Survey Review

ISSN: 0039-6265 (Print) 1752-2706 (Online) Journal homepage: http://www.tandfonline.com/loi/ysre20

Deformation Monitoring by GPS at Embankment

Dams and Deformation Analysis

E. Gökalp & L. Taşçı

To cite this article: E. Gökalp & L. Taşçı (2009) Deformation Monitoring by GPS at
Embankment Dams and Deformation Analysis, Survey Review, 41:311, 86-102, DOI:
10.1179/003962608X390021

To link to this article: http://dx.doi.org/10.1179/003962608X390021

Published online: 19 Jul 2013.

Submit your article to this journal

Article views: 17

View related articles

Citing articles: 8 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ysre20

Download by: [Library Services City University London] Date: 07 April 2016, At: 23:14
 Survey Review, 41, 311 pp.86-102 (January 2009)

DEFORMATION MONITORING BY GPS AT EMBANKMENT

DAMS AND DEFORMATION ANALYSIS
E. Gökalp1 and L. Taşçı2
1
Karadeniz Technical University, Trabzon, Turkey
2
Fırat University, Elazığ, Turkey

ABSTRACT
Deformation measurements are very important for inspecting huge engineering structures in order to
prevent disasters caused by deformation and thus protect national investment. These measurements
require the use of very accurate surveying equipment. In particular, the structure of a dam and the
physical effects it has on its surroundings need continuous monitoring in order to see if unexpected
changes have occurred. If any change is found, remedial work can be performed to protect the
structure. The Global positioning System (GPS) meets all of the requirements stated above and
therefore, GPS receivers were used in this research. The main aims of this work were the examination of
deformation at the crest of the dam caused by water load at different water levels and the weight of the
dam. Additionally, the study was undertaken to determine whether GPS measurements could reach the
Downloaded by [Library Services City University London] at 23:14 07 April 2016

accuracy requirements for dam deformation measurements. In order to monitor and examine the
deformation, a monitoring network consisting of 6 reference points and 11 object points was
established. Measurements were made four times over 2 years using dual frequency GPS receivers with
static methods. The measurements were processed and point coordinates were determined. Then
differences were calculated between periods and the differences were analyzed by iterative weighted
transformation and Fredericton methods to determine the stability of the points. Finally, the results from
the two methods were compared and discussed.

KEYWORDS: Deformation. GPS. Monitoring. Point stability.

INTRODUCTION

Dams need to be inspected and protected during their complete life cycle. At the
beginning, monitoring occurs are more frequent performed then as the dam ages the
periods between monitoring become longer. The reasons for inspecting the dams are as
follows:
• To check the structural and functional security of the dam
• To protect the lives and property of people living in the vicinity
• To acquire data for the design of similar projects in the future
Dams are affected and deformed because of internal and external loads. These
loads are not constant and can change over time. Deformation and seepage are a
function of these loads. The effects of deformation on Rock-filled dams are altogether
different from that of concrete dams. However, the deformation of both types is largely
characterized as being permanent. The weight of the dam and the hydrostatic pressure
of reservoir back water yield forces which result in a vertical movement on the
structure. The hydrostatic pressure of reservoir can also cause a permanent horizontal
deformation that is perpendicular to centreline of the dam. Elastic behaviour is smaller
in a rock-fill dam [5].
Deformation measurements are one of the important activities in engineering
surveying. There are a variety of surveying instruments and methods with different
capacities and features that are used in deformation monitoring. High precision
measurements are required and have to be made several times. Since GPS fulfils these
requirements it is the most common instrument used in deformation monitoring.

Contact: E. Gökalp e-mail: ertan@risc01.ktu.edu.tr

© 2009 Survey Review Ltd. 86 DOI 10. 1179/003962608X390021
 E. GÖKALP AND L. TAŞÇI

Presently, GPS has the capability of a baseline measurement with an accuracy of 3 mm

+ (10-6 * S). Here, S is the baseline length in km.
In order to determine the deformation at the Altınkaya dam, 3 dual frequency (L1,
L2) receivers were used. The deformation network has 6 reference points and 11 object
points. All baselines were measured using static methods, and were processed with
GeoGenius2000 software. Deformation measurements were taken four times over two
years. The purpose of this project was to monitor and analyze the deformation at the
crest of the Altınkaya dam which was caused by the water load at different water levels
combined with the dam’s weight. A secondary goal was to determine whether GPS
measurements could meet the accuracy requirements for dam deformation
measurements. Deformation analyses were made using both an iterative weighted
transformation (IWT), which does not require previous information about deformation
behaviour, and the Fredericton method. Then, the results from the two methods were
compared and discussed.

APPROACHES FOR ANALYZING THE MOVEMENTS AT REFERENCE AND OBJECT

POINTS IN DEFORMATION MEASUREMENTS

Iterative Weighted Transformation (IWT) Method

Downloaded by [Library Services City University London] at 23:14 07 April 2016

The IWT method is used when there is no previous information about the
movement of points within the network. Calculated displacement values could be
affected from datum selecting or from defining two different datums while adjusting
the measurements taken at two different periods [1]. Therefore, the weight matrix is
obtained iteratively. The IWT method can be used as follows:
• Adjusted coordinates of the points X1, X2 in the deformation network and their
cofactor (covariance) matrices Qx1, Qx2 are calculated with two separate
adjustments.
Displacement values (d1) and the cofactor matrix of d (Qd1) are calculated as

d1 = X2 – X1 (1)

Qd1 = Qx1 + Qx2 (2)

In the beginning, the weight matrix (W) is accepted as W = I. This indicates that all
the points in the network have the same importance. Therefore, the solution is similar
to the Helmert transformation, if some points are given unit weight and the others a
zero weight, i.e., W = diag {I, 0} [2].
Then d is calculated as

d = S (W) d1 (3)

Here, S(W) shows that S matrix, calculated with W=I, can be obtained as

S = ( I - H (HT W H)-1 HT W) (4)

