A GIS-BASED STUDY OF HANOI LAND SUBSIDENCE, INCLUDING
A REVIEW OF THE MONITORING TECHNIQUE INSAR

by

Nguyen Thi Tam

A thesis submitted in partial fulfillment of the requirements for the
Degree of Master of Science
in Geotechnical and Geoenvironmental Engineering

Examination Committee:

Nationality:
Previous Degree:

Dr. Pham Huy Giao (Chairman)
Prof. Freek van Der Meer (Co-chairman)
Dr. Noppadol Phien-wej

Vietnamese
Bachelor of Science in Geological Science
Hanoi University of Science
National University Hanoi, Vietnam

Scholarship Donor: The BRIDGE Project

Asian Institute of Technology
School of Engineering and Technology
Thailand
May 2008

i


Acknowledgments
I am truly grateful to many individuals who helped me make my research work a
challenging, enjoyable and rewarding experience.
Firstly, I would like to express my deepest gratitude to my advisor, Dr. Pham Huy Giao who
exceptionally shared his valuable time, excellent support and professional expertise with
understanding and patience in the course of my research.
I would like to express my profound gratitude, utmost sincere appreciation and thanks to
my co-advisor, Prof. Freek van der Meer for his guidance, valuable suggestions and
encouragement in the whole research period especially for his guidance and suggestions
related to chapter 4 of my thesis work.
Special gratitude and utmost thanks is extended to my examination committee members,
Dr. Noppadol Phien-weij for his valuable suggestions and encouragement given to me
from the proposal period that helps direct this thesis in a proper way.
My sincere gratitude and thanks goes to Dr. Mark van der Meijde for his kind supports
related to InSAR technique.
For accomplishing this thesis, I would like to express my thanks to Dr. Jirathana
Worawattanamateekul, Ms. Tran Van Anh, Dr. Nguyen Dinh Duong, Mr. Tran Anh Tuan,
Mr. Tran Trong Thang for giving me the valuable data related to InSAR method and GISbased study.
I also would like to take this opportunity to express my gratitude and utmost thanks to
Prof. Mai Trong Nhuan who always support and encourage me from my undergraduate
study to graduate study period. This gives me the confidence to cope with any new
challenge.
My gratitude and thanks is extended to Prof. Dang Van Luyen for his kind supports and
encouragement during my study in AIT and ITC.
My sincere thank is extended to all the staffs of Earth Resource Exploration stream specific
in ITC for the valuable lectures and sincere helps when I was studying there. I would like

to extend my thanks to all faculty members, office staffs and friends of Geotechnical and
Geoenvironmental Engineering Program for their friendships and continuous
encouragements.
I also gratefully acknowledge the kind assistance given by Hanoi Institute of Building
Technology, Hydrogeological Division No 2. Special thanks go to Dr. Tran Thi Ha, Mr.
Dang Tran Trung, and Dr. Nguyen Van Dan for providing an amount of useful data used
in this thesis.
I would like to express my deeply gratitude to the Asian Institute of Technology (AIT),
International Institute for Geo-Information Science and Earth Observation (ITC) and in
particular the BRIDGE project for offering me a full scholarship to study in the Geosystem
Exploration
and
Petroleum
Geoengineering
program,
Geotechnical
and
Geoenvironmental Engineering Field of Study, AIT with a pleasant semester of exchange
study at ITC. Without the support of BRIDGE project, I would not have chance to
accomplish my MSc study at both well known institutes AIT and ITC.
Last but not least, my deepest gratitude is for my grandparents, parents, brother, Nguyen
Tu Trong, his wife, and especially my two younger sisters Nguyen Thi Dung and Nguyen
Thi Hang for their love, sacrifices and continued encouragement during my study.

All this work is dedicated to you, my two younger sisters,

ii


Abstract

Hanoi is located in the Red River delta, about 100km from the East Sea. The city’s
population is 3 million and most of water supply comes from groundwater (Giao, 1999).
Groundwater exploitation in Hanoi by both public and private wells may have reached
750,000m3/d and the groundwater level has lowered in many locations, say to 35m at Ha
Dinh station (Ha, 2007). Over-abstraction of groundwater has caused land subsidence,
which in turn affects the upper infrastructures of the city. As the network of monitoring
station in Hanoi is still limited, a new tool that can monitor land subsidence over a larger
area is needed to be studied.
This study has reviewed on a technique known as Satellite-based Interferometric Synthetic
Aperture Radar (InSAR) for monitoring land subsidence. Differential InSAR (DInSAR) is
a conventional InSAR method, which removes the phase signal caused by relief to yield a
differential interferogram in which the signature of surface deformation (in turn land
settlement) can be detected. Raucoules and Carnec (1999) had generated differential
interferograms from ERS-1/2 SAR data and detected signatures of ground vertical
movement as much as 15 mm within the time period of 2,5 years; Van Anh (2005) had
used JERS-1 SAR data in the period of 1995-1998 and detected the settlement rate in the
range of 27-33mm/y. The comparison of DInSAR results with the land subsidence records
by leveling shows that the settlement from DInSAR was higher than the leveling
measurements and the spatial pattern distribution of high settlement rate between them was
quite different.
In addition, this study has proposed a GIS-based procedure to create a land subsidence
susceptibility map for Hanoi, which is resulted from overlaying of 7 layers, i.e., surface
settlement, Quaternary formation thickness, groundwater drawdown, groundwater dynamic
regime, soft clay distribution, population density, and distribution of industrial areas. A
weighted scale from 1 to 5 was proposed and assigned to each layer. The suitability map is
classified into four susceptibility zones of occurring land subsidence: the very high
susceptibility zone is in the centre of southern Hanoi area; the high susceptibility zone is in
the southwestern Hanoi area, the medium susceptibility zone is in the middle of southnorth Hanoi area; and the low susceptibility zone is in the northern Hanoi area.

iii



Table of Contents
Chapter

Title

Page

Title Page
Acknowledgements
Abstract
Table of Contents
List of Tables
List of Figures
List of Illustrations

i
ii
iii
iv
v
vi
viii

1

Introduction
1.1 Problem statement
1.2 Objectives of the study

1.3 Scopes of the study

2

Literature Review
2.1 Backgrounds on land subsidence
2.2 Study area
2.3 Basics of Geographical Information System (GIS)
2.4 Basics of Remote Sensing
2.5 From SAR data to Information: Status, Trends and Future
(Rudiger Gens, 2006)
2.6 Important Parameters of SAR Satellite System

3

4

5

6

7
8

1
1
2
2

Methodology

3.1 Flowchart carried out the thesis work
3.2 A study of InSAR technique in monitoring land subsidence
3.3 Land subsidence susceptibility zonation
InSAR technique and its application for Hanoi land subsidence
4.1 Introduction
4.2 Interferometry Synthetic Aperture Radar (InSAR)
4.3 Detection of Hanoi land subsidence by DInSAR method
4.4 Groundwater level lowering and subsidence appearance in Hanoi
area
4.5 Comparisons and Discussions
4.6 Limitations of DInSAR method, a step to PSInSAR technique
Land subsidence susceptibility zonation
5.1 Introduction
5.2 The factors control Hanoi land subsidence
5.3 Weighted overlay the land subsidence control factors
5.4 Concluded remarks
Conclusions and Recommendations
6.1 Conclusions
6.2 Recommendations
References
Appendixes

iv

3
3
8
13
15
16

17
19
19
19
22
26
26
26
28
30
30
32
35
35
35
37
38
39
39
39
41
46


Table
2.1
4.1

4.2a
4.2b


List of Tables
Title
Comparison of different means for measuring subsidence
(Worawattanamateekul, 2006)
Used interferometric SAR combinations from ERS1/ERS2 data on Hanoi and
perpendicular components of the baselines in meters (Raucoles and Carnec,
1999)
Descriptions of five images of JERS-1 (Anh, 2005)

5.1

Characteristics of the produced (SLC) and (MLI) data from four data sets
(Anh, 2005)
Two sets of leveling survey data (Anh, 2005)
Comparison of leveling data with DInSAR data. The annual rate of
subsidence estimated by DInSAR analyses were sampled at ten benchmarks
(Updated from Anh, 2005 and Toan, 2005)
Settlement and drawdown data at land subsidence monitoring station (HIBT,
2004)
The thickness of soft clay at some area inside Hanoi

