Long and short term monitoring of ground deformation in Thessaly basin

using space-based SAR Interferometry

byFalah Atta Fakhri

Supervisors: Dr. Issaak Parcharidis

Dr. George Mighiros Dr. Efthimis Karymbalis

Tuesday, June 25, 2013

PhD Candidacy Oral ExaminationHarokopio University of Athens

Department of Geography

Agenda Chapter One: Introduction Chapter Two: Data acquisition and SAR

interferometric techniques and processing Chapter Three: Impact of groundwater on ground

deformation. Chapter Four: Impact of fault movement and

earthquakes on ground deformation Chapter Five: Impact of lithology types on ground

deformation Chapter Six: Impact of soil on ground deformation Chapter Seven: The conclusions derived from this

research study and Discussion .

Natural hazards comprise:

Hydro-meteorological hazards, which include floods and flash floods, droughts, wildfires, tropical cyclones and hurricanes, and severe storms. Geological hazards, which include tectonic movement, earthquakes, tsunamis, volcanoes and explosive crater lakes, landslides, mudflows, erosion, and siltation.

Human-induced hazards comprise:which include wars, groundwater and oil withdrawal, mining, and land degradation.

Together, all of these hazards contribute to serious environmental problems which in consequence affect and destroy the economic development of countries and finally, in turn, impact on all walks of life.

Chapter One: Introduction PREFACE

Study area

The study area is located in the eastern part of the northern Thessaly Plain in central of Greece.

Thessaly plain, indicating that the study area exists within the frames of ascending and descending radar image tracks

Legend

Pineios_River

Ascending_Frame_Track 143

Descending_Frame_Track 279

Prefectures of Thessaly

P. EYRYTANIAS

P. FUIVTIDAS

P. GREBENVN

P. IVANNINVN

P. KARDITSAS

P. KOZANHS

P. LARISAS

P. MAGNHSIAS

P. PIERIAS

P. TRIKALVN

Study area problems and constraints

The study area suffers from a ground deformation phenomenon which affects civil construction and agricultural activity.

Furthermore it is a very complicated area due to the distribution of human hazards, which include groundwater withdrawal, and due to the presence of natural hazards for instance active tectonics.

SAR Interferometry

This new geodetic technique calculates the interference pattern caused by the difference in phase between two images acquired by spaceborne SAR at two distinct times.

SAR Interferometric Techniques

Repeated pass Interferometry (Conventional InSAR)

In repeat-pass InSAR, two or more SAR images are acquired at different times with the same or a corresponding sensor from almost identical aspect angles.

Interferometric Stacking

The basic idea of interferogram stacking is to combine multiple observations into a single result.

Persistent Scatterers Interferometry (PSI)

This technique exploits temporal and spatial characteristics of interferometric signatures, collected from point targets.

Objectives of the Research Study

1- To evaluate the possibility of applying SAR interferometric techniques to monitor and map ground deformation in urban and agricultural lands over the long and short terms.

2- To investigate and identify the causes of ground deformation.

3- To evaluate the possibility of recognizing each individual cause of ground deformation by monitoring the time series behavior of ground deformation using the statistical results of SAR interferometric techniques.

4- To apply spatial and qualitative correlations between ground deformation and parameters (precipitation, groundwater, fault movement, earthquake, lithology, and soil) to reveal the reality of ground deformation within the study area.

Flowchart of study research methodology

SAR interferometry

Data managementDevelopment of geographic

information system

Results

Create statistical correlation

Create spatial correlation

Meteorology

Thematic maps

Hydrology and hydrogeology

Soil Geology

Additional data collection

Field- work

Data acquisition

Preprocessing

Production andinterpretation of deformation

maps

Repeated pass interferometry

Persistent scatterer

interferometry

Stacking interferometry

Chapter Two: Data acquisition and SAR interferometric techniques and processing

Data and Methodology

SAR Data Selection and Interferometric Processing (Ascending Track 143)

The total dataset consists of 24 Single Look Complex (SLC) SAR C-band images of ERS-1/2, during 1995–2000. Additionally, 15 SLC images of ENVISAT ASAR acquired during 2003–2008 by ESA, which cover the study area, have also been selected along this track.