Where the H matrix for the 3 D network is written as

e 0 0 0 z0 − y0 x0 

H = 0 e 0 − z0 0 x0 y0

  (5)
0 0 e y0 − x0 0 z0 
3m*7

87
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

where eT = (1,..........,1), m is the number of the points in the network and z 0 , y0 , x0

are approximate coordinate vectors with respect to the centre of the network [7].
In the IWT, some points in a reference network cannot be accepted as stable, in
other words, not every point has equal importance [1], [2]. Hence the weight matrix
can be obtained iteratively as

Wk = diagonal [1/di(k)] (6)

where, di(k) is the ith component of the vector dk after kth iteration
dk+1 is calculated as
dk+1 = S (Wk) dk (7)

dik+1- dik < ε (8)

If the difference is greater than ε, Wk is calculated again using the value obtained
from equation (7). The iterative procedure continues from equation 5 until the
differences between the successive estimated datum parameters approach the user
defined threshold (ε).
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Then, Qd(k+1) is calculated as

Qd(k+1) = S (Wk) Qd1 ST (Wk) (9)

Note that during the iterative procedure, some dik may be close to 0. In that case,
Wk becomes very large. This could cause numerical instabilities in calculation of Wk.
Therefore, in order to avoid this problem Wk is calculated as

Wk = Diag [1 / dik+ c] (10)

where, c is a small constant.

In order to determine unstable reference and object points in the deformation
network, following procedure could be used.
Values of ci are determined using calculated parameters in equations (7) and (9) for
each point as in equation (11).

di2 /qdi σ20 = ci (11)

Here, σ20 is determined as

σ20 = (df1 σ201 + df2 σ202) / (df1 + df2) (12)

Where df1 and df2 are degrees of freedom used while adjusting first and second
period measurements respectively. σ201 and σ202 are variance factors

ci < F(1-α, 1,df ) (13)

If the above equation is accepted, it may be said that the point is stable otherwise it
is unstable. In equation (13), α is significance level, df is the degree of freedom which
is calculated from df = df1 + df2.

88
 E. GÖKALP AND L. TAŞÇI

Fredericton Approach
In order to analyze deformation measurements, a generalized approach has been
developed by the Fredericton group. The approach is applicable to any type of
geometrical analysis, both in space and in time domains. Additionally, it can be used
for the detection of unstable points in reference networks and the determination of
strain components and relative rigid body motions in relative networks [3], [2].
The Fredericton approach determines the unstable points within the network by
analyzing the changes (∆l) in length and/or angle between two measurement periods
which are derived from a least squares adjustment of coordinates.
This approach can be applicable as follows:
All possible measurements are derived from adjusted coordinates.
∆l and its covariance matrix C∆l are calculated as

∆l = l2 – l1 (14)

Where, l1 and l2 are derived measurements from first and second period measurements
respectively.
C∆l = C11 + C12 (15)
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Test statistics (∆l/ σ∆l) are calculated using the parameter C∆l. Then, compared with
the values taken from F tables as below.

∆l/ σ∆l > F (1,df,α)0,5 (16)

Significance levels (α) selected are 0.05, 0.025, and 0.01 respectively. In other
words, the confidence levels selected are 95%, 97.5%, and 99% respectively.
The measurements with test statistics greater than their F tables values are listed.
Then, most repeated measurements are determined from the list. A decision is given
with respect to repeating results whether the point is stable or unstable.

APPLICATION

Definition of Work Area

The Altınkaya dam is 27 km south west of Bafra, Samsun and is the second largest
embankment dam in Turkey. The dam is convex towards the water. The specifications
for the dam are in Table 1.

Building of monuments of reference and object points

A reference network should have at least 6 reference points otherwise the
determination of stable points could be very complex furthermore it is not always
possible [4]. Hence, 6 reference points were built as pillars on the stable ground which
surrounds the dam. In order to monitor and measure possible displacements at the
crest, 10 object points were established at the crest when the dam was built. Since that
time only one object point, numbered 0023, was added to the deformation network
(Figures 1., 2.). This point was built because of the physical changes that had been
observed in the surrounding area.

Previous measurements in the working area

There exist deformation measurements realized by alignment and precise levelling
methods at the crest of the Altınkaya dam before this study. The previous
measurements were not considered because of the large gaps between the periods of

89
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

the measurements, inconsistency among the accuracies of the measurements, and lack
of information about accuracies of some of the measurements.

Table 1 . Specifications of Altınkaya Dam

Properties

Location Samsun

Purpose Energy

Construction (starting and completion) year 1980 -1988

Embankment type Rock-fill

Dam volume 16000000 m3

Height (from river bed) 140.00 m

Reservoir volume at normal water surface elevation 5763.00 hm3

Downloaded by [Library Services City University London] at 23:14 07 April 2016

Reservoir area at normal water surface elevation 118.31 km2

Capacity 700 MW

Annual generation 1632 GWh

Measurement of reference and object points

Two separate measurements were made at the Altınkaya dam. One between
reference points and the other within the object points. The measurement information
related to the deformation network is given in Table 2.

Table 2. Measurement information related to the deformation network

Period Measurement Date Water
GPS Receivers GPS Antenna
No Beginning End Level

1 21.09.2000 23.09.2000 Ashtech Z Surveyor 700700B_Mar.III_L1/L2 170.34 m.

2 05.06.2001 08.06.2001 Ashtech Z Surveyor 700700B_Mar.III_L1/L2 167.53 m.

3 20.09.2001 22.09.2001 Ashtech Z Surveyor 700700B_Mar.III_L1/L2 164.20 m.

4 27.05.2002 29.05.2002 Ashtech Z Surveyor 700700B_Mar.III_L1/L2 177.23 m.

Before commencing deformation measurements, all the equipment was calibrated.

In order to avoid or diminish any equipment errors the same GPS receivers and
antennas were used at the same points in all periods. Measurements at reference points
were made on pillars therefore, forced centring was achieved. The object point
measurements were taken using tripods with optical plummets and string plumb-bobs
used for centring.
Baseline lengths vary between 60 m. and 2 km. in the deformation network (Figure
1). The measurements related to the reference network were made with 3 GPS
receivers. The measurement plan on the reference network is given in Table 3. The

90
 E. GÖKALP AND L. TAŞÇI

observation period was 45 minutes with a sampling rate of 10 seconds. The satellite
elevation mask was selected at 150 in order to avoid multi-path effect and cycle slip
error.