5.2
5.3

Assigned weight of soft clay distribution factor
Assigned weight for population density factor

4.3
4.4


4.5

v

Page
52
53

53
54
55
56

57
58
58
58


Figure
1.1
2.1

2.2
2.3
2.4a
2.4b
2.5
2.6a

2.6b
2.6c
2.6d
2.7
2.8
2.9
2.10
2.11a
2.11b
2.11c
2.11d
4.1a

4.1b

4.1c

4.2
4.3

4.4

4.5

List of Figures
Title
(a) Vietnam; (b) Study area, Hanoi, the capital of Vietnam
Total stress, effective stress, and fluid pressure (pore water pressure) on
an arbitrary plane through a saturated porous medium (after Freeze and
Cherry, 1979)

General schematic of a pipe extensometer for site-specific measurement
of subsidence
Above-ground portion of a cable extensometer with recording equipment
Lithofacies section of the Red River Delta (W-E) (Nghi et al, 1991)
Lithofacies section of the Red River Delta (S-N) (Nghi et al, 1991)
Hanoi subsoil profiles (Hien, 2006
Ha Noi aquifer system (Giao and Ovaskainen, 2000)
Hydrogeological map of Hanoi area (K2, 2003)

Page
59
59

60
60
61
61
62
62
63

Sketch of Hanoi aquifer system (HWBC, 2000)
National groundwater monitoring stations and Hanoi groundwater
monitoring stations
Locations of Well Fields and Land Subsidence Monitoring Stations in
Hanoi (Toan, 2005)

64
65


Well fields in Hanoi city: (a) well field; (b) Groundwater exploitation 1
(Anh, 2005)
A conceptual GIS Model showing thematic layers
Electromagnetic energy has been classified by wavelength and arranged
to form the electromagnetic spectrum
SAR system from a satellite
InSAR imaging geometry
Geometric parameters of a satellite interferometric SAR system
Geometry of a satellite InSAR system
Geocoded unwrapped interferogram resulting from the fusion of the
interferograms 23925/4753, 4252/4753,24426/4753, and 4252/24426
(Raucoules and Carnec, 1999)
Interferograms 24426/7258 and 4753/7258 with possible movement
related features. The phase variation is about half fringe i.e 15 mm
movement (during a 6 months period) (Raucoules and Carnec, 1999)
Interferograms 24426/15274 and 7258/15274 (2 years apart). A possible
vertical movement (less than half fringe) is located (Raucoules and
Carnec, 1999)
Interferograms and coherence maps generated from Image 4
(1995/09/18), Image 3 (1995/08/05) and Image 5 (1998/09/22).
Differential interferogram Δφ. One fringe corresponds to the change of
distance along the line of sight by one half of wavelength (11.76 cm)
(Anh, 2005)
Estimation of total subsidence during the period 1995-1998 based on
the assumption that there has been no horizontal displacement (Anh,
2005)
Subsidence along the profiles X-X’ and Y-Y’ given in Figure 4.4 (a)
Profile X-X’ across the area A, (b) Profile Y-Y’ across the areas B and C.

67


vi

66

68
68
68
69
69
70
71

72

72

73
74

75

76


4.6
4.7
4.8
4.9
4.10a

4.10b
4.11
4.12a
4.12b
4.13
5.1

Distance is measured in meters from the bank west of the Red River
towards N60°E direction (Anh, 2005
Annual rate of subsidence during the period 1995-1998 (Anh, 2005)
Well field locations and respective discharge capacity
Three dimensional drawdown view of Hanoi area, showing the cone of
depression (data provided by K2, 2004)
Land subsidence map period 1988-1995: (a) total view of Hanoi, (b)
areas detected land subsidence (K2, 2001)
Surface settlement at 10 land subsidence monitoring stations (data
provided by HIBT, 2004)
Settlement rate at 10 land subsidence monitoring stations
Overlaying settlement data at monitoring stations on ERS differential
interferogram
Locations of benchmarks listed in Table 4.3. Numerals represent annual
rates of subsidence revealed by leveling survey in the period 1994-1995
Leveling settlement data versus DInSAR settlement data
Overlaying settlement rate at monitoring stations on JERS-1 differental
interferogram
Land subsidence inventory map period 1988-1995: (1) settlement rate is
less than 10mm/y, (2) settlement rate is between 10-20mm/y, (3)
settlement rate is between 20-40 mm/y (updated after K2, 2001)

5.2a

5.2b

Quaternary thickness contour map over Hanoi area

5.2c

Quaternary thickness interpolated map over Hanoi area (kriging
interpolation from 77 observation boreholes
Drawdown map over Hanoi area
Drawdown interpolated map
Groundwater dynamic distribution over Hanoi area
Index map of groundwater dynamic regime factor: (0) natural
groundwater dynamic zone; (1) medium destructive groundwater
dynamic zone; (2) strongly destructive groundwater dynamic zone.
Soft clay distribution over Hanoi area (update data from Toan, 2005): (0)
no data; (1) area without clay; (2) area distributed by soft clay
Population densities over Hanoi area
Industrial zones in Hanoi city, buffered map by 5 km
Land subsidence monitoring locations, well field locations
Weighted overlay computation
Weighted overlay of 3 layers (drawdown, quaternary sediment thickness
and land subsidence inventory map); (1) Low susceptibility of occurring
land subsidence; (2) Medium susceptibility of occurring land subsidence;
(3) High susceptibility of occurring land subsidence; (4) Very high
susceptibility of occurring land subsidence
Weighted overlay seven factors controlling Hanoi Land subsidence
Weighted overlay result of 6 factors (including soft clay, quaternary
sediment thickness, drawdown, population density, groundwater dynamic
regime factor, land subsidence inventory map) controlling Hanoi land
subsidence


5.3a
5.3b
5.4a
5.4b

5.5
5.6
5.7
5.8
5.9
5.10

5.11
5.12

3D view of quaternary sediment distribution over Hanoi area

vii

76
77
77
78
79
79
80
81
82
82

83

84
85
83
84
85
86
87

88
89
90
91
91
92

93
94


List of Illustrations
Term

Description

DEM
ERS
ESA
GPS

HIBT
HWBC
InSAR
JERS-1
JICA
K2
LOS
PS
PSC
PSI
SAR
SLC
SRTM-(X)

Digital Elevation Model
European Remote Sensing satellite
European Space Agency
Global Positioning System
Hanoi Institute of Building Technology
Hanoi Water Business Company
Interferometric Synthetic Aperture Radar (also radar interferometry
Japanese Earth Resource Satellite – 1
Japan International Cooperation Agency
Hydrogeological Division No. 2
Line of Sight direction of the satellite (also slant range direction)
Permanent Scatterer
Permanent Scatterer Candidate
Permanent Scatterer Interferometry
Synthetic Aperture Radar
Single Look Complex

Shuttle Radar Topography Mission (X stand for X- band of the
microwave wavelength

viii


Chapter 1
INTRODUCTION
1.1

Problem statement

Ground deformation, and in particular subsidence, are major issues in terms of damage to
infrastructures, implying remarkable costs for prevention and compensation. The causes of
subsidence phenomena can be numerous; among others ground settlement due to liquid
extraction is very popular over the world. Aiming at an effective risk management,
monitoring techniques and integrated research are required in order to allow a better
knowledge of ground deformation phenomena and of their possible evolution.
Hanoi (figure 1.1), the Capital of Vietnam, is located in the Red River delta, a low-lying
area in northeastern country, around 100km from the Gulf of Tonkin, the East Sea. The
city is the centre of education, culture, economy, trade, travel, transportation of Vietnam.
The development along with the continuously increasing of population requires the
increasing consumption of water resources. The city’s population is 3 million and the water
for domestic and industrial use in Hanoi comes from wells located within and around the
city. The groundwater exploitation capacity has risen from 40,000m3/d in the 1950s to
750,000 m3/d currently and is expected to exceed 1 million m3/d in the 2020s. Overexploitation of groundwater in Hanoi is suspected to cause land subsidence which in turn
may affect the upper infrastructures of the city.
Hanoi area is underlain by young quaternary sediments including organic and inorganic
clays, silts, peats that are highly compressible, will in turn manifest the vulnerability of
land subsidence phenomenon in Hanoi area. The more the compressible of the subsoil the