Processing raw data to obtain SLC format

images

Add the parameters of orbits

Estimate perpendicular baselines of all the

selected images

Resample images by estimating initial range

and azimuth offsets

Corregister of SLC images

DEM , 90 (m) (SRTM)

crop image;

Multi-look

Geometrical correction

Flowchart of preprocessing ASR images

Simulation SLC SAR images with the

Multi-look average image ascending track highlighting the study area and the mountains around the basin and urban area

Results and discussion

Repeated pass interferometry processing

Master image Slave image B┴ (m) Interval Days

19960228 19960403 -66.80 35

Coherence map for time interval 19960228_19960403 ascending track

LARISA

Differential interferogram for time interval 19960228–19960403 ascending track

LARISA

Interferometric stacking processing

Average coherence for time interval 1995–2008 ascending track highlighting the coherence of the reference point inside the red circle

LARISA

Ground deformation rates along LOS direction deduced by interferometric stacking, for the considered time intervals (1995–2008) Ascending track and different acquisition. Background is an average of multi-look SAR intensities. The selected reference point is marked with a green Bp is 0-200 m, 29 interferograms

LARISA

Persistent (Permanent) Scatterers Interferometric (PSI)

Distribution of geo-coded radar targets (persistent scatterers) in Larissa basin before expansion. The average in line of sight (LOS) velocity for the period (1995–2006) number of points are 1866

LARISA

Distribution of geo-coded radar targets (persistent scatterers) in Larissa basin after expansion. The average in line of sight (LOS) velocity for the period (1995–2006) number of points are 62551

LARISA

SAR Data Selection and Interferometric Processing (Descending Track 279)

The total dataset consists of 48 SLC SAR C-band images of ERS-1/2 from 1992 –2000 and additionally, 25 SLC images of ENVISAT ASAR from 2002 – 2010 acquired by ESA,, which cover the study area have been selected along this track

Repeated pass interferometry processing

Master image Slave image B┴ (m) Days

19980802 19980906 - 1.51 35

Coherence map for time interval 19980802–19980906 descending track

LARISA

Differential interferogram for time interval 19980802–19980906 descending track

LARISA

Interferometric stacking processing

Coherence map for time interval 1992–2010 descending track highlighting

the coherence of reference point inside the red circle

LARISA

Ground deformation rates along LOS direction deduced by interferometric stacking, for the considered time intervals (1992-2010), Bp 0-150, 73 inteferograms

LARISA

Persistent (Permanent) Scatterers Interferometry (PSI)

Distribution of geo-coded radar targets (persistent scatterers) in Larissa basin before expansion. The average in line of sight (LOS) velocity for the period (1992–2010) number of points are 1930

LARISA

Distribution of geo-coded radar targets (persistent scatterers) in Larissa basin after expansion. The average in line of sight (LOS) velocity for the period (1992–2010) number of points are 4801

LARISA

Chapter Three: Impact of groundwater on ground deformation

Location of groundwater monitoring network within study area, superimposed on SLC SAR image

Three point candidates of the PSI with different distances from borehole AD6 ASCENDING TRACK 1995 - 2006

0

5

10

15

20

25

Jan-

92Ju

l-92

Jan-

93Ju

l-93

Jan-

94Ju

l-94

Jan-

95Ju

l-95

Jan-

96Ju

l-96

Jan-

97Ju

l-97

Jan-

98Ju

l-98

Jan-

99Ju

l-99

Jan-

00Ju

l-00

Jan-

01Ju

l-01

Jan-

02Ju

l-02

Jan-

03Ju

l-03

Jan-

04Ju

l-04

Jan-

05Ju

l-05

Jan-

06Ju

l-06

Jan-

07Ju

l-07

Jan-

08Ju

l-08

Jan-

09Ju

l-09

Jan-

10Ju

l-10

Time

Wat

er L

evel

[m]

AD6

0

10

20

30

40

50

60

70

80

90

100

Jun_

1995

Dec

_199

5

Apr

_199

6

Mar

_199

7

May

_199

7

Dec

_199

7

Aug

_199

8

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr

_200

3

Aug

_200

3

Feb

_200

4

Apr

_200

4

Aug

_200

4

Sep

_200

4

May

-200

5

Aug

_200

5

Dec

_200

6

Time

Mo

nth

ly p

reci

pit

atio

n [

mm

]