1006

Reference
Object points
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Fig. 1. Altınkaya Dam deformation network

Two different measurement plans were applied to the survey object points.
Receiver locations and sessions related to the measurements are given in Table3. In the
first plan, two receivers were set over points 1003 and 1004. Then, the third receiver
was set over each object point about 30 minutes (Figure 2a). In the second plan, a
receiver was set over point 0003 during the all observation periods, and the other
points were measured using a leapfrog method (Figure 2b). The main goal of this
measurement plan was to correlate the observations and make loop closures.

PROCESSING OF GPS OBSERVATIONS

The measurements were processed with GeoGenius 2000 software. The accuracy of
the baselines showed a maximum value of 1.2 mm horizontally and 3.2 mm vertically
for the 4 observational periods. The adjusted coordinates and their covariances were
obtained from a free network adjustment. The variance values are given in Table 4. In
order to determine the gross error of the baseline measurements, the Tau test was
applied. The t criteria are extended by a minimum error definition based on a typical
GPS error specification. By default, no errors less than a given value are flagged. The t
criteria are extended by the fixed offset. The Minimum Error Offset is given in the
current distance unit and a relative value (Minimum Error Scale) is given in [ppm].
The default values are 5 mm (offset) + 0.5 ppm (relative). All the baselines passed the
test for selected default values in this research. Additionally, the Minimum Norm
Quadratic Unbiased Estimation (MINQUE) method was used to determine the
accuracy of the baseline measurements instead of taking the values that are given for
the GPS receiver and by software manufacturers as above [6]. The accuracy of
baselines were calculated 4 mm horizontally and 6 mm vertically by MINQUE [8].

91
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

There was little difference between the given and calculated values. Therefore, the
default values given above were used.

N
N
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Fig. 2a. First measurement plan Fig. 2b. Second measurement plan
of the object points of the object points

Table 3. Reference and object point measurement plans

Reference Network Object Network
Number of
Receiver locations Receiver locations
Session No.
Receiver locations 1st Plan 2nd Plan
1 1003-1005-1006 1003-1004-0003 0003-0005-0007
2 1003-1004-1006 1003-1004-0005 0003-0007-0009
3 1003-1004-1002 1003-1004-0007 0003-0009-0011
4 1003-1001-1002 1003-1004-0009 0003-0011-0013
5 1001-1004-1005 1003-1004-0011 0003-0013-0015
6 ---------------- 1003-1004-0013 0003-0015-0017
7 ---------------- 1003-1004-0015 0003-0017-0019
8 ---------------- 1003-1004-0017 0003-0019-0021
9 ---------------- 1003-1004-0019 0003-0021-0023
10 ---------------- 1003-1004-0021 ----------------
11 ---------------- 1003-1004-0023 ----------------

A 2-dimensional deformation analysis was made because of the lower accuracy of

the Z components with respect to the X and Y components. It was also assumed that
the vertical deformation caused by the weight of the dam did not exist. Therefore, only
horizontal displacements were determined and analyzed in this study. In order to see

92
 E. GÖKALP AND L. TAŞÇI

the real directions of the displacements, all WGS 84 Cartesian coordinates were
transformed to a local topocentric coordinate system.

Table 4. The variance values obtained from free network adjustment

Variances (σ2)
Period No
(mm2)
1 0.765
2 0.787
3 0.813
4 0.620

DETERMINATION OF UNSTABLE POINTS IN THE DEFORMATION NETWORK USING IWT

Point coordinates E, N (Appendices I., 1, 2, 3, 4) and their cofactor matrices Qx1,

Qx2 were calculated with two separate free network adjustments. The IWT determined
the displacement values (d) with two iterations (Table 5, 6, 7). In these tables the
Downloaded by [Library Services City University London] at 23:14 07 April 2016

required statistical tests were also given to determine unstable points.

Table 5. Stable and unstable points determined using IWT between the first and second
periods
Displacements (d) di2/qdi σ20 = ci ci<F(1-.05,1,200)
Point No
dN(m) dE(m) ci for dN ci for dE ci<3.90

1001 -0.0049 0.0010 1.4873 0.1571 STABLE

1002 0.0001 0.0048 0.0011 5.9790 UNSTABLE
1003 0.0008 0.0004 0.3916 0.1537 STABLE
1004 -0.0023 0.0009 2.5651 0.5130 STABLE
1005 -0.0035 -0.0000 2.9078 0.0046 STABLE
1006 -0.0062 0.0026 4.1942 1.3235 UNSTABLE
0003 0.0039 -0.0049 2.2032 4.7467 UNSTABLE
0005 0.0018 -0.0012 0.4742 0.2786 STABLE
0007 0.0046 0.0004 4.9887 0.0738 UNSTABLE
0009 0.0049 -0.0001 3.3720 0.0246 STABLE
0011 0.0044 0.0036 4.8758 4.5500 UNSTABLE
0013 -0.0024 0.0047 1.2589 7.2013 UNSTABLE
0015 0.0002 -0.0027 0.0163 2.5315 STABLE
0017 -0.0000 -0.0025 0.0000 2.3457 STABLE
0019 -0.0048 -0.0043 4.7971 6.1042 UNSTABLE
0021 -0.0006 0.0003 0.0691 0.0421 STABLE
0023 0.0021 -0.0045 0.4623 3.6080 STABLE

When the deformation ellipses are drawn from the results of the IWT (App. II., 1,
2, 3) and the tables above are examined the following results can be seen.
Points 0003, 0007, 0011, 0013 and 0019 on the crest, and also points 1002 and
1006 on the reference network, were unstable between the first and second periods.
During these periods, the water level decreased from 170.34 m to 167.53 m.
Points 0011 and 0013 had additional movements. Also, points 1001, 0015 and 0017
showed significant movement between the first and third periods, during which time
the water level decreased from 170.34 m to 164.20 m.

