ACCURACY AND
EFFICIENCY OF BATHYMETRIC SURVEY TO BE IMPROVED WITH RTK TECHNIQUE
Dr. Nguyen Ngoc Lau
(1), Dr. Truong Ngoc Tuong (2),
Eng. Nguyen Tan Son (3) and
Eng. Nguyen Quang
(3)
(1)
Geodesy and GIS Department
– HCMC University of Technology
(2)
Ports and Waterways – HCMC
University of Technology
(3)
Geomatics Department 
Portcoast Consultant Corporation
(The first
Scientific Conference of Geodesy and Mapping and Land Management in Hanoi,
Dec.2004)
Abstract:
Measuring water level near shore to determine sea/river – floor height in
hydrographic surveying has provided low precision in limited areas. In this
paper, we present an application of RTK GPS technique at PortCoast Consultant to
experimentally survey a part of Soai Rap river, Ho Chi Minh City, the results
showed that the use of traditional method can cause an error in height about
three times bigger and much less efficiency than RTK technique.
1.
Introduction:
Measuring bathymetry is
formerly a complicated procedure and it could give us low accuracy and
efficiency. This procedure has been improved so much thanks to Global
Positioning System.
Figure 1. DGPS
Technology
One of GPS techniques used
widely is DGPS (Differential Global Positioning System). The technique can be
described briefly in Figure 1. The first receiver located at a fix location
onshore whose coordinate is known clearly. It is called base station or
reference station transmitting distance correction into rover station via radio
links. The second one located on a boat called rover station will receive
distance correction from base station and adjust distance measurements from it
to satellite to process and supply coordinate and height of its antenna in real
time. Measurements used to process are pseudorange measurements C/Acode or
Pcode with accuracy 3m and 0.3m respectively. In case of dynamic measuring, the
measurements can get accuracy of coordinate positioning form 0.5m to 3m.
Accuracy of height determined by GPS technology is 3 times smaller than that of
coordinate positioning so that it does not meet requirements and GPS technology
is not used for height determination. Depth of seabed is determined in measuring
water level method as the following figure:
Figure 2.
Measuring water level with leveling machine and target.
Above figure height of water level near river
bank is calculated by formula:
h_{w} = h_{0} – b + a –
c (1)
The principle of
determination of riverbed height in measuring water level assumes that water
level at points near river bank equals to that at location of a boat. At that
time height of riverbed is calculated by:
h = h_{w}
– d (2)
In this report, we would like to present the
proceed of test measuring with RTK technology on Soai Rap river in Ho Chi Minh
City at the same time with the proceed of water level measuring. Analysis
results will give information about accuracy of depth determination with water
level method in the area. Thanks to the experiment, advantage of accuracy and
efficiency of RTK technology for bathymetric survey is also emphasized.
2.
What is RTK Technique?
RTK is
the abbreviation of
RealTime Kinematic (i.e. dynamic survey
in real time). RTK technology is almost similar to DGPS technology. However, in
the case of RTK technology, base station will transmit phase measured value to
rover station. Rover station establishes phase measured value in the form of
double difference for processing. Therefore, the technology is also called DGPS
technology using phasemeasured value. Because phasemeasured value has the
accuracy in a few millimeters so that the positioning accuracy can get from a
few centimeters to decimeters and that depends on to finding the integer of
ambiguity resolution number in phase measured value. If 100 percent of
multivalue parameters are solved, we will get the
fixsolution with the highest accuracy. If multivalue parameters can be solved
partially, we will get the floatsolution.
Solution of finding out multivalue parameters
only based on phase measured values at the measuring time is a big challenge to
GPS equipment including theirs hardware and software. Besides that depends on
distance between base station and rover station, the number of satellite
observed
simultaneously and velocity of the boat.
Figure 3  RTK method
(Real Time Kinematic)
Above figure describes principle of determining
the river floor height by using RTK method, where:
Dh:
different height between two antennas of the base station and the rover station,
determined by RTK with cm accuracy.
a: antenna height at the base
station, measured by tape with about 5mm accuracy.
h_{o}: The point’s height
compared with the reference ellipsoid (assume no error).
f: antenna height of the rover
station compared with water level at a boat position, measured by a tape with cm
accuracy.
d: riverbottom depth, measured by
an echo sounder.
h: riverbottom height, calculated
by formula:
h = h_{w} –
d
h_{w} is the instant
height of water level and h_{w} = h_{o} + a 
Dh
– f.
Apparently, RTK provides both
exact height and position of antenna at measuring time. It helps avoid
shortcomings of bathymetric process by measuring water level and enhances the
accuracy of bathymetric surveying process.
3.
Equipment and a sample of
measuring area:
GPS receiver used in this
case is dual frequency Topcon Legacy H for base station and Topcon Hiper for
rover station. Radio links with 38400 Baud frequency and RTCM format of
transmitted data. According to [4], accuracy with RTK technique is 10mm+1.5ppm
in coordinate positioning and 20mm+1.5ppm in height determination. It means that
if distance between base station and rover station is 3 kilometers, accuracy of
coordinate positioning and height determining is 14.5mm and 24.5mm respectively.
Topcon Legacy H Topcon
Hiper Odom Echotrac MKIII
Figure 4.
Hardware for bathymetric survey
Odom Echotrac MKIII dual frequency Echo sounder
is used with its accuracy:
m=0.01m+ 0.1% X depth (m) at the frequency 200
kHz.
Figure 5.
Depth measuring area and measuring line.
Mean square error of
determining level water is calculated by:
(m)
If maximum depth of river bed is
50m, mean square error of determining river bed depth is
(m)
If we use the equipment, accuracy of
determining water level at boat location and river bed height is about 27mm and
66mm respectively.
