An Appropriate GPS Technology for Landslide Monitoring at East-West Highway, Perak, Malaysia

Khamarrul A.R
Email: [email protected]

Wan Aziz W.A
Email: [email protected]

Faculty of Geoinformation Science & Engineering
University Technology Malaysia
Skudai 81310, Johore

1. Introduction
The assessment of landslide behavior is usually undertaken by means of monitoring scheme. Usually, the measurement of superficial displacement is the simplest way to observe the history of a landslide and to analyse the kinematics of the movement. In all cases, measurements have to be made efficiently in terms of time, manpower and budget.

In the past, a variety of surveying techniques have been used to detect the superficial movements of unstable area (Mikkelsen, 1996). For examples, tapes and wire devices have been used to measure changes in distance between points or crack walls (Gulla et al., 1988). Levels, theodolites, Electronic Distance Measurement (EDM), and total station measurements provide both the coordinates and changes of target, control points and landslide features (Ashkenazi et al., 1980). In addition, aerial or terrestrial photogrammetry provides point coordinates, contour maps and cross-section of the landslides. Photogrammetry compilation enables a quantitative analysis of the change in slope morphology and also the determination of the movement vectors (Oka, 1998). A comprehensive summary of the main methods and their precisions is shown in Table: 1.
Table:1 Overview of methods used in measuring surface displacement and their precision a (Mikkelsen, 1996)

a Note: 1ppm means one part per million or 1additional millimetre per kilometre of measured line.
Nowadays, the Global Positioning System (GPS) has been fully operational. The GPS equipment is more reliable, cheaper, faster, and easier to use compared to conventional instruments. New hardwares, field procedures and softwares have also been developed to assist users in data collection and processing purposes. Thus, the GPS equipment has become more progressive and used for a wide range of monitoring applications. This paper therefore highlights the performance of GPS technology in landslide monitoring encompassing a specific large-scale area.

2. The use of GPS Technology in Landslide Mapping
A landslide phenomenon is one of the terms to be used in describing the down-slope of soil, rock and organic material under the influence of gravity. Landslide studies can be organized into three phases, detection and classification of landslides; monitoring activity of existing landslides; analysis and prediction of slope failures in space (spatial distribution) and time (temporal distribution). The second stage in a landslide study is typically monitoring the movement of a landslide. This involves the comparison of landslide conditions over time, including the aerial extent of the landslide, the speed of movement and the change in the surface topography. Using either geotechnical methods or GPS technology the magnitude, direction of the slide, and the boundary of the landslide block can be determined.

The GPS is a radio navigation, timing and positioning system with a wide set of applications. By tracking the electromagnetic waves that are sent continuously to the earth, the system can obtain the 3D coordinates (F , l , h or x, y, z). The GPS system has become a valuable complement or extension to the conventional surveying methods (theodolite, tapes, EDM, total stations, etc). Field surveys are usually carried out within a frequently given period. The results are discontinuous over time, and related to the cumulative movements of the surface points.

The accuracy required for the measurement of landslide displacement should be, in many cases, at least in the order of centimeters (Josep A. Gili et. 2000). Therefore, the basic question that arises is whether the satellites orbiting 20,200 km above the earth can be used to measure coordinates or displacement of landmark points located at the ground surface with centimeter accuracy.

Measurement of landslide displacements can be undertaken by means of either static or kinematic method. The choice depends on the practical considerations : (i) the accessibility, (ii) number of points, (iii) precision and (iv) distance from point to point. Nowadays, the most productive methods i.e. modern positioning technique available for determining single points with precision of milimetres or centimeters is Rapid Static (RS) and Real Time Kinematic (RTK).

The RS method is a development of the classical static method, with improved algorithms that accelerates the ambiguity resolution procedures. For instance, measurement of one baseline with six and more satellite available is required for only a few minutes of data logging. This time increases to 15 and 20 min with five and four satellites, respectively. In this method, post-processing must be carried out. The data files from different receivers are merged in order to obtain the solution of the baselines between station points.