93
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

Table 6. Stable and unstable points determined using IWT between the first and
third periods
Displacements(d) di2/qdi σ20 = ci ci<F(1-.05,1,200)
Point No
ci<3.90
dN(m) dE(m) ci for dN ci for dE
1001 0.0029 0.0097 0.4721 8.0325 UNSTABLE
1002 -0.0103 0.0052 6.1449 3.8158 UNSTABLE
1003 0.0021 0.0000 1.4975 0.0006 STABLE
1004 -0.0017 -0.0007 0.9620 0.3295 STABLE
1005 0.0024 -0.0014 0.8072 0.6271 STABLE
1006 -0.0070 0.0002 4.0799 0.0133 UNSTABLE
0003 0.0064 -0.0050 7.1295 7.8394 UNSTABLE
0005 0.0046 0.0015 2.7285 0.6077 STABLE
0007 0.0058 -0.0029 6.2107 2.7319 UNSTABLE
0009 0.0033 -0.0017 1.0282 0.4601 STABLE
0011 0.0075 0.0003 2.1515 0.0175 STABLE
0013 -0.0018 0.0031 0.3367 1.7885 STABLE
0015 0.0002 -0.0060 0.0150 8.5250 UNSTABLE
0017 -0.0053 0.0022 4.0967 1.3495 UNSTABLE
0019 -0.0076 -0.0010 6.9080 0.3137 UNSTABLE
Downloaded by [Library Services City University London] at 23:14 07 April 2016

0021 -0.0021 0.0010 0.4181 0.2324 STABLE

0023 -0.0017 -0.0045 0.1943 2.0689 STABLE

Table 7. Stable and unstable points determined using IWT between the first and
fourth periods
Displacements (d) di2/qdi σ20 = ci ci<F(1-.05,1,200)
Point No
dN(m) dE(m) ci for dN ci for dE ci<3.90

1001 -0.0057 -0.0006 2.3763 0.0691 STABLE

1002 -0.0013 -0.0011 0.5584 0.3176 STABLE
1003 -0.0009 0.0023 0.6948 1.3540 STABLE
1004 0.0014 -0.0028 0.9184 3.4756 STABLE
1005 -0.0018 0.0014 0.3943 0.2627 STABLE
1006 0.0001 0.0000 0.0123 0.0215 STABLE
0003 0.0075 -0.0042 15.0231 4.4901 UNSTABLE
0005 0.0047 -0.0007 3.7995 0.1225 STABLE
0007 0.0042 0.0018 4.9409 0.9578 UNSTABLE
0009 0.0076 0.0026 8.9717 1.3279 UNSTABLE
0011 0.0031 0.0100 3.4081 29.6775 UNSTABLE
0013 -0.0029 0.0073 2.2329 15.2726 UNSTABLE
0015 0.0010 -0.0045 2.6374 5.4270 UNSTABLE
0017 -0.0028 0.0019 2.7679 1.3061 STABLE
0019 -0.0054 -0.0030 7.6885 2.6844 UNSTABLE
0021 -0.0005 -0.0033 0.0760 3.0798 STABLE
0023 -0.0010 -0.0022 0.1427 0.8921 STABLE

Points 1001, 1002 and 1006 had additional movement with points 0009, 0011, 0013
showing significant movement. Despite the fact that some object points seemed not to
be moving, the F test statistics of displacements were very close to F table values (see
Table 7). Indeed, all the points on the crest were moving between the first and fourth
periods. In this time periods, the water level increased from 170.34 m to 177.23 m.
From the results above, the following conclusions can be drawn
• Movement on the dam’s crest was mostly effected by water loading at different
water levels. Water levels were 170.34 m. in first period, 167.53 m. in the
second period, 164.20 m. in the third period, and 177.23 m. in the final period.

94
 E. GÖKALP AND L. TAŞÇI

• In this study, the most significant movements were seen at points 0003, 0007,
0013 and 0019 in the middle of the dam’s crest and at the ends of the crest.
• Some of the reference points (1001, 1002, and 1006) were determined unstable
especially between the first and second and the first and third periods. This
shows that the method has the capability of determination of unstable reference
points in reference networks. This will help in selection of the best minimum
constraints and the best deformation models at the later stages of the
deformation analysis. Since the mentioned reference points placed at the left
side of the dam’s reservoir, there is a suspicion that this part of the dam may be
affected by water load at different water levels.

DETERMINATION OF UNSTABLE POINTS USING THE FREDERICTON APPROACH

Table 8. Derived baselines from the first and second periods and F test results
∆l/σ∆l ∆l/σ∆l ∆l/σ∆l
Baselines ∆l/σ∆l <F(1,200,0.05)0.5 <F(1,200,0.025)0.5 <F(1,200,0.01)0.5
=1.974 =2.236 =2.645
1002-1004 2.227 1.974 2.236 2.645
1002-1005 2.213 1.974 2.236 2.645
1002-0015 3.030 1.974 2.236 2.645
Downloaded by [Library Services City University London] at 23:14 07 April 2016

1002-0017 3.085 1.974 2.236 2.645

1002-0019 4.377 1.974 2.236 2.645
1002-0023 2.654 1.974 2.236 2.645
1003-1006 -2.075 1.974 2.236 2.645
1004-0003 -2.879 1.974 2.236 2.645
1004-0007 -2.126 1.974 2.236 2.645
1004-0009 -2.005 1.974 2.236 2.645
1004-0015 -2.160 1.974 2.236 2.645
1004-0017 -2.229 1.974 2.236 2.645
1004-0019 -2.801 1.974 2.236 2.645
1004-0023 -2.369 1.974 2.236 2.645
1005-0007 2.897 1.974 2.236 2.645
1005-0009 2.429 1.974 2.236 2.645
1005-0011 3.307 1.974 2.236 2.645
1006-0003 -2.553 1.974 2.236 2.645
1006-0019 -2.221 1.974 2.236 2.645
0003-0013 -2.743 1.974 2.236 2.645
0003-0019 -2.385 1.974 2.236 2.645
0007-0013 -2.712 1.974 2.236 2.645
0007-0019 -2.785 1.974 2.236 2.645
0009-0013 -2.482 1.974 2.236 2.645
0009-0019 -2.595 1.974 2.236 2.645
0011-0013 -2.291 1.974 2.236 2.645

