Presentation Layout
HIGH ACCURACY REAL-TIME DAM MONITORING
USING LOW COST GPS EQUIPMENT
GPS in Engineering Monitoring.
GPS Carrier Phase Multipath Error
Multipath Sidereal-day Correction Technique
Proposed Technique for using Low-Cost GPS in EM
Initial Tests of this Proposed Technique
Dam Monitoring with GPS
Conclusions and Suggestions for Further Works
Reda Ali, Paul Cross, and Ali Elsharkawi
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
GPS Versus Conventional methods
Conventional methods
Advantages
Inter-visibility not necessary
Automated operation- with high
Advantages
High accuracy can be obtained
High redundancy of observations
Gives absolute displacements of
●
● Drawbacks
● Drawbacks
●
selected points
GPS Carrier Phase Multipath Error(1/4)
GPS
observation rate
All weather condition- in real
time
S
d
S
= V cos θ
Sr = α
= Sd + Sr
S
High Cost !!!
stations
Low accuracy in real time
Bad weather conditions limit
applications !!!
observations
Are labour intensive and
required skilful observers
Not all methods suitable for real
time
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
= β V cos ( φ + ψ )
β = 1+ 2α cos(θ ) +α2
Path delay due to MP
Need inter-visibility between
V cos(φ +θ ) 0 ≤α ≤ 1
ψ = tan-1 (
=
2*π
λ
α sin ( θ )
)
1.0 + α cos ( θ )
* ψ
Geometric Aspects of Carrier Phase
Multipath
ψ
 1

tan
θ r sin θ s
−


= d r cosθ r
− cos θ cos ( φ − φ )
s
s
r 

θs
φs
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
GPS Phase Multipath Error (2/4)
GPS Phase Multipath Error (3/4)
Characteristics of GPS Phase Multipath
Characteristics of GPS Phase Multipath
Highly dependant on Site location (Not reduced
through differential processing )
It can reach magnitudes of 0.25 of wave length
(nearly 5 cm of L1 carrier)
Its value and shape depends on many factors such as:
Antenna Reflector distance
Antenna Reflector geometry
Satellite Elevation Angle and Azimuth
Signal Wave length
θr
φr
It has nearly zero mean so it may be cancelled
by averaging the residuals
It has little effect on final accuracy of positioning
for a long period of observation , but it causes a
disaster for GPS real time applications such as GPS
Engineering Monitoring
It repeats itself every Sidereal day (Mean Sidereal
day = 23h 56m 4s of mean solar day)
Reflection Coefficient
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
GPS Carrier Phase Multipath Error (4/4)
GPS Carrier Phase Multipath Repeatability (1/2)
Extraction of GPS Phase Multipath
Double Differenced 6-15
0 .0 3
= R sr + C ( dt r − dt
φ rs
R sr =
● GPS
φ
Carrier Phase observation equation
s
)+T r
− I rs + λ N sr + M sr + Noise
( X r − X s ) + ( Y r −Y s ) + ( Z r − Z s )
2
2
2
J
A
− φ
J
B
)− (φ
K
A
− φ
0 .0 1
0 .0 0
- 0 .0 1
- 0 .0 2
- 0 .0 3
13 7464
K
B
σ = ± 7.4 mm
0 .0 2
phase Double Differencing Technique
= (φ
JK
AB
s
Day 7 January 2002
Multipath error (m)
● GPS
)
13 8364
1 39264
1 40164
14106 4
G P S T im e (S e c ) [G P S W e e k 1 1 4 8 ]
0 .0 3
σ = ± 7.8 mm
Day 8 January 2002
φ
JK
AB
=
JK
JK
R + C ( dtAB−dt )+T AB − I
JK
JK
AB
AB
[
Multipath error
Multipath
error(m)
(m)
0 .0 2
+ λNAB + M AB + Noise
JK
JK
] [
]
f
f
f
jk
φ = RABj k + ρ&Aj dtA −ρ&Bj dtB − ρ&Ak dtA −ρ&Bk dtB +λ NAB
+MABjkφ +εABφ
rec
mult
c
c
c
jk
AB
0 .0 1
0 .0 0
-- 0 . 00 1
-- 00 .. 00 22
Correlation Factor = 0.895 GPS Time Diff = 86152
D ay8
-- 0
0 .. 00 3
3
223616
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
D ay 8 -D ay 7
2 24516
22 5416
σ = ± 2.6 mm σafter
= ±15
3.8second
mm
226 316
22721 6
G P S T i m e ( S e c ) [G P S W e e k 1 1 4 8 ]
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
I- Data From Previous Day (1/2)
Block Diagram of the Suggested Technique for
GPS Engineering Monitoring Software (GPSEM)
Data Input
Data from recent
day
(I)
Data from previous
day
(II)
Sidereal day
correction technique
(IV)
Quality Control
Process (CUSUM)
Compute Satellite Position
Initial Processing Steps
Variance Covariance matrix
Construct Normal Equation
Double difference
Observation Equation
No
Yes
Integer Ambiguity Fixing
Is data sufficient
(III)
Kalman Filter
Yes
Apply Kalman Filter
Detected
movements
Transfer from DD to SD
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
II- Sidereal day Correction Technique (1/3)
Format of Multipath file
Estimation of Sidereal day time shift between two days (T)
● Approximate Coordinates of Rover Station
●
Time of observation and number of used satellites