higher the susceptibility of occurring land subsidence. As the elevation of most part of the
city is lower comparing to the Red River water level, the city was isolated by the
embankment and dam systems, and land subsidence problem becomes more serious.
The investigation of land subsidence in Hanoi area is still sporadic and in its infancy stage
with very few monitoring data of good quality. The techniques used to monitoring land
subsidence in Hanoi area are poor and were not enough to reflect the reality status of Hanoi
land subsidence.
The techniques based on revisiting ground-based benchmarks such as optical leveling or
Global Positioning System (GPS) generally offer precise measurements but limited by
their spatially sparse and infrequent sampling. As a result, the ground-based bench-marks
over Hanoi area are very few; there are only 10 land subsidence monitoring stations for
now. Denser and more frequent subsidence observations are essential for understanding the
ground subsidence distribution over Hanoi area.
Satellite-based Interferometric Synthetic Aperture Radar (InSAR), developing during
1980-1990s, offers the potential for acquiring displacement observations over wide areas
with a high resolution both spatially and temporally. Permanent Scatterer Interferometric
Synthetic Aperture Radar (PSInSAR) is a newly developed surface displacement
observation technique based on conventional InSAR. It is an operational tool for precise
ground deformation mapping on a sparse grid of phase stable radar targets, acting as a
“natural” geodetic network. This technique identifies, estimates and removes atmospheric
distortions, leaving the PS displacement as the only contribution to the signal phase shift.
SAR data and, in particular, PSInSAR can be very useful for geological hazard assessment
(e.g. landslides, subsidence, earthquakes). This approach provides fast and updatable data
acquisition over large areas, and can be integrated with conventional methods of
investigation (e.g. field surveys, air photos interpretation). Of particular interest is the
possibility to combine deformation measurements with geological data in a geographical
information system (GIS).

1



Unfortunately, PSInSAR even InSAR techniques are still not common used in
Southeastern countries especially in Vietnam. This technique requires the user with good
knowledge of interferometry, it requires data archive, time and experience for data
processing. The complex atmospheric conditions, the unavailable data, in additions with
lack of InSAR experts are the main limitations for applying this technique in Vietnam.
However many challenges, to reach and explore about the new and feasibility quantitative
method for land subsidence assessment, this study will review on the theory of PSInSAR
technique with possible application for monitoring and assessing Hanoi land subsidence as
a primary work.
As the PSInSAR technique is not yet being used for measuring Hanoi land subsidence, an
effective solution is the integration of the previous studies and available data of
geographical settings; geological settings; hydrological settings; groundwater extraction;
geotechnical settings, and the land use information then identify the significant factors that
control Hanoi land subsidence using Geographical Information System (GIS) to produce
the integrated assessment of land subsidence susceptibility for Hanoi area.
1.2

Objectives of the study
1) To review on a technique known as Satellite-based Interferometric Synthetic
Aperture Radar (InSAR) for monitoring land subsidence
2) To create a land subsidence susceptibility map for Hanoi area based on the results
of previous studies, combining with updated data.

1.3

Scopes of the study

To perform the objectives of this study, the following scopes need to be done:
1) To overview InSAR technique in monitoring land subsidence; the initial studies

resulted on Hanoi land subsidence using DInSAR technique obtained by Raucoules
and Carnec (1999) and Anh (2005) need to be revisited in a comparison with some
ground monitoring record to determine the ground subsidence for Hanoi area.
2) To review the previous and existing studies on Hanoi land subsidence; collection of
update information.
3) Building a land subsidence database including geological, subsoil, hydrogeological
data based on the tools of Exel, Access, Grapher, Rockwork 2006, Map info 80,
Arc GIS 9.2
4) Proposing a procedure to create a land subsidence susceptibility map for Hanoi
area.

2


Chapter 2
LITERATURE REVIEW
2.1

Backgrounds on land subsidence

2.1.1 A brief review of land subsidence in over the world
The worldwide exploitation of groundwater resources and the consequent water-level
declines are creating many areas of land settlement associated distress to the infrastructure.
Examples of cities and countries experiencing land subsidence are Wilmington, Texas; San
Joaquin and Santa Clara Valley, California, Savannah, and Georgia, USA; London,
England; Mexico City, Mexico; Shanghai, China; Bangkok, Thailand; Venice, Italy; The
Netherlands; and Osaka and Tokyo, Japan (Poland et al, 1984).
Over 150 regions of contemporary settlement are known, some with as much subsidence as
10 m in countries such as Mexico, Japan, and the United States (Poland, 1984). Many more
areas of subsidence are likely to develop in the next few decades as a result of accelerated

exploitation of natural resources in order to meet the demands of increasing population and
industrial development.
Many major cities in Asia (Shangtai and Tianjin in China; Japanese cities in Saga, Nobi
and Northern Kanto plains; Taipei of Taiwan and Bangkok of Thailand etc) were
developed in deltaic areas, which are commonly underlain by young and compressible
clayey soils in alternation with sandy and gravel layers. These environments are very
sensitive to occurring land settlement.
Land subsidence in Shangtai started in 1921 and had undergone a rapid rate of 23 to
100mm/year during the period from 1921 to 1965. The rate of settlement was reduced to 2
- 16mm/year from 1965 to present. However, it is noticed that due to the latest economic
and industrial development of Shanghai the subsidence rate started increasing again after
1990.
In Tianjin city (China), land subsidence was recorded early. Before 1950s, the rate of
settlement is about 0.7 to 7.9mm/year which then increased to about 50mm/year in the
early 1960s, to 170mm/year in period of 1974 - 1977, and over 100mm/year to a maximum
of 250mm/year in the early 1980s. From 1983 to 1997, land subsidence rate decreased to
about 20 - 30mm/year due to the reduction of pumping rate.
Subsidence in Bangkok has been noticed since 1970s when groundwater started to be
extensively exploited. There are about 10,000 wells in total that have been drilled to tap the
groundwater from the Bangkok aquifer system. According to Phien-wej, impact of land
subsidence to Bangkok city has been particularly critical because of the flat low-lying
topography and the presence of a thick soft clay layer underneath that augment flood risk
situation and foundation problems. The subsidence reached the most critical value in the
early 1980s when it occurred at the rate as large as 120 mm/year. After that, it decreased
but the subsidence zone expanded following the uncontrolled growth of the city. Despite of
various attempts implemented to remedy the crisis groundwater pumping from the thick
confined aquifer system continued to increase from 1.2millions m3/day in the early 1980s
to more than 2.0millions m3/day at the turn of the century. Piezometric levels in the main
aquifer layers have been lowered down as much as 70 m. The monitoring data showed a
clear correlation between the trends of surface subsidence and the decline of the

piezometric levels occurring during the last 20 years.
In Japan, subsidence had been identified in the early of the 20th century (Tokyo, 1910s;
Osaka, 1920s). War world II disrupted industrial activities, greatly reducing the industrial
usages of groundwater and thus halting land subsidence. However, subsidence began again
in the 1950s, particularly in metropolitan areas where industries revived and groundwater

3


demands increased rapidly. At present, land subsidence still persists in the suburbs of
Tokyo (in the Northern part of the Kanto plain), in the rural area of Shiroshi region of the
Saga plain where groundwater is extensively used for agriculture. In Shiroshi, land
subsidence has been observed since 1957 and its accumulated amount has reached 123 cm
over the period from 1960 to 1998, affecting an area of 324 km2.
Land subsidence in Taiwan occurs mostly in the SW coastal areas in the Lanyang and
Tapei basins. It started in mid 1950 with the rate about 3mm/year during the 1954 to 1979
period. The land subsidence rate was found accelerating during January to June, 1981 to
reach 50mm/month when there was over-pumping of groundwater for industrial usages.
From 1992 to 1994 a maximum magnitude of land subsidence of 15.4 cm was found as
reported by Liao et al. In February 1971, groundwater control zones were delineated with
associated regulations published to manage groundwater utilization and to prevent land
subsidence and seawater intrusion in Taiwan. Among those subsiding areas, groundwater
pumping in Taipei basin has been controlled since 1978, and thus subsidence in the basin
has been stabilized. However, in the other areas with excessive use of groundwater land
subsidence has not been controlled yet. The total subsidence area in Taiwan reaches 1,747
km2, which is nearly 6.4 times of the plain area in Taiwan.
2.1.2 Land subsidence mechanism
2.1.2.1 The causes of land subsidence
Land subsidence is a gradual settling or sudden sinking of the Earth's surface owing to
subsurface movement of earth materials.