0

2

4

6

8

10

12

14

16

18

20

Wat

er t

able

[m

]

Monthly amount of Precipitation[mm] Water table [m] AD6

LOS Displacemnt of point candidates of PSI corresponding to monthly precipitation amount. Displacement time series of point candidates are rescaled to the first acquisition (i.e. 28 June 1995). ASCENDING TRACK 1995 - 2006

-120

-100

-80

-60

-40

-20

0

20

40

Jun_

1995

Dec

_199

5

Apr

_199

6

Mar

_199

7

May

_199

7

Dec

_199

7

Aug

_199

8

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr

_200

3

Aug

_200

3

Feb

_200

4

Apr

_200

4

Aug

_200

4

Sep

_200

4

May

-200

5

Aug

_200

5

Dec

_200

6

Time

LO

S d

isp

lace

men

t [m

m]

0

10

20

30

40

50

60

70

80

90

100

Mo

nth

ly a

mo

un

t o

f P

reci

pit

aio

n [

mm

]

Monthly amount of Precipitation[mm] Deformation_rate_mm(p68587)_90m

Deformation_rate_mm_p(68496)_179m Deformation_rate_mm_p(69756)_219m

LOS Displacemnt of point candidates corresponding to the groundwater level of borehole AD6. Displacement time series of point candidates are rescaled to the first acquisition (i.e. 28 June 1995). ASCENDING TRACK 1995 - 2006

-120

-100

-80

-60

-40

-20

0

20

40

Jun_

1995

Dec

_199

5

Apr

_199

6

Mar

_199

7

May

_199

7

Dec

_199

7

Aug

_199

8

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr

_200

3

Aug

_200

3

Feb_

2004

Apr

_200

4

Aug

_200

4

Sep

_200

4

May

-200

5

Aug

_200

5

Dec

_200

6

Time

LOS

dis

plac

emen

t [m

m]

0

2

4

6

8

10

12

14

16

18

20

wat

er ta

ble

[m]

p68587_90m p68496_179m p69756_219m water_table_AD6

Conventional SAR Interferometry Seasonal Deformation 19960228–19960403 ASCENDING TRACK

Borehole Groundwater level (m) Interferometric fringes

SR72 20.62 Significant

SR77 18.50 High Significant

Conventional SAR Interferometry Seasonal Deformation 19980802–19980906 DESCENDING TRACK 279

Borehole Groundwater level (m) Interferometric fringes

SR72 34.00 High Significant

SR77 36.43 High Significant

Impact and interference type of clay minerals with fluctuation of groundwater level on land deformation

RainfallRising groundwater

level

Groundwater withdrawal

Presence of clay minerals which have the capability to swell and shrink

Water enters between clay layers

Activation of swelling operation

Water leaves from among mineral layers

Activation of shrinking operation

Subsidence of the ground

Uplift of the ground

Compression of materials

Decline of groundwater level

Maybe are caused microseismic (3–4) magnitude?

Chapter Four: Impact of faults movement and earthquakes on ground eformation

In order to examine and investigate the correlation between fault movements and ground deformation by implementing three techniques, conventional SAR interferometric, interferometric stacking and persistent scatterers interferometry (PSI),

Fault traces which are distributed within the study area of the eastern part of northern Thessaly were digitized from the papers by (Caputo, 1993), (Caputo and Pavlides, 1993), (Caputo et al., 1994), (Caputo et al., 2004), (Caputo and Helly, 2005) and (Caputo et al., 2006).

Thereafter these were corrected and rectified depending on 7 geological maps of Thessaly at a scale of 1:50,000 issued by the Greek Institute of Geology and Mineral Exploration, which were used along with field observations.

In addition, by using a seismotectonic map of Greece with seismogeological data at a scale of 1:500,000, a shape file was consequently created and identified utilizing GIS software ArcGIS 9.3.

Earthquake events data within the study area were collected by utilizing the earthquake catalogue of the (Institute of Geodynamics), National Observatory of Athens,

An attribute table was then created from this catalogue.

