
Strozzi, T. and Wegmller, U., 1999. Land subsidence in Mexico city mapped by ERS
differential SAR interferometry, Proc. IGARSS 99, Hamburg, Germany, June 28
July, 1999.
Thacher, W. and Massonnet, D., 1997. Crustal deformation at Long Valley caldera,
eastern California, 1992-1996 inferred from satellite radar interferometry, Geophys.
Res. Lett., 24, 2519-2522.
32

Vadon, H. and Sigmundsson, F., 1997. Crustal deformation at Mid -Atlantic Ridge, SW
Iceland, mapped by radar interferometry, Science, 275, 194-197.
Vuorela, A., 1999. SARDEM digital elevation models based on interferometric ERS
SAR images, Lecture delivered in International Seminar on Recent Advances in
Differential SAR Interferometry and Applications organized by CSRE, IIT Bombay.
Wegmller, U. and Werner, C.L., 1996. Land applications using ERS-1/2 tandem data,
Fringe 96, Zrich, Switzerland, available at http://www.geo.unizh.ch/rsl/fringe96/
papers/ wegmuller-werner/
Williams, C.A. and Wadge, G., 1998. The effect of topography on magma chamber
inflation models: Application to Mt. Etna and radar interferometry, Geophys. Res.
Lett., 25, 1549-1552.
Yonezawa, C. and Takeuchi, S., 1998. An attempt to detect crusta l deformation by ERS-1
SAR interferometry in Omeazaki area, Central Japan, 32 nd RESTEC News Letter,
Japan, March, 1998.
Zebker, H. A. and Goldstein, R. M.., 1986. Topographic mapping from interferometric
SAR observations, Journal of Geophysical Research, 9 1(B5), pp.4493-4999.
Zebker, H.A., Rosen, P.A., Goldstein, R.M., Gabriel, A. and Werner, C.L., 1994. On the
derivation of coseismic displacement fields using differential radar interferometry:
The Landers earthquake, J. Geophys. Res., 99, 19617 -19634.

33