According to Poland et al., (1984), the major causes of land subsidence phenomenon are:
(1) loading of the land surface; (2) vibration at or near the land surface; (3) compaction due
to irrigation; (4) solution due to irrigation; (5) drying and shrinkage of deposits; (6)
oxidation of organic materials; (7) decline of the water table; (8) decline of artesian
pressure in water sand; (9) decline of pressure in oil zones due to the removal of oil and
gas; and (10) tectonic movements.
Nowadays, the land subsidence resulted from fluid (e.g. groundwater, oil) withdrawal in
weakly consolidated materials is causing many serious damages. It therefore, attracted a lot
of study over the world. The loss of fluid causes consolidation of the empty pore spaces,
which means that any void in the soil previously filled with fluid are compressed by the
mass of the overlying materials, effectively decreasing the soil volume and resulting in
land subsidence. There are 3 types of fluid withdrawal by man that have caused noticeable
subsidence under favorable geologic conditions are: (1) the withdrawal of oil, gas, and
associated water; (2) the withdrawal of hot water or stream for geothermal power; and (3)
the withdrawal of groundwater. Each of the three types of withdrawal has produced
maximum subsidence of the same order of magnitude.
2.1.2.2 One-dimensional consolidation theory
In the field of soil mechanics, Karl Terzaghi (1925) developed the theory of primary onedimensional consolidation of clays that has served as the basis for solution of most
practical soil mechanics and settlement problems in the past. This theory commonly is used
to estimate the magnitude and rate of settlement or compaction that will occur in finegrained clayey deposits under given change in load (stress).
According to the theory, compaction results from the slow escape of pore water from the
stressed deposits, accompanied by a gradual transfer of stress from the pore water to the
granular structure of the deposits. The basic principle of effective stress was shown in
equation 2.1 and Figure 2.1.
σ’ = σ – u

4

(2.1)



Where
σ’ = effective stress (effective overburden pressure or grain-to-grain load)
σ = total stress (geostatic pressure), and
u = pore water pressure (fluid pressure or neutral stress)
The increase in vertical effective stress is equal to the change in pore-water pressure:
(2.2)
dσ ' = − du
As we know:
(2.3)
u = ρ g ψ = ρg ( h − z )
Then, dσ ' = − ρgdh
(2.4)
Where ρ denotes water density
g denotes gravity acceleration
ψ denotes pressure head
z denotes elevation head and it is constant at a point under interest
h denotes the hydraulic head
Equation 2.4 implies that a change in effective stress at a point in the system is governed
by a change in the hydraulic head at that point.
2.1.2.3 Effects of land subsidence
Potential effects of land subsidence include differential changes in elevation and gradient
of stream channels, drains, and water transport structures; reducing flood protection;
failure of water well casings due to compressive stresses generated by compaction of
aquifer systems; tidal encroachment in lowland coastal areas; tensional or compression
strain in engineering structures and houses.
2.1.3 Soil Compression due to groundwater pumping
Land subsidence caused by groundwater pumping is a sum of compression of the sand and
clay layers. Clay layer contributes a major part in the total compression of the soil profile
and its calculation is based on Terzaghi’s one-dimensional consolidation theory. However,

in his theory, the creep behavior of the soil skeleton is not considered and the thickness of
clay layer is assumed not to be changed during the consolidation process.
To calculate the compression of soil layer, the changes in the piezometric head have to be
calculated from hydrogeologic model and then input as a boundary condition to solve
equation (2.1) to find the distribution of pore pressure in the clay layer adjacent to the
pumped aquifer. The decrease in the pore pressure is equal to the increase in the effective
stress. Then the compression of a clay layer can be calculated as (Duc, 1999):
⎛ σ 0 + Δσ ' ⎞
⎟⎟
S c = RR(orCR)b log⎜⎜
(2.5)
'
⎠
⎝ σ0
Where: Sc = the compression of a clay layer, RR = Recompression Ratio, and CR =
Compression Ratio of clay determined from consolidation test (the Recompression Ratio,
RR, is used when the increased effective stress is still less than the maximum past
pressure), b = thickness of clay layer (m), σ’0 = initial effective stress at the depth of the
soil layer and Δσ’ = increase in effective stress (kPa).
The compression of an aquifer (sand layer) is rather instantaneous with groundwater
extraction and can be calculated by means of following relationship:
1
(2.6)
S a = .b.γ w .dh = b.mv .γ w .dh = b.S s .dh
E

5


Where E = modulus of compression (kPa), γw=unit weight of water, mv = the coefficient of

volume compressibility (1/kPa), Ss = coefficient of specific storage (1/m), and dh = the
head drop (m).
The subsidence at a point on the ground surface (St) is the summation of the compression
of all aquifers (Sa) and clay layers (Sc) beneath that point as:
S t = ∑ (S c + S a )
(2.7)
2.1.4 Methods of monitoring land subsidence
There are a various methods available to monitor land subsidence. They include vertical
extensometers, baseline and repeated surveys of benchmarks using Global Positioning
System (GPS) or conventional survey methods; recently, Interferometric Synthetic
Aperture Radar (InSAR) method becomes a powerful and reliable method for monitoring
land subsidence over urban areas.
Vertical extensometers provide site-specific measurements of subsidence. These
instruments consist of a pipe or cable anchored at the bottom of the borehole. The pipe or
cable extends from the bottom of the borehole, through the geologic layers that are
susceptible to compaction, to the ground surface. The pipe or cable is then connected to a
recorder that frequently measures the relative distance between the bottom of the borehole
and the ground surface. Figure 2.2 provides a schematic of a pipe extensometer and Figure
2.3 shows the above ground portion of a cable extensometer with data recording
equipment. These instruments detect changes in land surface elevation to 1/100 of a foot on
a daily basis. The relationship between subsidence, groundwater level fluctuation and the
physical properties of the various aquifer systems can be evaluated by installing a multicompletion monitoring well adjacent to the vertical extensometer.
Leveling using GPS surveying or conventional leveling are alternatives to vertical
extensometers. GPS surveying is used to monitor subsidence over greater distance or at a
regional scale. Benchmarks or “geodetic stations” are used along a network. Ground
elevations at each benchmark can be obtained with plus or minus one inch of accuracy
with GPS surveying. For regional scale surveys of this type, conventional leveling is less
accurate. The land surface elevations are initially surveyed and then resurveyed every few
years to track changes in elevation at the benchmarks and monitor trends over time.
InSAR, interferometric synthetic aperture radar, is an increasingly popular alternative to

extensometers and GPS. InSAR is space-borne, remote sensing technique that uses
changes in satellite radar signals created by interferences on the earth’s surface to measure
changes in land surface elevation. It is used to measure and track deformation on the
earth’s surface to measure changes in land surface elevation. It is used to measure and
track deformations in the earth’s surface caused by earthquake, volcanoes, and by
groundwater and fossil fuel extraction and injection. Similar to GPS, InSAR enables
measurement of subsidence on a regional scale and, like extensometers, the accuracy of
elevation measurements with InSAR can be within a fraction of an inch. InSAR is a cost
effective means of monitoring subsidence in the appropriate situation. Unfortunately,
InSAR is not well suited for the predominantly agricultural lands.
2.1.5 Land subsidence analysis
Practically, an analysis of land subsidence for a multi-aquifer system consists of two parts:
groundwater modeling and consolidation settlement analysis. The former is used to analyze
ground water flow in the pumped aquifers, while the later is used to calculate the
compression of the adjacent aquitards.
The mechanism of land subsidence due to pumping can be briefly described as: drawdown
from the pumped aquifer induce pore pressure dissipation from the overlying and
6


underlying clay layers, thus, increasing in this way their effective stress and consequently
causing their compression.
The drawdown in the pumped aquifer, calculated from a ground water model, can serve as
boundary conditions for the consolidation model. The drawdowns are actually timedependent, but for a small area like that of a well field, a pseudo-steady state flow in a
pumped aquifer would reach quickly comparing to process of pore pressure dissipation
from adjacent aquitards. Therefore, in many situations, the drawdown can be assumed as
the fixed head boundary conditions. For a multi-aquifer system of Bangkok or a bi-aquifer
system of Hanoi, a quasi-3D groundwater model (Giao 1997; Giao et al. 1999) or a 3-D
groundwater model would be more appropriately used. However, for the purpose of a
preliminary assessment and analysis of land subsidence at location of a Hanoi well field,