●
●
approximate coordinates of rover station.
Number used in the stochastic model indicates the strength of
signal
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Revolution Period of slave satellite is computed using
Kepler’s third law of planetary motion (t)
t =  2π
For every satellite, data are stored in one single line:
Satellite identifiers number (PRN)
Satellite position coordinate in X-direction in WGS84
Satellite position coordinate in Y-direction in WGS84
Satellite position coordinate in Z-direction in WGS84
Residual of single difference L1 phase observations with respect to
Output files
E-N-UP file
Multipath file
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
I- Data From Previous Day (2/2)
● The Stored Data for Each Epoch :
Is ambiguity fixed
●
r3
GM




G
M
r
is the universal gravitational constant
is the mass of the Earth
is the radius from the center of the Earth to
the center of the satellite.
Every GPS satellite makes two revolutions per one day so
T = 2* t
Search through previous day data in certain searching volume
centered by T, to find the time of epoch which gives the smallest
distances between location of this satellite between the two days
If this satellite does not exist in the stored file, this satellite will not
be used for any further calculations
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
II- Sidereal day Correction Technique (2/3)
II- Sidereal day Correction Technique (3/3)
Computing the Rover station positions for two days
Necessary Coordinate Transformations
●
Formation DD observation equations using data from previous day
●
Formation Weight matrix of DD Phase observations
CDD =
1
σ 2DD( r ,1 )
 2
 σ SD( r )
M
σ 02 
 2

 σ SD( r )
σ 2SD( r )
σ 2DD( r ,2 )
....
E 
N 
UP 
σ 2SD( r ) 
L M
M O M
L Lσ




2
DD( r ,N −1 )

σ
●
= F 1* F 2* F 3
F3 =e( d / D)
i
Formation of Normal matrix and solve for position associated
with their covariance matrix using LS Technique
●
Subtracting positions of two days
2
 −sin(λ )
−Cos(λ ) sin(ϕ )
Cos(λ ) Cos(ϕ )
=










UP direction up
Xˆ i (+) = X̂i (-) + Gi v̂i (−)
6- Update state covariance matrix
= [I - Gi H i ] C ˆ
C Xˆ
X i (−)
i (+ )
Time Update
Measurement Update
5- Update state vector
{ }
Q EOS={R }QWGS84R T
●
State Vector
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
M
= [E
=




N
I
L
0
UP
M
L
M
vE
S
L
T
vN
vUP ]T




{
}
S
T = diag {t } = diag {e −α∆t}
Covariance of dynamic model noise
= diag{s} = diag α -1( 1 − t )
Q1∆ t + S 22Q2

C y( ω ) =


 S 23Q2
L
M 23 2
L L
M 33 2
-α∆t − e-2α∆t ) / 2α 3
 S 22 = diag{S 22} = (-3+ 2α ∆t + 4 e