Consequently, a shape file of earthquake events was created utilizing Arc GIS 9.3 for the period 1964 – 2010 with magnitude M >= 3 and depth varying between 0 – 30 km.

Distribution of faults and earthquakes within study area

Legend

earthquake_1964_2010_Depth_30km

MAGNITUDE

3

4

5

(( (( Normal Faults

Thessaly_Settlements

The interference effects of fault movement on ground deformation will be discussed and interpreted in a probability approach depending on spatial correlation, for the reason that no statistical correlation or model-building has been done between ground deformation and fault movement.

Ascending track 143

Interferometric stacking

Total deformation at Larissa estimated with interferometric stacking technique, June 1995-March 2008

Descending track 279

Interferometric stacking

Total deformation at Larissa estimated with interferometric stacking technique, November 1992 – October 2010

ASCENDING TRACK 1995 – 2006

Persistant Scatterers Interferometric (PSI)

Frequency of deformation rate of points targets at LARISA 1995-2006

Frequency of deformation rate of points targets at Tyrnavos 1995-2006

Frequency of deformation rate of points targets in Giannouli 1995-2006

Location of selected candidate points minimum and maximum deformation rate, ascending track 143, settlement of Larissa

-30

-25

-20

-15

-10

-5

0

5

10

Jun_

1995

Dec_

1995

Apr_

1996

Mar

_199

7

May

_199

7

Dec_

1997

Aug_

1998

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr_

2003

Aug_

2003

Feb_

2004

Apr_

2004

Aug_

2004

Sep_

2004

May

-200

5

Aug_

2005

Dec_

2006

Time

LOS

disp

lace

men

t [m

m]

-350

-300

-250

-200

-150

-100

-50

0

50

100

Jun_1995

Dec_1995

Apr_

1996

Mar_

1997

May_1997

Dec_1997

Aug_1998

Jan_1999

Jun_1999

Oct_

1999

May_2000

Apr_

2003

Aug_2003

Feb_2004

Apr_

2004

Aug_2004

Sep_2004

May-2

005

Aug_2005

Dec_2006

TimeL

OS

dis

pla

cem

en

t [m

m]

Conventional SAR Interferometry ASCENDIGN TRACK

SEASNAL DEFORMATION

19960228_19960403

Conventional interferogram corresponding to a 7 km cross-section of Larissa in the period 19960228_19960403

Spatial profile showing the displacement field as observed by conventional interferometry within a 7 km cross-section of Larissa, in the period 19960228_19960403 red lines correspond to the faults

-2

-1

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7 8

Distance Km

Dis

plac

emen

t mm

/LO

S

SWNE

Conventional SAR Interferometry DESCENDIG TRACK

SEASNAL DEFORMATION

19980802_19980906

Conventional interferogram corresponding to 7 km cross-section of Larissa in the period 19980802_19980906

Spatial profile showing the displacement field as observed by conventional interferometry within a 7 km cross-section of Larissa, in the period 19980802_19980906 red lines correspond to the faults

-35

-30

-25

-20

-15

-10

-5

0

0 1 2 3 4 5 6 7 8

Distance Km

Disp

lace

men

t mm

/LO

S

NE SW

Chapter Five: Impact of lithology types on ground deformation

The type of lithology has an important impact on ground deformation for the reason that any physical or chemical change of shape or size of materials will be reflected in the stability of objects.

Spatial correlation between lithology type and ground deformation has been created to verify the impact of lithology on ground deformation, taking into account the general type of lithology within the study area.

Thirty settlements were identified. However, just 19 were selected to examine and investigate the influence of lithology type on ground deformation.

The geological formations are grouped into three classes taking into account their lithology, consolidation, origin and age.

FIRST CLASS IS Alluvial = al

SECOND CLASS IS Fluvio-lacustrine deposits = Pt2

THIRD CLASS IS Terrestrial fluvio-torrential deposit = pl-pt

Geological map of Thessaly, map is modified from IGME. Faults are modified according of Caputo.