the use of a 2-D groundwater model in combination with a 1-D consolidation model would
be enough. The FEM formulation of the models is briefly presented below:
FEM 2-D Groundwater Model
The governing equation of 2-D groundwater flow in a confined aquifer
∂ ⎛ ∂h ⎞ ∂ ⎛ ∂h ⎞
∂h
⎟⎟ ± Q = S
(2.8)
⎜ Tx
⎟ + ⎜⎜ T y
∂x ⎝ ∂x ⎠ ∂y ⎝ ∂y ⎠
∂t
Where, h is the head in the confined pumped aquifer; Tx, Ty are the aquifer transmissivities
along x and y-directions; S is the aquifer storage; Q is pumping rate (the sign is minus for
discharge and plus for recharge, respectively)
The FE formulation of Eq. (2.8), by Galerkin's method, for a generic domain, is as follows:
⎛ ∂N k ∂N l
∂N k ∂N j ⎞
⎛ ∂hk ⎞
e
e
∫e⎜⎜⎝ Tx ∂x ∂x + Ty ∂y ∂y ⎟⎟⎠hk dΩ + ∫e SN k N l ⎜⎝ ∂t ⎟⎠hk dΩ
Ω
Ω
(2.9)
⎛ ∂N k
∂N k ⎞
e
e
⎟hk dR = 0

− ∫ QN l dΩ − ∫ N l ⎜⎜ Tx
+ Ty
∂x
∂y ⎟⎠
⎝
Ωe
Re
FEM 1-D Consolidation Model
Applying Green's Lemma to Terzaghi's consolidation equation (2.1) can obtain the weak form
for a generic element, which is linear in this case, as follows:
∂N j
∂u j
∂N i ∂N j
e
(2.10)
c
u
d
Ω
+
c
N
u
−
N
N
dΩ e = 0
e
v
i

j R
i
j
∫e v ∂z ∂z j
∫
∂z
∂t
Ω
Ωe
The second term is the flux term, which can be neglected, thus Eq. (2.8) becomes:
∂u j
∂N i ∂N j
e
(2.11)
c
u
d
Ω
−
N
N
dΩ e = 0
v
j
i
j
∫e ∂z ∂z
∫e
∂t
Ω

Ω
Calculation of Time-dependent Compression
Calculation of primary consolidation of a clay layer due to pore pressure change which has
been induced by the drawdown in the underlying aquifer can be done as follows: the
compressible layer, which is overlying the aquifer, is divided into a number of sub-layers of
small thickness, e.g., 1 or 2 m. The pore pressure change inside the compressible layer due to
pumping can be calculated by a FEM 1-D consolidation model, using TZP program
developed by Giao (1997). Based on the pore pressure change, the compression of each sub-layer can be calculated using the relationship given below (Giao, 1997):
P ' (z )
σ ' (z ) + Δu (z , t )
dS c ( z , t ) = bi .[RR(z ). log c'
+ CR( z ). log v 0
(2.12)
σ v0 (z )
Pc' ( z )
The primary consolidation of a clay layer consisting of n sub-layers is as follows:
n

S c (t ) = ∑ dS c ( z , t )
z =1

7

(2.13)


Where, dSc(z,t) is the consolidation settlement of the sub-layer at z depth, time t; Sc(t) is the
total consolidation settlement of the whole clay layer at time t; bi is the thickness of the
considered sub-layer; P'c(z) is the preconsolidation pressure of the considered sub-layer;
σ’v0(z) is the effective stress of the considered sub-layer; Δu(z,t) is the dissipation of pore

pressure of the considered sub-layer at time t; CR(z) is the compression ratio of the
considered sub-layer; RR(z) is the recompression ratio of the considered sub-layer.
2.2
2.2.1

Study area
General setting

Hanoi, with population 3.5 million, is the capital of Vietnam. The name Ha Noi means the
“the city inside river” which is located on the right bank of the Red River, at 20053’21023’N, 105023’-105043’E. Previously, Hanoi has an area of 913 km2 per total 332,000
km2 of Vietnam area. The city includes nine urban districts of Ba Dinh, Cau Giay, Dong
Da, Hoan Kiem, Hai Ba Trung, Hoang Mai, Long Bien, Tay Ho and Thanh Xuan, and five
suburban districts of Dong Anh, Gia Lam, Tu Liem, Thanh Tri and Soc Son. Hanoi's
population is constantly growing, reflecting the fact that the city is both a major
metropolitan area of Northern Vietnam, and also the country's political centre. However,
this population growth puts a lot of pressure on the infrastructures, as well as water
resources of Hanoi.
Hanoi experiences the typical climate of northern Vietnam, where summers are hot and
humid, and winters are relatively cool and dry. The summer months from May to
September receive the majority of rainfall in the year (1,682 mm rainfall/ year). The winter
months from November to March are relatively dry, although spring then often brings light
rains. The minimum winter temperature in Hanoi can dip as low as 6-7°C (43°F) not
including the wind chill, while summer can get as hot as 38-40°C (100-104°F). Autumn
reason (from September to November) with average temperature of 250 to 360 C is cool,
clear and dry. This is the best season in Hanoi, but is short, lasting no more than 50-60
days. Most of the Hanoi area is flat with elevation below 20m excepted in the northwest
hills are up to 462 m high, it is rather being declined from Northwest to Southeast along
flow direction of the Red River. Along two sides of the Red River, the dike system has
been constructed for 1000 years; the new alluvial sediments are mainly in the part out site
the dike creating higher elevation than one in side of the dike. The study area is composed

of unconsolidated sediments with 50-80m of thickness overlying the Neogene sediments.
There are many rivers flowed through Hanoi. Many of them changed their currents from
time to time, as the results, many lakes were formed. Hanoi was also called the city of
Lakes and rivers. The rivers and lakes give Hanoi a natural beauty. The Red River is the
largest river in the northern part of Vietnam. It passes through Hanoi and in the center of
the city is joined by the Duong River. The average volume of water transported by these
two rivers through Hanoi is 3500m3/s. In the flood season, the water level of these rivers
rises high. From the past, the Vietnamese people had built thousands of kilometers of
dykes by the river banks.
2.2.2

Geological setting

The Hanoi area is located in the center of the Hanoi depression which was filled by weak
consolidate Neocene rocks as sandstone, argillite, shale and unconsolidated quaternary
soils as gravels, pebbles, sands, sandy clay, clay, mud and other. It is flat and filled with
8


alluvial soils brought by the Red River. The stratigraphical formations of Hanoi area were
divided as below (see also Figure 2.4a, 2.4b):
Neogene Sediments
These deposits cover most of the area along the Red River. In the vicinity of Hanoi it had
been established a great thickness of Neogene strata comprising about 2,800m of Miocene
and 200m of Pleistocene beds. The Miocene beds are composed of sandstones, mustones
and bed of lignite which yielded plant remains and molluscas.
Pleistocene Sediments
The Pleistocene sediments are formed by gravel, sand and clay. They are between Neogene
and Holocene formation. The upper contact with the Holocene formation is conformable
while the lower contact of the formation with Neogene formation is angularly