S = S 32 = diag S 32 =(1− 2 e α∆t + e-2α∆t ) / 2α 2
 23
Q

S
 S 33 = diag {S 33 } = (1 − e -2 α ∆ t ) / 2 α
S
Q
{ }
If α is very small due to long
correlation length
1

Q ∆t + Q2 ∆t 3
 1
3

C y(ω ) =

1
Q2 ∆t 2

2

L
If α is very large due to
short correlation length
M 12 2 2
L L
M 2
Q ∆t
Q ∆t






C y (ω ) =
0
QE

0


 0
QN
0
0
0 
QUP
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
IV- Quality Control Process
Control chart
● Failure Detection and Identification (FDI) Process
XT
Corresponding transition matrix
●
-1

Constant Velocity Model
●
= φ i - 1 C X)
φT +
i − 1 i -1 C yi − 1
i (− )
]

1 UP Local
0
III- Kalman Filter Technique (2/2)
1- Predict State Vector
X̂i ( −) = φ i - 1 X̂i - 1( +)
2- Compute Covariance matrix of prediction
C li + H i C x̂ i(-) H T
i
0
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Equations
[

EOS
∆ EOS ={R } ∆ WGS84
III- Kalman Filter Technique (1/2)
G i = C x̂ i(-) H T
i
  X − Xo 
 Y −Yo 
 Z −Zo 
Transforming the difference between position and their covariance
matrix from WGS84 to Engineering Object coordinate System (EOS)
movements and noise
3- Compute predicted residuals
vˆi (−) = bi − Hi Xˆ i (−)
4- Compute gain matrix
0
 cos α sin α 0  E 
= - sinα cos α 0  N 
Along deformation axis e
Perpendicular axis n 
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
C Xˆ
Cos(λ )
−Sin(λ ) Sin(ϕ ) Cos(ϕ )