Table 9. Point repetition numbers for the first and second periods

Point Numbers
F TEST
1001
1002
1003
1004
1005
1006
0003
0005
0007
0009
0011
0013
0015
0017
0019
0021
0023

F(95%)
0 6 2 8 4 2 4 0 4 4 3 4 2 2 7 0 2
Repetition
F(97.5%)
0 4 0 3 3 1 4 0 3 3 3 4 1 1 6 0 2
Repetition
F(99%)
0 4 0 2 2 0 2 0 3 0 2 2 1 1 4 0 1
Repetition

95
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

Table 10. Derived baselines from the first and third periods and F test results
∆l/σ∆l ∆l/σ∆l ∆l/σ∆l
Baselines ∆l/σ∆l <F(1,200,0.05)0.5 <F(1,200,0.025)0.5 <F(1,200,0.01)0.5
=1.974 =2.236 =2.645
1001-1003 2.3409 1.974 2.236 2.645
1001-1004 2.7442 1.974 2.236 2.645
1001-1005 2.3847 1.974 2.236 2.645
1001-1006 2.4778 1.974 2.236 2.645
1001-0003 3.6948 1.974 2.236 2.645
1001-0005 2.0193 1.974 2.236 2.645
1001-0007 3.2342 1.974 2.236 2.645
1001-0009 2.6720 1.974 2.236 2.645
1001-0015 3.9526 1.974 2.236 2.645
1001-0019 2.5788 1.974 2.236 2.645
1001-0023 3.0530 1.974 2.236 2.645
1002-1003 -2.2799 1.974 2.236 2.645
1003-0017 -2.6682 1.974 2.236 2.645
1003-0019 -2.9423 1.974 2.236 2.645
1004-0003 -3.3358 1.974 2.236 2.645
1004-0007 -2.4693 1.974 2.236 2.645
1004-0015 -2.3278 1.974 2.236 2.645
1005-1006 -2.1053 1.974 2.236 2.645
Downloaded by [Library Services City University London] at 23:14 07 April 2016

1005-0019 -2.5889 1.974 2.236 2.645

1006-0003 -2.1018 1.974 2.236 2.645
0003-0013 -2.7427 1.974 2.236 2.645
0003-0017 -4.0208 1.974 2.236 2.645
0003-0019 -4.0325 1.974 2.236 2.645
0003-0021 -2.3600 1.974 2.236 2.645
0005-0017 -2.5720 1.974 2.236 2.645
0005-0019 -2.7881 1.974 2.236 2.645
0007-0013 -2.3162 1.974 2.236 2.645
0007-0017 -3.6053 1.974 2.236 2.645
0007-0019 -3.7613 1.974 2.236 2.645
0007-0021 -2.0605 1.974 2.236 2.645
0009-0017 -2.1883 1.974 2.236 2.645
0009-0019 -2.4485 1.974 2.236 2.645
0011-0017 -2.2010 1.974 2.236 2.645
0011-0019 -2.4510 1.974 2.236 2.645

In order to determine unstable points in the deformation network, all possible baselines
were derived from coordinates in two separate adjustments of the measurement
periods. Then, the change in baselines, ∆l, and its covariance matrix, C∆l, were
calculated. Test statistics were calculated and compared with the values taken from F
tables. Here, confidence levels selected were 95%, 97.5%, and 99% respectively.

Table 11. Point repetition numbers for the first and third periods
Point Numbers
F TEST
1001
1002
1003
1004
1005
1006
0003
0005
0007
0009
0011
0013
0015
0017
0019
0021
0023

F(95%)
Repetition 11 2 4 4 3 3 7 3 6 3 2 2 3 6 8 2 1
F(97.5%)
Repetition 10 1 4 4 2 1 6 2 5 2 1 2 2 4 8 1 1
F(99%)
Repetition 6 0 2 2 0 0 5 1 3 1 0 1 1 3 4 0 1

96
 E. GÖKALP AND L. TAŞÇI

Table 12. Derived baselines from the first and fourth periods and F test results
∆l/σ∆l ∆l/σ∆l ∆l/σ∆l
∆l/σ∆l <F(1,200,0.05)0.5 <F(1,200,0.025)0.5 <F(1,200,0.01)0.5
Baselines
= 1.974 = 2.236 = 2.645
1001-0011 -4.0023 1.974 2.236 2.645
1001-0013 -2.8187 1.974 2.236 2.645
1002-0007 -2.4739 1.974 2.236 2.645
1002-0009 -2.8581 1.974 2.236 2.645
1002-0011 -5.7543 1.974 2.236 2.645
1002-0013 -3.3123 1.974 2.236 2.645
1003-1004 3.1962 1.974 2.236 2.645
1003-0003 4.0081 1.974 2.236 2.645
1003-0005 2.1389 1.974 2.236 2.645
1003-0007 2.2796 1.974 2.236 2.645
1003-0009 2.8807 1.974 2.236 2.645
1004-0003 -2.2123 1.974 2.236 2.645
1004-0011 5.9469 1.974 2.236 2.645
1004-0013 5.9021 1.974 2.236 2.645
1004-0017 3.7958 1.974 2.236 2.645
1005-0009 2.4140 1.974 2.236 2.645
1005-0011 2.1665 1.974 2.236 2.645
1006-0011 4.0202 1.974 2.236 2.645
1006-0013 2.6028 1.974 2.236 2.645
0003-0007 -2.2831 1.974 2.236 2.645
Downloaded by [Library Services City University London] at 23:14 07 April 2016