unconformable.
According to Ky et al (1987), the Pleistocene deposits are divided into three formations:
The Thai Thuy Formation: consists of gravel interbeded with grayish clayey silt lenses in
the lower part and in the clayey silty matrix in the upper part. The Thai Thuy formation is
characteristic of the Early Pleistocene deposits. The thickness is range from 100 to 130m.
The Hanoi Formation: unconformably overlaps the Thai Thuy formation. The composition
consists of coarse-grained gravels mingled with sandy silty clay. Along the delta rims there
are sub-rounded coarse gravels mixed with sharp-edged fragments of alluvial proluvial
origin. The thickness is 40 - 50m. The Hanoi is of Middle Pleistocene to Early Late
Pleistocene age.
The Vinh Phuc Formation: consists of marine sediments of sand, silt mix with pebbles in
the lower part, and layers of yellowish brown red silty clay interbedded with fine sand in
the upper part. It contains abundant forminifera such as Ammonia becarri, Elohidium sp.,
Cibicides sp. Cibicdes sp. The thickness is 30 - 40 m and widely distributed along the rim
of the delta.
Holocene Sediments
The Holocene sediment covering the study area was brought by recent alluvium. The
composition consists of fine grained sediments such as sand, silt and clay, in which there
are sporadic occurrences of marine fossils. The formation is a mix of marine, estuarine,
and lake sediments with the dominance of alluvial origin. The thickness may reach 40m
and decreases upstream of the Red River toward northwest (Fontaine & Workman, 1978).
The Holocene sediments are divided into:
The Hai Hung Formation: consists of clayey silt and sandy layers. Foraminifera are
abundant, including Ammonia becarii, Elphidium hispidulum, Quinqueloculina sp. The
upper part of these sedimentary formations consists of a layer of grayish brownish red
marine clay, overlapping beds with fossil trees, coastal swamp deposits containing plant
remains and with giant oysters. The thickness is from 10 to 30 m.
The Thai Binh Formation: consists of clayey silt mixed with pebbly sand layers, of
alluvial-proluvial genesis, distributed along the river valleys on the first terrace. The
alluvial swamp sediments consists black grayish clay and peat distributed in the plain

center and along its rim outside the perimeter shoreline of the Hai Hung transgression.

9


2.2.3

Hanoi Subsoil Characteristics

According to previous studies, the subsoil profiles of the Hanoi area have established. It
consists of layers from ground surface as follows (Figure 2.5):
1

Top soil or weather soil layer ranges from ground surface to 3 m depth. It covers most
of the study area

2

Organic silt clay layer has a depth range from 3 to 15 m. The average thickness is about
5 m.

3

Soft silty clay layer with dark color range from 3 to 20 m. It’s about 10 m thick and in
dark grey color.

4

Stiff silty clay with yellowish to grayish brown color has the thickness of about 10-15m
and ranging from 1.5 to 25 m in the depth.


5

Silty sand has the thickness of about 2 to 5m

6

Fine to medium grained sand is yellowish brown to grey in color, the thickness ranges
from 5 to 10m.

7

Gravely sand layer is found about 30m or deeper which is light brown to dark grey in
color.

Soil investigation and characterization by both field and laboratory geotechnical testing are
very important for a correct assessment of land subsidence of an area.
2.2.4 Hydrogeological setting
2.2.4.1 Aquifer systems
The Hanoi area plays very important role in economy and life, hence the hydrogeological
study has being invested. The medium scale (1/200.000) and large scale (1/50000) of
hydrogeological maps have been compiled (Figure 2.5b). Almost of area has being
explored for construct of well fields. The national groundwater monitoring net has being
established, there are 13 monitoring points in the study area, the GW monitoring regime
has being observed since 1995. The local GW monitoring network has being established
with 80 points which are in the strong GW abstraction area (Figure 2.5c).
The Hanoi aquifer system (Figure 2.6) essentially consists of 4 units from the ground
surface to the depth of 80m to 100m: the Holocene aquifer, the aquitard, the Pleistocene
aquifer, the Neogene sandstone bedrock.
(1) The Holocene aquifer is an unconfined aquifer distributed widely over an area of

530km2. The upper part consists of clayey and sandy layers and has a thickness of up to
10m. The lower part is made up of various sands mixed with gravel with the average
thickness of between 9.2m in the north and 13.3m in the south of the Red River delta. The
transmissivity for the Holocene aquifer is from 20 to 800m2/day. The specific yield is
between 0.01-0.17. The water level is 3-4m below the surface. In the south of the Red
River, the water level is lower due to the groundwater pumping. The specific capacity in
test wells is from small to 4.5 l/sm. The recharge sources are rainfall, irrigation and river
water. Groundwater losses occur through discharge into the river in the dry season and
through evaporation and percolation into lower aquifers, also due to groundwater pumping.
(2) The aquitard underlying the Holocene aquifer consists of clay, clayey silt, and peat in
some places. This layer is not continuously extended, it tends to be laminated and
disappeared near the Red River. Where the clay layer is missing a direct hydraulic
connection, called hydraulic window, between the Holocene and Pleistocene aquifers is
formed.

10


(3) The Pleistocene or lower aquifer is situated lower in the stratigraphic sequence. The
depth to the top of this aquifer is 2-10m at the north, 5-22m at Gia Lam district and 1035m at the south. The Pleistocene aquifer is present in most part of the Hanoi area. It
consists of alluvial sand, gravel, pebble and cobble, being in general, coarser toward the
lower part. The initial groundwater level is from 2 to 5 m, but in many places it was much
lowering due to hundreds of extraction wells. The Pleistocene aquifer is the most important
productive aquifer, it has a hydraulic conductivity is about 150m/d, a stranmissivity from
600m2/d to about 3000m2/d.
(4) The Neocene formation is largely extended, consisting of consolidated cobbles, gravels
and sands, interbedded with siltstone and claystone. It is essentially an impermeable rock
layer. However in some locations, it may be fissured and become water-bearing.
Relationship between groundwater and Red River
Previous studied results showed that the close relationship between qp aquifer and Red

River in trip along Red River. In this zone, the fluctuation of groundwater (GW) level is
similar to one of the Red River, therefore, the annual amplitude is lower. In direction far
from the river, the annual amplitude is decreasing.
There are two types of relationship between GW and Red River. The first type is
discovered in North of Red River and Duong river, in the flood reason (July and August)
the river water level is being risen and recharges to GW, the GW flow direction is from
North to South. The second type is being discovered in the South of Red River and Duong
River, the river always recharges to GW, so the GW flow direction is northwest –
southeast.
2.2.4.2 Groundwater extraction
Hanoi area has big population with many industrial and agricultural sectors. The water
demand is very high. The groundwater resources in Hanoi area is rather plenty so the water
supply is based mainly on GW resources. There are 3 types of GW extraction in Hanoi
urban area:
(1) Public extraction is managed by Water Business Companies for public water supplies.
The exploitation wells are being located in zone which is called well-field. There are about
150 wells in 10 well fields. The capacity is increasing with time. GW extraction in Hanoi
started in 1909 with small capacity, it increased to 25,000 m3/d (1954), to 150,000 m3/d
(1960s), to 175, 000 m3/d (1970s), and to 450,000 m3/d nowadays.
(2) Private extraction which is managed by organizations, enterprises, factories … to
supply for their own water demand. The investigated data show that there are 500 wells
with 120,000m3/d.
(3) Groundwater extraction in rural area. In the past, the groundwater abstraction is via dug
wells (5-10m of depth). In recently time, rural groundwater is produced from shallow small
diameter wells pumping water from qh and the upper qp aquifer. Normally each house has
one well.
The demand for drinking water and for domestic water use is still increasing with time.
According to the Water Master Plan for Hanoi this demand can reach 700,000 m3/day in
2010 and 1,400,000m3/day in 2020.
The GW abstraction with large amount will cause negative impacts to environment

including lowering of water level, creating cone of depression, land subsidence, and GW
pollution.
2.2.5 Previous studies on Hanoi subsidence
The study on the land subsidence in Hanoi area is still in preliminary state, but land
subsidence phenomena have been discovered. In 1988, the Hydrogeological and
Engineering Geological Division No 2 (K2) had constructed 32 benchmarks to the South