Sin(λ ) Cos(ϕ ) Sin(ϕ ) 
Transformation from Local Coordinate System (E-N-UP) to
Engineering Object Coordinate System (EOS)
σ 2DD( r ,i ) = σ 2SD( r ) + σ 2SD( i )
one-way path between satellite and antenna σ i2 may be computed
as the product of 3 functions such as:
●
Transformation from WGS84 to Local Coordinate System (E-N-UP)
IV- Quality Control Process (2)
CUmulative SUM Control chart (CUSUM)
● CUSUM First proposed by Page [1954]
-µ )
C0 = 0
Sn = Sn−1 + ( X n - µ ) / σ
S0 = 0
C n = C n −1 + ( X
n
● CUSUM Statistical Control Limits
● Control chart
S i+ = max[ 0 , yi − K + S i+−1 ]
Si−
= max[0 ,
- yi − K + Si−−1
]
x − µ0
yi = i
σ
● Chosen reference or allowance value (K),
decision interval (H) and Average Run Length (ARL)
K = 0.5 for ARL = 500
H = 4.389.
Decrease the false alarm rate by pushing the in-control ARL out
to1000 readings, then the required H would become 5.071
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Initial Test of the system(1/2)
● Location:
Initial Test of the system(2/2)
Time Series of Day71
UCL main library roof
- 9 .1 3
Zone I
Z o n e II
Z o n e II I
- 9 .1 4
Along deformation axis (m)
● Equipment: UCL main library roof
Two GPS Leica 500 - AT502
Base line length 81.7 m
● Period of observations:
Days 69, 70, and 71, of 2003
Nearly three hours /day
M e a s u r e d C o o rd in a t e s
- 9 .1 6
- 9 .1 7
RMS = ± 3.6 mm
- 9 .1 8
- 9 .1 9
3 00 60 0
● Observation rate
1 Hz, mask angle 7.5 degree
T r u e C o o r d i n a te s
- 9 .1 5
30 240 0
3 042 00
3 06 000
30 780 0
G P S T im e (S ec ) G P S W e ek # 1 20 9
30 96 00
0.030
Along deformation axis (m)
Zone I
● Data Collection scenario :
Days 69 and 70 no movement
Day 71 movement had been
introduced with certain pattern
Zone II
Zone III
0.020
0.010
0.000
-0.010
Filtered coordinates
-0.020
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
RMS = ± 1.6 mm
True Coordinates
Delay of nearly 3 min
-0.030
300600
302400
304200
306000
307800
GPS Time (Sec) GPS Week #1209
309600
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Monitoring Pacoima Dam with GPS (1/2)
Monitoring Pacoima Dam with GPS(1 /2 )
Water Level and Studying Cases
PACOIMA DAM
1902
Location : California - USA
Uses : Flood Control- water
conservations
1900
1898
Water Surface Elevation (Feet)
1896
GPS Operator : SCIGN
DAM1
Baseline Length =103.7 m
DAM2
1894
1892
1890
1888
Case Study II
1886
Case Study I
1884
1882
Receiver Ashtech ZXII3 and Choke
ring antenna
1880
1878
15-Feb
16-Feb
0.00
-0.01
0.02
0.01
0.00
-0.01
-0.02
-0.02
-0.03
172800
-0.03
432000
Day 18-Feb
518400
Day 16-Feb
345600
Day 17-Feb
432000
-103.77
-103.78
-103.78
-103.79
Day 16- Feb
345600
Day 17- Feb
432000
Day 18- Feb
GPS Time (S ec)
-103.80
518400
172800
-0.68
-0.68
-0.69
-0.69
-0.70
-0.70
-0.71
-0.71
-0.72
-0.73
-0.74
-0.75
Day 16- Feb
345600
Day 17- Feb
432000
-0.78
Day 15-Feb
259200
Day 16-Feb
345600
Day 17-Feb
26-Feb
27-Feb
28-Feb
29-Feb
432000
Day 18-Feb
518400
518400
GPS Time (Sec)
-0.72
-0.73
-0.74
-0.75
-0.78
172800
2.5 minutes MAF
259200
259200
Day 16-Feb
345600
Day 17-Feb
432000
345600
Day17-Feb
432000
Day 18-Feb
Day18-Feb
518400
0.02
0.01
0.00
-0.01
-0.02
-0.03
172800
Day 15-Feb
259200
Day 16-Feb
345600
Day 17-Feb
432000
Day 18-Feb
0.01
0.00
-0.01
-0.02
Sidereal-daycorrections
2.5minutesMAF
-0.03
-0.04
Day15- Feb
259200
Day16- Feb
345600
Day17- Feb
432000
Day18-Feb
GPSTime (Sec)
518400
0.03
0.02
0.01
0.00
-0.01
-0.02
-0.03
-0.04
172800
0.05
0.05
0.04
0.04
0.03
0.03
0.02
0.01
0.00
-0.01
-0.02
-0.03
Sidereal-daycorrections
Day 15-Feb
259200
Day 16-Feb
345600
2.5 minutes MAF
Day17-Feb
432000
Day 18-Feb
GPSTime (Sec)
518400
518400
GPS Time (sec)
0.04
0.02
-0.05
172800
Day 15-Feb
Day16-Feb
0.03
-0.04
GPS Time (Sec)
From Pharaohs to Geoinformatics : FIG Working
Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005GPS Time (Sec)
0.03
-0.04
Day15-Feb
Day 18- Feb
-0.77
-0.77
172800
259200
-0.76
-0.76
Sidereal-daycorrections
-0.03
172800
Day 15- Feb
Deformation in UPdirection (m)
UP coordinates (m)
-103.79
259200
-0.02
0.04
deformation in perpendicular to deformation