0003-0011 -4.0661 1.974 2.236 2.645

0003-0013 -5.0513 1.974 2.236 2.645
0003-0017 -4.4591 1.974 2.236 2.645
0003-0019 -4.3994 1.974 2.236 2.645
0003-0021 -2.5704 1.974 2.236 2.645
0003-0023 -2.3014 1.974 2.236 2.645
0005-0011 -1.9982 1.974 2.236 2.645
0005-0013 -3.0952 1.974 2.236 2.645
0005-0017 -2.5520 1.974 2.236 2.645
0005-0019 -2.7506 1.974 2.236 2.645
0007-0013 -3.0472 1.974 2.236 2.645
0007-0017 -2.6047 1.974 2.236 2.645
0007-0019 -2.9389 1.974 2.236 2.645
0009-0011 -2.1729 1.974 2.236 2.645
0009-0013 -3.4407 1.974 2.236 2.645
0009-0017 -3.1661 1.974 2.236 2.645
0009-0019 -3.5364 1.974 2.236 2.645
0009-0021 -2.0717 1.974 2.236 2.645
0009-0023 -2.0845 1.974 2.236 2.645
0011-0019 -2.5715 1.974 2.236 2.645
0011-0023 -2.0521 1.974 2.236 2.645
0013-0015 1.9125 1.974 2.236 2.645
0013-0023 -2.6025 1.974 2.236 2.645
0015-0019 -2.0950 1.974 2.236 2.645

The measurements with test statistics greater than their F tables values are listed in
Table 8, 10, 12. Finally, the most repeated measurements determined from the list and
the repetition numbers obtained are in Tables 9, 11, 13. From these tables, the points
that have the highest repetition numbers were accepted as unstable. When examining
the tables, there are very close repetition numbers among the points. Therefore, to
avoid misinterpretation the points with repetition numbers which were 50% of the
most repeated point were accepted unstable. For example, the repetition number is 8
for point 1004 with a confidence level of 95% (Table 9). Hence, when the points with
repetition numbers equal to 4 were accepted as unstable otherwise the points were
classified as stable. The summary of the stable and unstable point is listed in Table 14.
Some of the points were accepted as unstable in spite of sudden reduction of their
repetition numbers at confidence levels of 97.5% and 99% respectively. For example,
the repetition numbers of points 1004 and 0009 between first and second periods. This
occurred because the values of ∆l/σl were very close to the F table values. Therefore,

97
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

there is need for prior information about points in order to avoid misinterpretation
concerning the stability and instability of the points.

Table 13. Point repetition numbers for the first and fourth periods
Point Numbers

F TEST

1001
1002
1003
1004
1005
1006
0003
0005
0007
0009
0011
0013
0015
0017
0019
0021
0023
F(95%)
2 4 5 5 2 2 9 4 6 9 10 9 1 5 6 2 4
Repetition
F(97.5%)
2 3 4 4 1 2 8 3 6 6 6 9 0 5 5 1 2
Repetition
F(99%)
2 3 3 4 0 1 5 2 2 5 5 7 0 3 4 0 0
Repetition

Table 14. The summary list of the stable and unstable points
Point
1-2 Periods 1-3 Periods 1-4 Periods
No
1001 STABLE UNSTABLE STABLE
Downloaded by [Library Services City University London] at 23:14 07 April 2016

1002 UNSTABLE STABLE STABLE

1003 STABLE STABLE UNSTABLE
1004 UNSTABLE STABLE UNSTABLE
1005 UNSTABLE STABLE STABLE
1006 STABLE STABLE STABLE
0003 UNSTABLE UNSTABLE UNSTABLE
0005 STABLE STABLE STABLE
0007 UNSTABLE UNSTABLE UNSTABLE
0009 UNSTABLE STABLE UNSTABLE
0011 UNSTABLE STABLE UNSTABLE
0013 UNSTABLE STABLE UNSTABLE
0015 STABLE STABLE STABLE
0017 STABLE UNSTABLE UNSTABLE
0019 UNSTABLE UNSTABLE UNSTABLE
0021 STABLE STABLE STABLE
0023 STABLE STABLE STABLE

CONCLUSION AND RECOMMENDATIONS

When examining the displacement vectors, obtained by the IWT method, the points
with a displacement greater than or equal to 4 mm were accepted as unstable. This
coincides with the horizontal accuracy of the observations that were calculated by
MINQUE.
When comparing the Fredericton method with the IWT method, the Fredericton
method has the determination capacity of the same unstable points 86% in 1-2 and 1-4
periods, 65% in 1-3 period with respect to IWT.
The main reason of comparing the IWT and the Fredericton methods is to control
the obtained results. Besides, both methods have the capability of determination of the
single point displacements free from the choice of the datum in the reference networks.
The selection process of the unstable point/points became less conclusive when the
repetition of the failure was similar for most of the points in the Fredericton method.
This requires interpretation of the results.
IWT is a robust method used when there is insufficient knowledge about the
behaviour of the deformable body. This is not actually an advantage. Several methods
have been developed for preliminary identification of the deformation model in the
literature.

98
 E. GÖKALP AND L. TAŞÇI

In this work, the hypothesis that the maximum horizontal movements caused by a
water loading effect could occur in the middle of the crest of the dam in arch dams was
proved by the GPS measurements and deformation analysis methods.
A 2-dimensional deformation analysis was made because of the lower accuracy of
the Z component with respect to the X and Y components. Therefore, for the further
work, in order to determine the movement in the vertical dimension, use of precise
levelling or another method that has the potential to provide the vertical accuracy equal
to or greater than the horizontal accuracy of GPS is recommended.
In conclusion, although significant movements have been determined mainly in the
middle parts of the dam’s crest and at the ends of the crest, the Altınkaya dam was
shown to be normal in relation to the water load effect. However, in regard to
magnitude of the displacements there is neither threat to the structural and functional
security of the dam nor to the lives and property of people of people living in the
vicinity.
ACKNOWLEDGEMENT

We are grateful to Karadeniz Technical University Research Fund and Fırat

University Research Fund in Turkey for financially supporting this research.
Downloaded by [Library Services City University London] at 23:14 07 April 2016