11


of Red River for measurement of land subsidence. From 1991 to 1995, the Transportation
Public Work Service managed these benchmarks with sponsorship from the Finnish
Government, and added 13 benchmarks.
The causes of a possible land subsidence in Hanoi area were listed by Giao (1997) namely:
(1) Hanoi water extraction; (2) the piezometric level in the exploited aquifers have clearly
been lowered, it has been reduced by 15 to 20 m in many locations; (3) private
groundwater wells are on a totally uncontrolled rise in Hanoi; (4) when groundwater
extraction would bypass a limit say 1 or 1.5 million m3/d, or the groundwater level in the
pumped aquifer is more than a certain critical drawndown from the land subsidence point
of view, the compression of the confining clay layer will drastically occur; (5) the
confining layer (consisted of clay, silty clay, mud), adjacent to the pumped aquifer, is
spreading all over Hanoi and was reported to be quite compressible; (6) land subsidence
signs like well protrusion, damages to the buildings were found in a number locations in
Hanoi, near the water plants, well fields of Phap Van, Ha Dinh, Mai Dich, The Children’s
Hospital, Chien Thang Garment Factory etc.
The first seminar related Hanoi land subsidence was presented by the Vietnam soil
mechanics and Foundation Society, Vietnam Geotechnical Institute, Saskachewan
University and Associated Clifton, Canada in May 1999, Hanoi.
The results of land subsidence of six monitoring stations installed by Hanoi Institute of
construction (HIC) were presented by Dung (1999). According to this study, after 7 years

since the first monitoring station started working, no clear pore pressure profile at any of
these six stations has been obtained, so they gave recommend about the quality of
piezometer sensors.
Thu at al (1999) presented some results of a consolidation analysis at Phap Van and Mai
Dich locations, using programs by Geoslope, i.e., Seep/w and Sigma/W and Modflow.
Modflow is a software program for pseudo-three dimensional groundwater modeling. This
model was synthesized based on well logs, geotechnical, and topographic information.
Using the Seep/W combined the Sigma/W to predict settlement in the immediate vicinity
of the pumping well and find out the boundary condition of well field.
Ngu (1999) studied land subsidence based on coupling between analytical groundwater
model, Theis’ solution, and an analytical consolidation model. He also gave the way to
determine the settlement by Terzaghi’s consolidation solution. This approach is classic
from the point of soil mechanics, and can be used for simple situations of a single pumping
well, homogeneous aquifer, homogeneous clay etc.
Tien at al (1999) presented some monitoring results of settlement of the construction works
in Hanoi as part of the project named “Land and Water Management in Vietnam” jointly
conducted by Saskatchewan University (Canada), TDC Co, and Vietnam Geotechnical
Institute. Three locations are Institute of Journalism near Mai Dich Water Plant, the
Sweden Children Hospital (Thanh Cong) near the Ngoc Ha Water plant, and the Chien
Thang (Thanh Cong) near the Ngo Si Lien water plant were selected for land subsidence
investigation.
Giao (1999) has reviewed on Hanoi land subsidence and groundwater resources. Hanoi
water supply is mainly from the groundwater. Land subsidence has been caused in some
areas of Hanoi. Experience in some cities in Asia provides that the over pumping
groundwater is reason lend to land subsidence. Monitoring stations of groundwater and
land subsidence were installed by K2 and HIC. According to his study, land subsidence has
manifested in a number of location in Hanoi and to control this problem, three
requirements have to be solved composed of good monitoring data of piezometer levels in
the pumped and adjacent aquifers as well as those of soil compression at different depth
levels; proper determination of geotechnical parameters of the soil layers to be in analyses


12


of the sand and clay layers; and good tools of groundwater and land subsidence analysis.
Giao (1997) had introduced a practically reliable technique based on a bilinear
compression model couple with a FEM solution of Terzaghi’s consolidation equation,
which has been successful using in calculation of Bangkok land subsidence.
2.2.6 Hanoi land subsidence monitoring network
The fist pore pressure monitoring stations were installed and run by K2. The project was
funded by the Hanoi Water Supply Program (HWSP). For elevation monitoring, 30 surface
settlement indicators were installed by K2 in 1988, and increased to 43 due to the new
funding of HWSP. The settlement indicator is the plate buried at 0.5 m depth with a
ceramic piece on top as the leveling point; this type was installed at the base of a preexisting groundwater well.
Since 1988, the leveling was carried out one a year. The reference benchmark is National
benchmark, located at mountainous area at Do Lo, Chuong My. Besides, 3 pore water
pressure monitoring stations, by means of electrical vibrating wire piezometers, were
installed by K2 and HWSP at Mai Dich, Phap Van and Luong Yen. In each location,
piezometers were installed at 4 depths and recorded 5 times per month on the 6th, 12th, 18th,
24th, and 30th days. The land subsidence monitoring by K2 had stopped since 1995 due to
lack of funding.
Hanoi Institute of Building Technology (HIBT) is assigned by Hanoi City Administration
as the principal investigator of Hanoi land subsidence. In 1990, HIBT installed 6 land
subsidence monitoring-stations following AIT prototypes. There are now 10 monitoring
stations built by HIBT, most of them are near the water plants (Figure 2.7). The following
parameters are measured at the monitoring stations: (i) Surface settlement; (ii) Piezometer
head; and (iii) Layer compression of various depths. In order to understand the state of
Hanoi land subsidence, a sample data are recorded as well as collected at each monitoring
station.
2.3

2.3.1

Basics of Geographical Information System (GIS)
Basics of GIS

A geographic information system (GIS) is a system for capturing, storing, analyzing and
managing data and associated attributes which are spatially referenced to the earth. In the
strictest sense, it is a computer system capable of integrating, storing, editing, analyzing,
sharing, and displaying geographically-referenced information. In a more generic sense,
GIS is a tool that allows users to create interactive queries (user created searches), analyze
the spatial information, edit data, maps, and present the results of all these operations.
Geographic information system technology can be used for scientific investigations,
resource management, asset management, Environmental Impact Assessment, Urban
planning, cartography, criminology, history, sales, marketing, and logistics. For example,
GIS might allow emergency planners to easily calculate emergency response times in the
event of a natural disaster; GIS might be used to find wetlands that need protection from
pollution etc.
During the late 1980s, the increasing availability of new friendly operating systems and
personal computers led to the use of Geographic Information Systems (GIS) by an
increasing number of researchers interested in land-use planning and assessment and hence
a progressively wider use of different commercial GIS packages. With the exception of a

13


very few cases, most papers about GIS applications for mapping in landslide regions were
published from the beginning of the 1990s. In some cases, the majority of the analyses and
map modeling are fully achieved through a given GIS package but in many other cases, the
use of GIS is only partial. Some essential parts of the analysis and modeling are performed
using external statistical packages or additional computing tools.

Geographic Information systems (GIS) have become important tools in efficiently solving
many problems in which spatial data are important. Natural resources and environmental
concerns, including groundwater have benefited greatly from the use of GIS. It is
becoming powerful computer tools for varied applications ranging from sophisticated
analysis and modeling of spatial data to simple inventory and management. GIS
incorporates data that describes population characteristics, socio-economic conditions and
the landscape and analysis the spatial relationship of these factors.
The most significant difference between GIS and other information systems and databases
is the spatial nature of the data in a GIS. The analysis functions in a GIS allow
manipulation of multiple themes of spatial data to perform overlays (Figure 2.9), buffering
and arithmetic operations on the data. With its spatial analysis capabilities, GIS technology
can play an important role in human services research thereby ensuring better service
delivery for clients.
2.3.2

Geostatistics

Geostatistics is a point-pattern analysis that produces field predictions from data points. It
is a way of looking at the statistitical properties of those special data. It is different from
general applications of statistics because it employs the use of graph theory and matrix
algebra to reduce the number of parameters in the data. Only the second-order properties
of the GIS data are analyzed. To determine the statistical relevance of the analysis, an
average is determined so that points (gradients) outside of any immediate measurement
can be included to determine their predicted behavior. This is due to the limitations of the
applied statistic and data collection methods, and interpolation is required in order to
predict the behavior of particles, points, and locations that are not directly measurable.
Interpolation is the process by which a surface is created, usually a raster data set, through
the input of data collected at a number of sample points. There are several forms of
interpolation, each which treats the data differently, depending on the properties of the
data set. In comparing interpolation methods, the first consideration should be whether or

not the source data will change (exact or approximate). Next is whether the method is
subjective, a human interpretation, or objective. Then there is the nature of transitions
between points: are they abrupt or gradual. Finally, there is whether a method is global (it
uses the entire data set to form the model), or local where an algorithm is repeated for a
small section of terrain. Interpolation is a justified measurement because of a Spatial
Autocorrelation Principle that recognizes that data collected at any position will have a
great similarity to, or influence of those locations within its immediate vicinity.
2.3.3

Kriging technique

Kriging is group of geostatistical techniques to interpolate the value of a random field (e.g.
the elevation Z of the landscape as a function of the geographic location) at an unobserved
location from observations of its value at nearby locations. The theory behind interpolation
and extrapolation by Kriging was developed by the French mathematician Georges
14