axis (m)
Perpendicular to deformation axis (m
-103.77
UP coordinates (m)
perpendicular to deformation axis coordinates. (m)
-103.76
Day 15 - Feb
25-Feb
GPS Time (sec)
518400
GPS Time (sec)
-103.75
-103.76
-0.01
Day 18-Feb
-103.74
-103.75
0.00
172800
259200
-103.73
-103.74
0.01
-0.04
Day 15-Feb
GPS Time (sec)
-103.73
0.02
deformation in perpendicular to deformation
axis (m)
0.01
172800
24-Feb
Kalman filter
0.03
Deformation in UPdirection (m)
0.02
-103.80
23-Feb
0.04
Deformation in along deformation axis (m)
Deformation in along deformation axis (m)
0.03
Day 17-Feb
22-Feb
0.04
0.03
Along deformation axis (m)
Along deformation axis coordinates (m)
0.04
345600
21-Feb
Analyses on the First Case Study ( 2 / 5 )
0.04
Day 16-Feb
20-Feb
Sidereal day correction
After outliers Detection
0.05
259200
19-Feb
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Analyses on the First Case Study (1 / 5 )
Raw Coordinates
Day 15-Feb
18-Feb
Local Time (Year 2000)
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
172800
17-Feb
518400
Day15- Feb
259200
Day 16- Feb
259200
Day 16-Feb
345600
Day 17- Feb
432000
Day 18- Feb
432000
Day 18-Feb
GPS Time (Sec)
518400
0.02
0.01
0.00
-0.01
-0.02
-0.03
-0.04
-0.05
172800
Day 15-Feb
345600
Day 17-Feb
GPS Time (Sec)
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
518400
Analyses on the First Case Study ( 3 / 5 )
Analyses on the First Case Study ( 4 / 5 )
20
15
Tempreature (C)
To assess the precision of the system, we look at the
behaviour of the system from day to day, and also the factors
that cause these movements between the two days..
The system is said to be precise if it reacts in the same way
under similar conditions.
10
5
0
-5
A c tu a l te m p re a tu re
-10
1 728 00
D a y 1 5 -F e b
T e m p e ra tu r e d if fe r e n c e
D a y 1 6 -F e b
259 20 0
34 56 00
D a y 1 7 -F e b
D a y 1 8 -F e b
4 32 000
51 840 0
G PS T im e (Sec )
The two possible factors that can cause dam deformation
are: an increase in water level in the dam reservoir or change
in temperature.
The average of the differences in temperatures between two
successive days are in order of 2o C. The effect is negligible .
There is no cause for any movements to occur between any two
successive days.
The difference in water levels between two successive days
is in order of 0.1 feet (i.e. ≈ 3.4 cm). This has no impact on the
dam deformation.
The question now is: does the output from the system reflect this
reality?
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Analyses on the Second Case Study ( 1 / 5)
Analyses on the First Case Study ( 5/ 5)
Raw data from 10 days
Maximum µ
(mm)
0.6
Maximum σ
(mm)
± 13.75
0.0
± 0.1
0 .0 5
The results show no deformation is detected with high precision
in real-time. The standard deviations of the time series are in
order of ±0.1 mm
Along deformation axis coordinates (m)
Type of time series and
analysis
Raw Coordinates
Sidereal-day corrections +
Kalman filter (real-time)
Along deformation axis direction
0 .0 4
0 .0 3
0 .0 2
0 .0 1
0 .0 0
-0 .0 1
-0 .0 2
D a y 1 9 - F eb
-0 .0 3
5184 00
6 0480 0
D ay 2 8 -F e b
69 1200
7776 00
8 6400 0
95 0400
10 3680 0
112 3200
12096 00
12 9600 0
138 2400
G P S T im e (s ec )
0 .0 0 1 5
Fourier analysis
Thus, high precision can be obtained by applying the proposed
system in such deformation monitoring applications.
0 .0 0 1 0
Amplitude (m)
The behaviour of the dam from day to day is consistent with
the Geo-technical measurements within ±0.04 mm.
0 .0 0 0 5
0 .0 0 0 0
1 .0 E -0 6
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Analyses on the Second Case Study ( 2 / 5 )
1 .0 E - 0 2
0.020
0 .0 0
-0 .0 1
-0 .0 2
-0 .0 3
D ay 19-F eb
-0 .0 4
5 18400
6048 00
1 .0 E -0 3
Along deformation axis direction
0 .0 1
D a y 2 8 -F e b
69 1200
77760 0
864 000
9 5040 0
10 3680 0
1123 200
12 09600
12960 00
138 2400
G P S T i m e ( se c )
If the data shown is filtered, the result will only be the
deformation that occurred between the two successive days.
To overcome this:
The deformations are considered to be zero on the first day
For any recent epoch, the computed deformation values are added
to those in the file. The process is carried out for the next epoch,
and so on till the end.
In other words, the output will be an accumulated time series of
the filtered deformation.
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Deformation at along deformation axis (m)
Deformation at along deformation axis (m)
0 .0 2
1 .0 E - 0 4
Analyses on the Second Case Study ( 3 / 5 )
After Kalman Filter
After sidereal day corrections
0 .0 4
0 .0 3
1 .0 E - 0 5
r e q u e n c2005
y (H z)
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. AprilF 16-21,
0.016
Along deformation axis direction
0.012
0.008
0.004
0.000
-0.004
-0.008
-0.012
-0.016
Day 19-Feb
-0.020
518400
604800
Day 28-Feb
691200
777600
864000
950400
1036800
1123200
1209600
1296000
1382400
GPS Time (sec)
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Analyses on the Second Case Study ( 4 / 5 )
Analyses on the Second Case Study (5 / 5 )
Validation of the System
Deformation Consistency with the Water Surface Elevation
Validation of GPSEM
Deformation consistency with water level
There is a strong correlation (≈ 98%) between the
deformation occurred along principle deformation axis
and the water surface level in the reservoir
LEICA SKI-Pro GPS software is used to process the data for
whole 24-hour to overcome the problem of multipath error.
The output is the three coordinates in WGS84. These coordinates
are transformed to bring these coordinates in the same frame
0 .0 2 0
Using Leica Ski-Pro GPS software
Deformations at along deformation axis (m)
0 .0 1 6
0 .0 1 2
0 .0 0 8
0 .0 0 4
0 .0 0 0
-0 .0 0 4
RMS = ± 0.43 mm
-0 .0 0 8
-0 .0 1 2
-0 .0 1 6
D ay 1 9 -F e b
D a y 28 - Fe b
-0 .0 2 0
51 8 4 0 0
604800
6 9 1 2 00
7 77 6 0 0
864000
9 5 0 4 00
1 0 3 68 0 0
1 12 3 2 0 0
1209600
1 2 9 60 0 0
1 38 2 4 0 0
G P S T i m e ( Se c )
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Conclusions (1 / 3 )
GPS overcomes all conventional methods drawbacks, but it
has its own. These drawbacks are low accuracy in real-time
and high cost of the equipment
The phase Multipath error is the dominant error that
limits the implementation of GPS in real-time EM
applications
The GPS phase Multipath error is highly dependent on
antenna reflectors, satellite relative geometry, and almost
repeats itself every sidereal day
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Conclusions (2 / 3 )
It detects the movements occurring in engineering objects
by applying the sidereal-day correction technique for phase
GPS multipath errors, a Kalman filter and a Cumulative
SUM (CUSUM) control chart.
Based on the results obtained from the investigations
carried out on Pacoima Dam, The system shows high
precision for detecting deformation of a real continuous
monitoring GPS project.
A methodology for GPS Engineering Monitoring using GPS
(GPSEM) has been developed and implemented
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Conclusions (3 /3)
There is a strong correlation (≈ 98%) between the
deformation that occurred along the principle deformation axis
computed from GPSEM and changes in the water level in the
reservoir.
The dam’s deformations computed from GPSEM in real
time are consistent with those computed from commercial
software (i.e. SKI-Pro) using 24-hour solutions. The RMS of
the differences between the two solutions in the direction of
deformation is ± 0.43 mm.
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
Recommendations and Suggestions for Further
Work
The system and software are needed to be tested under
operation circumstance on real engineering object ( like
high dam on Aswan).
More investigations are needed to study the behaviour of
the technique under high amplitude and frequency
deformation that may occur in long-span bridges for
example.
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005
THANK YOU
From Pharaohs to Geoinformatics : FIG Working Week 2005 and GSDI-8 Cairo, Egypt. April 16-21, 2005