Legend

Alluvial deposits; Lake deposits

Talus cones

Marls, clays

Conglomerates, Neogene

Calcareous formations-Marbles

Gneiss, Gneiss-Schist, Schists

Granites

Ophiolites

Metamoprhics

Shale-Chert formation

Flysch

Phyllite-Quarzite

Olistholites

Water

(( (( Normal Faults

Thessaly_Cities

-6

-4

-2

0

2

4

6

8La

rissa

a

l

Gia

nnou

li al

Cha

lki a

l

Ele

fthe

ron

al

Fal

anna

al

Mel

isso

chor

ion

al Gal

ini a

l

Pla

tyka

mpo

sal Gla

fki a

l

Itea

al

Fyl

lon

al

Pal

amas

al

Mar

athe

a al

Nik

aia

al

Ter

psith

ea p

l-pt

Tyr

navo

s P

t2

Rod

ia P

t2

Man

dra

Pl-P

t

Ele

fthe

rai P

l-Pt

Type of Lithology

LO

S D

isp

lace

men

t [m

m/y

r]

Minimum Subsidence (mm) Maximum Subsidence (mm) Minimum Uplift (mm) Maximum Uplift (mm)

Ascending track 143 1995 - 2008 Interferometric Stacking

Minimum and maximum deformation rates in LOS of interferometric stacking 1995-2008 of 19 settlements corresponding to type of lithology.

Fluvio-lacustrine deposits = Pt2 Terrestrial fluvio-torrential deposit = pl-pt

Alluvial = al

-5

-4

-3

-2

-1

0

1

2

3

4

Laris

sa a

l

Gia

nnou

li al

Cha

lki a

l

Ele

fthe

ron

al

Fal

anna

al

Mel

isso

chor

ion

al

Gal

ini a

l

Pla

tyka

mpo

s al

Gla

fki a

l

Itea

al

Fyl

lon

al

Pal

amas

al

Mar

athe

a al

Nik

aia

al

Ter

psith

ea p

l-pt

Tyr

navo

s P

t2

Rod

ia P

t2

Man

dra

Pl-P

t

Ele

fthe

rai P

l-Pt

Type of Lithology

LO

S D

isp

lace

men

t [m

m/y

r]

Minimum Subsidence (mm) Maximum Subsidence (mm) Minimum uplift (mm) Maximum uplift (mm)

Descending track 279 1992- 2010 Interferometric Stacking

Minimum and maximum deformation rates in LOS of interferometric stacking 1992-2010 of nineteen settlements corresponding to type of lithology.

Terrestrial fluvio-torrential deposit = pl-pt

Alluvial = al

Fluvio-lacustrine deposits = Pt2

Chapter Six: Impact of soil on ground deformation

The goal of this chapter is to examine the potential of using the PSI technique to identify the deformation of soil vertically (i.e., line of sight, LOS), and to study the statistical behavior of deformation for each point target through the statistical time series schemes of the data set, as well as the effect of soil type on its deformation.

According to the Exploratory Soil Survey and soil classification system (Soil Survey Staff, 1998) and (Soil Survey Staff, 1999), the classification of soil units of the study area (north part of Larissa) has been completed and 5 different orders were recognized (Alfisols, Entisols, Inceptisols, Mollisols, Vertisols)

The soil properties of each order have been examined, such as texture, drainage, erosion and slope. Soil data has been manipulated using Arc GIS 9.3 software, and several maps, such as soil texture, soil drainage, slope, and erosion have been created

Map of exploratory soil survey depicts the distribution of soil orders within the study area in the northern part of Larissa. Based on SLC of SAR image.

Legend

soil_orders

Alfisols

Entisols

Inceptisols

Molisols

Vertisols

(( (( Normal Faults

Ascending track 143

PSI candidate points within the non-urban area, mean displacement rates 1995-2006, ascending track 143. Movements are in the satellite line-of-sight direction. Based on SLC of SAR image.