Matheron based on the Master's thesis of Daniel Gerhardus Krige, the pioneering plotter of
distance-weighted average gold grades at the Witwatersrand reef complex in South Africa.
The English verb is to krige and the most common adjective is kriging.
Kriging belongs to the family of linear least squares estimation algorithms. The aim of
kriging is to estimate the value of an unknown real function f at a point x *, given the
values of the function at some other points. A kriging estimator is said to be linear because
the predicted value is a linear combination that may be written as
n

f ( x * ) = ∑ λi f ( x i )

(2.14)


i =1

n

ε ( x ) = F ( x ) − ∑ λi F ( x i )

(2.15)

i =1

The weights λi are solutions of a system of linear equations which is obtained by assuming
that f is a sample-path of a random process F(x), and that the error of prediction is to be
minimized in some sense. For instance, the so-called simple kriging assumption is that the
mean and the covariance of F(x) is known and then, the kriging predictor is the one that
minimizes the variance of the prediction error.
2.4

Basics of Remote Sensing

In the broadest sense, remote sensing is the short or large-scale acquisition of information
of an object or phenomenon, by the use of either recording or real-time sensing device(s)
that is not in physical or intimate contact with the object (such as by way of aircraft,
spacecraft, satellite, buoy, or ship).
There are two kinds of remote sensing: Passive sensors which detect natural energy
(radiation) that is emitted or reflected by the object or surrounding area being observed.
Reflected sunlight is the most common source of radiation measured by passive sensors.
Examples of passive remote sensors include film photography, infra-red, charge-coupled
devices and radiometers; Active sensor, on the other hand, emits energy in order to scan
objects and areas whereupon a passive sensor then detects and measures the radiation that

is reflected or backscattered from the target. RADAR is an example of active remote
sensing where the time delay between emission and return is measured, establishing the
location, height, speeds and direction of an object.
Remote sensing makes it possible to collect data on dangerous or inaccessible areas.
Remote sensing applications include monitoring deforestation in areas such as the Amazon
Basin, the effects of global warming on glaciers and Arctic and Antarctic regions, and
depth sounding of coastal and oceanic depths. Military collection during the cold war made
use of stand-off collection of dangerous border areas. Remote sensing also replaces costly
and slow collection on the ground, ensuring in the process that areas or objects are not
disturbed.

15


2.5

Synthetic Aperture Radar (SAR) data: Status, Trends and Future (Rudiger
Gens, 2006)

The introduction of imaging radar can be regarded as one of the most spectacular
developments in remote sensing since the early 1960s. It opened a world of applications in
geosciences and astronomy, and provided an alternative to the traditional optical methods
of imaging, which need solar illumination and cloudiness skies. Imaging radar, obtained
using the synthetic aperture concept can be used for measuring spatial variations in the
distance to the earth, e.g., due to topography, using two radar images and the principle of
interferometry (Gabriel and Goldstein, 1988). This technique, known as interferometric
synthetic aperture radar (InSAR), applied on board of a satellite, is able to image an area
on earth with a typical revisit period of about 1 month. Temporal variations in the satelliteearth distance, due to deformations such as subsidence can be measured, with accuracies
starting at the sub-cm level. Two main groups of interferometric parameters, the design
parameters and the environmental parameters, detect the potential applications and

limitations in deformation monitoring.
SAR data has been available to researchers for nearly forty years and satellite InSAR data
has been acquired for the last fifteen years. This time has witnessed tremendous
development in processing techniques and applications on both SAR and InSAR data.
2.5.1 Characteristics of SAR data
Being acquired in the microwave part of the electromagnetic spectrum, SAR data often
provides information that is spectrally complementary to other data sources. The fact that
SAR sensors are active and, therefore, provide their own signal allows data acquisitions
during day and night as well as literally in all weather conditions. The sensor collects two
types of information when acquiring data: the strength of the signals that determines the
brightness of the image and the time when the signal is acquired. This defines where in the
image a certain signal is located. The wavelength of SAR sensors are between X- band (2.7
cm) and P-band (60 cm). Modern SAR sensors such as RADARSAT-1, ENVISAT and
ALOS provide a large variety of beam modes with various look angles and resolutions. A
number of beam modes can be combined in a so-called ScanSAR mode that allows the
coverage of much larger areas at the expense of spatial resolution. Furthermore, they allow
data to be acquired in different polarization.
Combining all these characteristics result in a variety of observation scenarios that are
feasible for a large number of applications, described in more detail in the following
section.
2.5.2 Traditional and Advanced Use of SAR
The use of SAR data for civilian purposes was established with the launch of the ERS-1
satellite in 1991. Initially the satellite, operating in C-band with the look angle of 23.5
degrees, was designed for ocean monitoring.
Although the repeat cycle was changed in the various science phases of ERS-1, the satellite
provided its users for the first time with data covering the same area on a regular basis.
This enabled scientists to pursue the idea of SAR interferometry for a number of
applications. Originally perceived as a valuable tool for topographic mapping and the
generation of digital elevation models (DEMs), InSAR soon exceed any previous
expectations for monitoring deformations. In 1993, first key papers for the interferometric

use of ERS-1 data studying earthquakes, volcanic eruptions and glacier motion were
published.
The launch of the ERS-2 and RADARSAT-1 satellites in 1995 provided scientists with a
lot more data to work with. The two ERS satellites were flown in a tandem mode (for
about nine months) that acquired data one day apart. This resulted in the best global data
set for the generation of DEMs at that time. With RADARSAT-1 scientists could now
16


study the effect of varying look angles and different spatial resolutions on the backscatter
behavior of SAR data.
With the development of the permanent scatterer technique (Ferretti et al., 2001) and the
launch of the ENVISAT satellite in 2002, the number of journal papers on InSAR related
topics consistently exceeded the ones on the traditional use of SAR. The use of permanent
scatterers, especially for studying any form of subsidence, provided a technique that
overcame many of the limitations related to traditional differential interferometric
techniques used for long-term monitoring of deformation processes. This also resulted in
an increasing number of publications on subsidence and the permanent scatterer technique
in general in recent years.
2.5.3 Status
The use of SAR went through some very dramatic changes over the years. However, the
basis SAR processing has not changed in the last forty years. Advanced processing
strategies such as InSAR have been understood and are part of the “tool box” of
geoscientific image processing. They are applied on a regular basis for hazard monitoring.
Enhanced processing capabilities make new processing strategies feasible. Current
processing capabilities allow the processing of more data sets in shorter time as well as
larger data sets. Especially for InSAR processing, larger swath data are used as they allow
greater flexibility in choosing processing parameters. The Radarsat Geophysical
Processing System (RGPS) for sea ice monitoring has continuously collected data covering
the majority of the Arctic basin in three-day to six-day snapshots during the winter seasons

starting in 1997.
2.5.4 Trends and Future
With the increased volume of data available long-term monitoring studies become feasible.
Concurring with this general trend regarding remote sensing data, systems for monitoring
natural hazards are being established. Taking a more interdisciplinary approach, they aim
at getting a far more complete picture of the conditions in the study area. Modeling is
usually an integral part of the data analysis. SAR interferometry with its continuous
measurement over large areas made this development possible. One aspect of the
monitoring of natural hazards is the desire to predict them. However, this requires the data
processing in near-real time. With increased computer power this option becomes more
and more viable.
Another noticeable trend is that user requirements are moving away from raw data towards
information. Especially commercial users ask for higher level SAR products. The demand
for geocoded or terrain corrected SAR data is growing. The general need for high quality
digital elevation models in the scientific community for the majority geoscientific
applications is there for years now. InSAR is just one of the sources to satisfy these needs.
Working on complex problems with an interdisciplinary team requires a knowledge base
broader than ever before. Mastering the own rapidly developing area of expertise is only
part of it. In addition, the relationship to the other disciplines involved in the project needs
to be understood to ensure an effective communication.
2.6

Important Parameters of SAR Satellite System

Table 2.1 gives an overview of the most important parameters. The main design
parameters are the radar wavelength (3-24 cm), the perpendicular baseline B ⊥ , which is
the effective distance between the satellites, the times of the image acquisitions known as
the temporal baseline BT, and the total number of available images N. Note that BT also
contains the information on the repeat interval ΔT and the temporal range of acquisitions,
see Table 2.2. Other parameters such as the incidence angle and the inclination are

important as well, but by using different combinations of interferograms their influence
17