In the RTK method, the information of code and carrier phase observable received at both extremes of the baseline (base station and rover station) is merged to compute the precise position on the spot. The base receiver transmits a message containing its position, the pseudo-ranges measured to the rover through code correlation and phase measurements of the carrier received from the available satellites. Prior to obtaining the first results, it is necessary to spend a few minutes to initialize the system. The RTK calculates new positions from the old ones, through continuous tracking of the satellites in real time. In this procedure, therefore, the post-processing of the field data is not required. Any problem encountered with the equipment can be localized and solved immediately. It works quickly and gives results with precision at centimeter level for every second, even during movement. In RTK, corrections are transmitted to the rover via a local UHF data link; this transmission is quite directional. Therefore, unless repeaters are used, the RTK method needs an almost direct line of sight between base and rover. In general, this method also needs four or more satellite to work. Due to the continuous tracking of satellite, it is very sensitive to the loss of reception and to the quality of the signal. Even short interruptions will produce a loss of the initialization. To recover it, at least five satellites are needed.

Study of mass movement has been monitored using satellite as instrument to monitor monuments that have been installed in a stable or deform area. It gives the reference point on monument in each acquired observable. The concept of monitoring the landslide with GPS technology (Georg Gassner et., 2002, Hitoshi Kondo et., 1996) is shown in Figure:1.

Figure: 1 – Monitoring scheme; reference station (R,K) in stable area; rover station in the deformation area.

3. The Experiments
The performance of the GPS technique in measuring ground displacements has been tested in the active landslide phenomena. This landslide area is located at the East-West Highway, namely km22 and km26, Eastern Perak, Gerik, Malaysia.

Figure: 2 – Location of Gerik Landslide, East-West Highway, (a) km22 and (b) km26
The East-West Highway is the only route that connected the district of Gerik, Perak to Jeli, Kelantan- see Figure: 2. The Gerik landslide is one of eight locations, which has been declared and identified by Malaysian government as a high-risk landslide activity area. The National Landslide Research Center (NASEC) has also classified this area as a high-risk landslide activity in the ROM format scale.

The GPS preliminary surveys for landslide study were made in February 2003. Eleven baselines ranging from 62m to 3200m in length were observed by using dual frequency GPS receivers, Trimble 4800. The static technique has been applied to set up the reference network in the landslide areas. Figure: 3 (a) and (b) shows the control network establishment and satellite-receiver geometry during the observation, respectively.

Figure: 3- (a) Control network for static observation; (b) Example of planning the geometry of available satellite in Gerik (Eastern Perak, Malaysia) on February 26, 2003.
The rapid static and RTK GPS positioning techniques were used in the landslide monitoring scheme. In our experiment, two control points have been established at each landslide areas, namely BM and GK 1 for km22 test site, and JL 1 and JL 2 for km26 test site. In order to form a complete deformation network, thirteen and fifteen monitoring stations have also been monumented at sites km22 and km26, respectively – see Figure: 4 (a) and (b). Figure 5 (a) and (b) show the rapid static observation at the landslide areas. The static and rapid static GPS data were carried out in post-processed mode using Trimble Geomatics Office (TGO) software while the GPS kinematic data were processed in real time mode.

Figure: 4 – Monitoring network for km22 (a) and km26 (b).

Figure: 5 (a) and (b) Rapid static GPS observation for km22 and km26, respectively.

4. Results and Analysis
In the main control network, the results of the static GPS observations for control points BM, GK1, JL1 and JL2 have been obtained . In this case, their corresponding coordinated were determined with respect to fixed stations P307 and P310 – see Table: 2.