Legend

soil_orders

Alfisols

Entisols

Inceptisols

Molisols

Vertisols

(( (( ((Normal Faults

-70

-60

-50

-40

-30

-20

-10

0

10

20

Jun_

1995

Dec

1995

Apr

_199

6

Mar

1997

May

1997

Dec

1997

Aug

_199

8

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr

_200

3

Aug

_200

3

Feb_

2004

Apr

_200

4

Aug

_200

4

Sep

_200

4

May

-200

5

Aug

_200

5

Dec

_200

6

Time

LOS

dis

plac

emen

t [m

m]

-70

-60

-50

-40

-30

-20

-10

0

10

20

Jun_

1995

Dec

1995

Apr

_199

6

Mar

1997

May

1997

Dec

1997

Aug

_199

8

Jan_

1999

Jun_

1999

Oct

_199

9

May

_200

0

Apr

_200

3

Aug

_200

3

Feb_

2004

Apr

_200

4

Aug

_200

4

Sep

_200

4

May

-200

5

Aug

_200

5

Dec

_200

6

Time

LOS

dis

plac

emen

t [m

m]

0

10

20

30

40

50

60

70

80

90

100

Mon

thly

am

ount

of P

reci

pita

ion

[mm

]

Monthly amount of Precipitation[mm] displacement_mm

Chapter Seven: The conclusions derived from this research study and Discussion

The data of SAR images ERS1/2 and ASAR ENVISAT which have been used in this research study are shown the possibility for investigating and identifying the temporal and spatial ground vertical movement within study areas of Larisa basin. However, the cons of these types of data were the spatial resolution which is 20 meters, consequently this spatial resolution does not was large enough to detect the ground deformation for objects which are located within large scale. However, the temporal resolution was applicable good enough to the objectives of this study.

The SAR interferometric conventional technique, has pros to investigate the ground deformation during short-term within urban and non-urban area. However, the cons of this technique are the deformation is limited by the atmospheric path delay term.

The SAR interferometric stacking technique has the advantages to bypass the cons of the atmospheric path delay, however no time series could be obtained for each single object by this technique.

The persistent scatterers technique has the advantages to obtain the ground deformation for each single object for long-term time series; however the disadvantage of this technique is the hard conditions ought to apply to get the candidates points specially within agricultural fields.

Approximately all correlation cases between fluctuation of groundwater level and land deformation point to non-continuous significant correlation through the short and long distances between boreholes and point candidates of PSI within ascending and descending tracks. This may be a reflection of the spatial complexity of aquifer systems, the variety of subsidence and uplift deformation, and the large number of illegal wells.

Significant interferometric fringes are observed within approximately all of the boreholes in two differential interferograms of two tracks, ascending and descending, through the fluctuation of groundwater level.

The Persistent Scatterers Technique, through the application of spatial correlation between the locations of points targets and fault traces, reveals or/and indicates the possibility of the influence of fault movements on ground deformation.

In spite of the controversy regarding the gap of the last large magnitude earthquake in Larissa (1941), which remains a major issue, nevertheless, fault movements, which are the main reason of earthquakes creation, may be attributed to the impact of mutual processes between the swelling and shrinkage of clay minerals.

SAR interferometry techniques successfully revealed the impact of lithology type on ground deformation through the ascending and descending tracks.

Subsidence could not be attributed to the sole impact of the type of lithology. This was because there are several nested and interconnected factors such as lithology, fault movements, type of clay minerals and amount of precipitation.

Papers have been published from the dissertation

Fakhri F, “Use Sar Interferometry DInSAR And PSI To Identify The Geohazard Risk Of Nato Airport North East Larissa (Central Greece) ” International Forum On Satellite Earth Observation For Geohazard Risk Management Santorini, 21-23 May 2012.

Fakhri F, Parcharidis I, Karymbalis E, Pavlopoulos K, Relationship Between Lithology And Ground Deformation Estimated Using SAR Interferometry Over The Settlements Of Thessaly Prefecture (Central Greece). 4th Workshop On Remote Sensing And Geology Earsel Mykonos, Greece, 24 – 25 May 2012.

Fakhri F, Psomiadis Emmanouil, Parcharidis Issaak1 Monitoring Soil Deformation Using Persistent Scatters Interferometry (PSI) Technique: The Case Study Of Thessaly Prefecture (Central Greece). 978-1-4673-1159-5/12/ Ieee 3903 Igarss 2012

FAKHRI F, Investigating Of causes Short and Long Term Ground Deformation by Implementing SAR Interferometric Techniques in Larissa. ESA Living Planet Symposium 9 - 13 September, 2013 Edinburgh, United Kingdom