APPENDIX I. SUMMARY TABLES OF ADJUSTED COORDINATES

Table 1. Summary table of adjusted coordinates in the 1st period (21.09.2000-

23.09.2000)
Adjusted
Pt Adjusted Local Topocentric
Geographical Adjusted WGS 84 Coordinates
No Coordinates
Coordinates
φ° λ° X(m) Y(m) Z(m) E(m) N(m) Up(m)
1001 41.36489 35.74816 3890777.0208 2800769.7960 4193006.2175 1862.6596 365.0306 -114.2394
1002 41.36820 35.73626 3891207.2589 2799852.7976 4193331.5341 866.9684 732.5919 -39.3580
1003 41.35448 35.72719 3892669.0992 2799969.0987 4192405.4796 107.6937 791.4190 290.2849
1004 41.36599 35.71796 3892270.4107 2798731.2610 4193187.8718 -664.4352 487.2131 21.8880
1005 41.35006 35.71938 3893389.9831 2799682.4001 4192119.3442 -545.8354 -1282.2716 415.0067
1006 41.36248 35.71258 3892746.2813 2798518.7981 4192899.4458 -1114.7494 97.3195 28.2250
0003 41.36160 35.72590 3892130.5971 2799449.0986 4192807.7515 0.0000 0.0000 0.0000
0005 41.36203 35.72546 3892127.6223 2799400.9403 4192843.4102 -37.3591 46.9414 0.6459
0007 41.36248 35.72507 3892119.4901 2799354.9403 4192881.1803 -69.9550 97.4011 0.4916
0009 41.36296 35.72474 3892106.7863 2799312.1828 4192921.0842 -97.2497 150.6636 0.3840
0011 41.36346 35.72448 3892089.6270 2799272.9182 4192962.8203 -119.1073 206.3437 0.3041
0013 41.36398 35.72428 3892068.2245 2799237.4516 4193006.0487 -135.4042 263.9548 0.2905
0015 41.36451 35.72416 3892042.7913 2799206.0866 4193050.5280 -146.0179 323.0846 0.4461
0017 41.36505 35.72411 3892013.4486 2799179.5479 4193095.2939 -150.4305 382.6647 0.5249
0019 41.36559 35.72412 3891980.2363 2799157.4372 4193140.0794 -148.9886 442.6272 0.2004
0021 41.36613 35.72421 3891943.1680 2799139.7680 4193183.9171 -141.6890 502.2334 -1.1525
0023 41.36421 35.72471 3892025.1429 2799250.1656 4193015.7416 -99.9263 289.4374 -13.9816

99
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

Table 2. Summary table of adjusted coordinates in the 2nd period (05.06.2001-08.06.2001)

Adjusted
Adjusted Local Topocentric
Geographical Adjusted WGS 84 Coordinates
Coordinates
Coordinates
φ° λ° X(m) Y(m) Z(m) E(m) N(m) Up(m)
1001 41.36489 35.74816 3890777.0314 2800769.7939 4193006.2252 1862.6651 365.0239 -114.2256
1002 41.36820 35.73626 3891207.2696 2799852.7998 4193331.5491 866.9773 732.5890 -39.3373
1003 41.35448 35.72719 3892669.1229 2799969.1071 4192405.5069 107.7000 -791.4220 290.3243
1004 41.36599 35.71796 3892270.4249 2798731.2612 4193187.8825 -664.4300 487.2060 21.9071
1005 41.35006 35.71938 3893390.0122 2799682.4125 4192119.3704 -545.8289 -1282.2798 415.0505
1006 41.36248 35.71258 3892746.3129 2798518.8135 4192899.4710 -1114.7420 97.3079 28.2709
0003 41.36160 35.72590 3892130.5988 2799449.0834 4192807.7551 0.0000 0.0000 0.0000
0005 41.36203 35.72546 3892127.6247 2799400.9301 4192843.4140 -37.3554 46.9392 0.6486
0007 41.36248 35.72507 3892119.4955 2799354.9342 4192881.1919 -69.9498 97.4017 0.5031
0009 41.36296 35.72474 3892106.7752 2799312.1639 4192921.0778 -97.2451 150.6645 0.3680
0011 41.36346 35.72448 3892089.6220 2799272.9084 4192962.8221 -119.0990 206.3440 0.3012
0013 41.36398 35.72428 3892068.2199 2799237.4434 4193006.0426 -135.3949 263.9483 0.2833
0015 41.36451 35.72416 3892042.7914 2799206.0724 4193050.5257 -146.0161 323.0807 0.4416
0017 41.36505 35.72411 3892013.4554 2799179.5387 4193095.2986 -150.4286 382.6606 0.5314
0019 41.36559 35.72412 3891980.2507 2799157.4311 4193140.0847 -148.9886 442.6183 0.2134
0021 41.36613 35.72421 3891943.1793 2799139.7654 4193183.9276 -141.6844 502.2287 -1.1365
0023 41.36421 35.72471 3892025.1552 2799250.1580 4193015.7539 -99.9263 289.4354 -13.9660

Table 3. Summary table of adjusted coordinates in the 3rd period (20.09.2001-22.09.2001)

Adjusted
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Point Adjusted Local Topocentric

Geographical Adjusted WGS 84 Coordinates
No Coordinates
Coordinates
φ° λ° X(m) Y(m) Z(m) E(m) N(m) Up(m)
1001 41.36489 35.74816 3890777.0184 2800769.7968 4193006.2467 1862.6738 365.0300 -114.2360
1002 41.36820 35.73626 3891207.2713 2799852.8028 4193331.5563 866.9776 732.5764 -39.3482
1003 41.35448 35.72719 3892669.1089 2799969.0984 4192405.5123 107.6999 -791.4231 290.2974
1004 41.36599 35.71796 3892270.4178 2798731.2554 4193187.8930 -664.4318 487.2039 21.8891
1005 41.35006 35.71938 3893389.9967 2799682.4017 4192119.3796 -545.8299 -1282.2764 415.0243
1006 41.36248 35.71258 3892746.3045 2798518.8061 4192899.4776 -1114.7443 97.3043 28.2489
0003 41.36160 35.72590 3892130.6006 2799449.0862 4192807.7790 0.0000 0.0000 0.0000
0005 41.36203 35.72546 3892127.6056 2799400.9213 4192843.4175 -37.3526 46.9395 0.6174
0007 41.36248 35.72507 3892119.4762 2799354.9177 4192881.1892 -69.9531 97.4004 0.4643
0009 41.36296 35.72474 3892106.7696 2799312.1595 4192921.0873 -97.2466 150.6604 0.3509
0011 41.36346 35.72448 3892089.6182 2799272.9032 4192962.8389 -119.1022 206.3447 0.2897
0013 41.36398 35.72428 3892068.2175 2799237.4412 4193006.0585 -135.3964 263.9464 0.2734
0015 41.36451 35.72416 3892042.7827 2799206.0635 4193050.5327 -146.0194 323.0782 0.4190
0017 41.36505 35.72411 3892013.4460 2799179.5392 4193095.3028 -150.4238 382.6527 0.5106
0019 41.36559 35.72412 3891980.2445 2799157.4321 4193140.0950 -148.9853 442.6130 0.1987
0021 41.36613 35.72421 3891943.1617 2799139.7550 4193183.9255 -141.6837 502.2247 -1.1712
0023 41.36421 35.72471 3892025.1363 2799250.1458 4193015.7470 -99.9263 289.4291 -14.0055