It can be seen from Table: 2 that the standard errors in horizontal component (latitude and longitude) are less than 7mm . This result has shown that the GPS technique can be obtained with high accuracy . Similarly, the standard errors in vertical component for control points at km22 and km26 test sites is about 3.2cm and 6cm, respectively. Here, one may noticed that the accuracies were decreased proportionally with the increase in the distance from base stations; i.e. P307 and P310 are 33km, in length. Furthermore the locations for control point JL 1 and JL 2 are also surrounded by hilly areas and close to transmission lines.

Table: 2 – The Static GPS coordinate is summarized for control points

* Standard errors are given as positional errors
Results of rapid static survey for East-West Highways at km22 and km26 are summarized in Table: 3 and Table: 4, respectively.
Table: 3 Rapid Static coordinates of monitoring station for East-West Highway, Km22

Table: 4 – Rapid Static coordinates of monitoring station for East-West Highway, Km26
It can be seen from Table: 3 and Table: 4 that the accuracy of vertical component for km22 in horizontal components lies in between 2cm-20cm whereby the largest values occurred for the monitoring of points JL12 and JL13. This problem was due to the sky view environment and poor satellite-receiver geometry (DOP) at these stations during the observation.

The results of RTK observations is also given in Figure: 5(a) and 5(b) in local coordinate (RSO). From this figure, one may noticed that the derived coordinates at the selected monitoring points were fluctuated within centimeter level. The results show that the RTK technique can also be utilized in landslide monitoring scheme.

Figure:5 (a) Results of RTK survey for station; GK 4 and GK 3 at Km22 test site.

Figure:5 (b) Results of RTK survey for station; JL14 and JL15 at Km26 test site.

5. Conclusion
GPS is a very useful tool to be utilized for a wide range of scientific applications. This technology increases the accuracy, productivity, monitoring capability, rapidity and economy with respect to size of the study area and it is often better than classical geodetic survey techniques. This paper evaluates the appropriate GPS technique i.e Rapid Static and RTK for monitoring the landslide behaviour at two test areas, km22 and km26, East-West Highway. The results indicated that the GPS modern techniques is very reliable for landslide monitoring survey whereby deformation of superficial displacement will be determined with another epoch of GPS data collections (Martin Vermeer, 2002).

References

• Ashkenazi, V., Dodson, A.H., Sykes, R.M., Crane, S.A. (1980): ‘Remote Measurement of Ground Movement by Surveying Techniques’. Civil Eng. Survey. 5 (4), 15-22.
• Georg Gassner, Andreas Wieser, Fritz K. Brunner (2002): ‘GPS Software Development For Monitoring of Landslide’. Proc. FIG XXII, Deformation Measurement and Analysis II, Congress Washington, D.C. pp. 12.
• Gulla, G., Nicoletti, P.G., Sorriso-Valvo, M. (1988): ‘A Portable Device for Measuring Displacements Along Fractures, Proc. 5 th Int. Symp. On Landslide, Lausanne Vol.1, 423-426.
• Hitoshi Kondo, Akira Sugiyama and M.Elizabeth Cannon (1996): ‘Precise Carrier Phase GPS and Its Application To Real-Time Landslide Detection’. IEEE TENCON- Digital Signal Processing Application, Vol.2, Page(s): 906-911.
• Josep A. Gili, Jordi Corominas, Joan Rius (2000): ‘Using Global Positioning System Techniques In Landslide Monitoring’. Engineering Geology 55 (2000) 167-192, Barcelona, Sepanyol.
• Martin Vermeer (2002): ‘Review of The GPS Deformation Monitoring Studies’. Report of Radiation and Nuclear Safety Authority (STUK-YTO-TR)186 Helsinki, Finland.
• Mikkelsen, P.E. (1996): ‘Field Instrumentation. In: Turner, A.K., Schuster, R.L. (Eds.), Landslides Investigation and Mitigation, TRB Special Report 247. National Academy Press, Washington, DC, pp.278-316, Chapter 11.
• Oka, N. (1998): ‘Application of Photogrammetry to the Field Observation of Failed Slopes. Eng.Geol. 50, 85-100.