Topcon Hiper and Topcon
Legacy are connected to two ports RS232 of a laptop, where Hypack Max software,
specializing in bathymetric survey, is installed. This software enables users
predesign surveying lines on map in form of DXF [5]. In the operation process,
boat’s location shall be regularly updated on the base map. This allows the boat
to be operated to follow design lines with errors of few metres even in offshore
areas, in condition of undefinable directions and being influenced by outside
factors (Figure 5). Data of GPS locations in form of NMEA 0183 version 3.0 and
bathymetry shall be automatically saved in computers’ hardware.
The surveying area is a
bend of Soai Rap river – Ho Chi Minh City with maximum depth of approximately
30m. Determined location of base station is the cadastral bench mark, which is
performed with a triangular sign in Figure 5. Measuring the water level at this
location is implemented in interval of 15 minutes. Bathymetric lines are
designed horizontally along the river and in interval of 100m. The recording
interval of GPS receivers and Odom Echotrac MKIII shall be 2 seconds.
The transfer of coordinates
from WGS84 to local coordinate system is a complicated process, which shall not
be mentioned in the report. Therefore, WGS84 shall be used in the study to avoid
errors arisen from the transfer.
After the period of
approximately 45 minutes necessary for installing base station and rover station
onto the boat, the boat shall be operated following the design lines from
upstream to downstream areas. The maximum distance from base station to rover
station is about 4.5km. The number of GPS maintained during the survey shall not
frequently be less than 7. With the speed of the boat approximately 10km/h, it
only takes 6 hours to finish the survey.
4.
Transferring elevation from
ellipsoid into chart datum:
Elevation of
sea/riverbed is practically performed upon chart datum different from Ellipsoid
used in GPS, to which these 2 following solutions based on Hypack Max Software
can be applied:
·
In case of small
distance from the rover station to the base station, Ellipsoid plane and
standard plane of chart datum in the surveying area shall be considered as 2
parallel planes, of which the distance between them shall be assumed as the
difference of 2 levels at the base station (Ellipsoid – Chart datum). This
difference shall be utilized by Hypack Max to transfer elevation of points
automatically.
·
At the site, control
points of which Ellipsoid datum and chart datum have been available shall be
input into Hypack Max software before measurement. In the measuring phase,
difference of two levels at the point shall be calculated for adjustment, which
is considered as the measurement for the height at the points for adjustment.
This measurement can be applied at large areas or stretched areas, where two
Ellipsoids and standard plane of chart datum are not parallel.
To this experiment, the
first measurement shall be applied. However, the second measurement can be
applied because along the river there are base cadastral points, of which chart
datum has been determined by geometric measurement.
5.
Processing and analysis of
results:
The postprocessing of depth
measuring data is totally implemented by Hypack Max software. There is
approximately 8000 bathymetric points, 100% of which is root fix. This shows the
high reliability of RTK Technique. The height of water level is calculated with
formula (1) at base station and formula (4) at rover station as shown in the
following diagram:
Figure 6.
Height of water level at base station and rover station
According to the diagram,
water level is performed by “*” sign at
base station and “.” sign at rover station with RTK Technique in interval
of 2 seconds. Water level changes complicatedly in periods because the antenna
of rover station on boat at calm water area is influenced by impacts of wind,
tide or current, which is unavoidable in bathymetric survey. However, with thick
density of data (in interval of 2 seconds or more), calm water level at the
point can be determined by filtering method. An interfering filter has been used
to collect value in average interval of 5 minutes (equivalent to 150 data
points). The result is dark points “.”as in figure 6. These results subtracted
by values collected by water level measurement near shore shall obtain
differences as shown in the following table:
Time 
Difference
(cm) 
Time 
Difference
(cm) 
Time 
Difference
(cm) 
01:15
01:30
02:00
02:15
02:30 
23
27
23
19
18 
02:45
03:00
03:15
03:30
03:45 
20
18
18
18
17 
04:15
04:30
05:00
06:00
06:30 
15
15
12
8
10 
As performed in the above table,
differences of two water levels vary from 10 cm to 30cm and increase along with
increasing distance between base station and rover station (Figure 7), which
well suits the hydraulic regime at narrow river sections.
In such case, when replacing water level at boat with
that near shore, an average error 0.2m of height should be accepted. This value
is approximately 3.3 times different from depth measuring error (0.06m) and
hence, mean square error for determining the height of bottom shall be
Figure 7. Correlation
of level difference and distance of stations
6.
Conclusions and
Recommendations:
Measurement of water level
near shore was demonstrated to create errors up to 3 times different from those
of depth measurement. However, it is still the result of surveying on a small
river section. Errors could be bigger when measurement is implemented at seas,
where impacts of tide and wind are much bigger.
RTK technique is the new instrument that
shall provide more accurate and efficient results for bathymetric survey. With
the above instrument, it shall take much less time to postprocess
with Hypack Max software for bathymetric map and DEM.
As stated in chapter 2, accuracy and
reliability of RTK Technique mainly depends on the successful solution of
multivalue parameter in phase measured value, which is affected by different
factors such as distance between base station and rover station, quantity of
simultaneously monitoring satellites and moving speed of boat. RTK Technique has
fully satisfied proposed requirements of this study. And RTK Technique shall be
applied at complicated areas, namely coastal areas, in which the distance
between base station and rover station shall be up to kilometers.
Reference
[1]. Chris Rizos, 1996, Principles and Practice
of GPS Surveying, University of NewSouth Wales, Australia.
[2]. Department of the US Army, 2002,
Hydrographic Surveying.
[3]. Tran Minh Quang, 2000, Viet Nam National
University Ho Chi Minh City.
[4]. User guide of using Topcon Legacy H and
Topcon Hiper
[5]. User guide of using software Hypack Max.