Table 4. Summary table of adjusted coordinates in the 4th period (27.05.2002-29.05.2002)

Adjusted
Point Adjusted Local Topocentric
Geographical Adjusted WGS 84 Coordinates
No Coordinates
Coordinates
φ° λ° X(m) Y(m) Z(m) E(m) N(m) Up(m)
1001 41.36496 35.74820 3890777.0308 2800773.4435 4193018.1477 1862.6627 365.0198 -114.2515
1002 41.36827 35.73630 3891207.2677 2799856.4483 4193343.4693 866.9705 732.5841 -39.3854
1003 41.35455 35.72723 3892669.1199 2799972.7548 4192417.4262 107.7014 -791.4273 290.3099
1004 41.36606 35.71800 3892270.4224 2798734.9101 4193199.8026 -664.4346 487.2061 21.8743
1005 41.35013 35.71942 3893390.0083 2799686.0598 4192131.2889 -545.8279 -1282.2818 415.0210
1006 41.36255 35.71262 3892746.3100 2798522.4619 4192911.3907 -1114.7452 97.3105 28.2191
0003 41.36167 35.72594 3892130.5958 2799452.7303 4192819.6744 0.0000 0.0000 0.0000
0005 41.36209 35.72549 3892127.6116 2799404.5732 4192855.3253 -37.3556 46.9385 0.6138
0007 41.36255 35.72510 3892119.4934 2799358.5833 4192893.1124 -69.9490 97.3977 0.4740
0009 41.36303 35.72478 3892106.7727 2799315.8123 4192932.9977 -97.2430 150.6636 0.3389
0011 41.36353 35.72451 3892089.6196 2799276.5562 4192974.7423 -119.0933 206.3392 0.2778
0013 41.36405 35.72432 3892068.2171 2799241.0880 4193017.9637 -135.3929 263.9441 0.2655
0015 41.36458 35.72419 3892042.7888 2799209.7201 4193062.4458 -146.0186 323.0779 0.4083
0017 41.36512 35.72414 3892013.4519 2799183.1867 4193107.2179 -150.4248 382.6541 0.5078
0019 41.36566 35.72416 3891980.2482 2799161.0792 4193152.0043 -148.9880 442.6141 0.1933
0021 41.36619 35.72424 3891943.1794 2799143.4140 4193195.8479 -141.6889 502.2251 -1.1682
0023 41.36428 35.72474 3892025.1523 2799253.8065 4193027.6734 -99.9247 289.4287 -13.9985

100
 E. GÖKALP AND L. TAŞÇI

APPENDIX II. DEFORMATION ELLIPSES FOR OBJECT AND REFERENCE POINTS

1002

1004
0021
0019 1001
0017
0015 0023 0021
0013
1006 0011
0009 0019
0007
0005 0017
0003
0015 0023
0013
0011
0009
1003 0007
0005
0003
100

Fig. 1. Deformation ellipses for periods 1 and 2

1002

1004
0021
Downloaded by [Library Services City University London] at 23:14 07 April 2016

0019 1001
0017
0015 0023 0021
0013
1006 0011
0009
0007 0019
0005
0003
0017
0023
0015
0013
0011
1003 0009
0007
0005

1005
0003

Fig. 2. Deformation ellipses for periods 1 and 3

1002

1004
0021 1001
0019
0017 0021
0015 0023
0013
1006 0011 0019
0009
0007
0005 0017
0003
0015 0023

0013
0011

1003
0009
0007
0005
1005 0003

Fig. 3. Deformation ellipses for periods 1 and 4

101
 DEFORMATION MONITORING BY GPS AT EMBANKMENT DAMS AND ANALYSIS

References

1. Chen, Y.Q., 1983. Analysis of Deformation Surveys- A Generalized Method.

Technical Report No: 94, University of New Brunswick Surveying Engineering.
Fredericton, N.B., Canada.
2. Chen,Y.Q., Chrzanowski A., Secord, J.M., 1990. A Strategy for the Analysis of the
Stability of Reference Points in Deformation Surveys. CISM / JOURNAL ACSGC
Vol: 44, No:2.
3. Chrzanowski A., Chen,Y.Q., 1986. Report of the Ad HOC Committee on the
Analysis of Deformation. XVIII. International Congress FIG, Toronto, Canada.
4. Chrzanowski A., 1992. Interdisciplinary Approach to Deformation Monitoring and
Analysis. IUSM of the Working Group Sessions, Washington, D.C. U.S.A.
5. Engineer Manual, EM-1110-2-1004, 1994. Deformation Monitoring and Control
Surveying. U.S. Army Corps of Engineers. Washington, DC.
6. Gökalp, E., 1991. Evaluation of GPS measurements by MINQUE and integration
of GPS and levelling in subsidence monitoring. M. Eng. Report. Department of
Surveying Engineering. University of New Brunswick. Fredericton, N.B., Canada.
7. Kuang, S., 1996. Geodetic Networks Analysis and Optimal Design Concepts and
Downloaded by [Library Services City University London] at 23:14 07 April 2016

Applications. Ann Arbor Press. Inc. Chelsea, Michigan.

8. Taşçı, L., Gökalp, E., 2002, Evaluation of the Accuracy of Altınkaya Dam GPS
Measurements by Minimum Norm Quadratic Unbiased Estimation (MINQUE).
International Symposium on GIS. September 23-26, 2002, Istanbul, Turkey.

102