GPS in Crustal Deformation

GPS in Crustal Deformation


Indian Institute of Geomagnetism, Colaba (P.O)
Email: [email protected]

The Global Positioning System (GPS) is a space based navigation system, consisting of a constellation of 24 satellites, in six orbital planes with 55° inclination to the equator. The satellites are placed at a height of about 20,200 km with 12 hours orbital period and operated by the United States Department of Defense (DOD) for accurate determination of position, velocity and time. All the GPS satellites are controlled by system tracking stations, ground antennae and the master control station. 

In each satellite two rubidium and two cesium atomic clocks with stability 1013 to 1014 are used to derive the fundamental frequency fo = 10.23 MHz. The GPS signals are transmitted at two frequencies, designated L1 (154 fo = 1575.42 MHz) and L2 (120 fo =1227.6 MHz) which are derived from the fundamental frequency (fo). Two codes are used, one of which is called C/A (coarse acquisition code, fo (10) and the other is called P (precise, fo) code. As the rate of P code is 10 times the rate of C/A code, its precision is 10 times better than C/A code. The L1 and L2 are modulated by Pseudo Random Noise (PRN) code, (each satellite is identified by this code) and transmitted after biphase modulation with the carrier.

The distance to GPS satellite is estimated by measuring the time a radio signal takes to reach us from the satellite. This is accomplished by cross -correlation of pseudo-random code generated by the satellite and the receiver. The distances from receiver to satellite measured in this way are called code pseudo ranges. Minimum four satellites are required for estimating the coordinates of a point on the Earth’s surface. The position accuracy that can be estimated this way depends on our ability to account for various error sources (Reddy, 2001). The textbooks, such as Seeber (1993, p. 209-348), Hofmann et al. (1994), Leick (1995), Parkinson and Spilker (1996), Kaplan (1996) provide very good reference on this subject. 

While the use of the GPS is extensive in defense, navigation and surveying applications, it is being used in geo-science, ionospheric & atmospheric studies, global climate changes, observing polar motion & earth rotation rate, mapping the gravity field, detecting seismo ionospheirc effects, transport and communications, environment management, for accurate time and frequency etc.

At present we are able to achieve much better accuracies due to processing techniques which circumvent the purposeful degradation of the GPS signals. This motivated earth scientists to use GPS for monitoring the slow and relentless crustal deformation by employing a technique called carrier tracking which allows to determine baseline length within a few millimeters. The methodology is that the changes in position coordinates and baseline lengths in three orthogonal directions computed with GPS data during successive visits will enable us to assess the crustal deformation. Changes in deformational rates have intrinsic value in understanding the physics of the earthquake processes. 

Crustal deformation studies have received new impetus all over the world with the full complement of satellites for adequate coverage, availability of comparatively low-cost receivers, sophisticated post processing softwares and international cooperation through International GPS Services for Geodynamics (IGS). In many countries the receivers are used permanently in a network mode with data telemetered and processed continuously to have upgraded baseline vectors regularly. For the past few years, regional GPS networks designed mainly to monitor strain for earthquake research and forecasting have been operated in many countries all over the world and have proved useful in detecting the crustal displacements. 

GPS data collection and analysis
Exposed bed rocks or well settled concrete pillars were chosen with unobstructed view of the sky and with non reflective environment as GPS sites. The monuments which were made on vast expanses of bed rocks are considered to have least site instabilities. Dual frequency GPS receivers were used in re-occupation mode in collection GPS data. Generally, every year the data has been collected during winter when the humidity is very low to minimize the effects of troposphere. The sampling interval and elevation were fixed at 30 sec and 15o respectively throughout the survey.

The GPS data were organized into 24 hours segments covering a UTC day to facilitate the combination of data from some of the surrounding IGS sites; IISC, DGAR, BAHR, KIT3, LHAS, YAR1, KERG to constrain the site co-ordinates. Then the data were processed using the GAMIT software developed at MIT and SIO (King and Bock, 1991) to produce estimates and an associated covariance matrix of station positions for each session with loose constrains on the parameters. To get a combined solution (site positions and velocities), all such covariance matrices are input to GLOBK which is a Kalman filter. The basic algorithms and a description of this technique are given in Herring et al. (1990) and its application to GPS data is in Feigl et al. (1993). By introducing global h-files, we have obtained coordinates and velocity vectors at each site in the ITRF96 reference frame (Boucher et al., 1998. The horizontal components of these velocity vectors are further used to estimate the horizontal strain field by Least-Squares Prediction (LSP) method (Reddy et al., 2000).

Some case studies

Deccan trap region in Western Maharashtra 
The Deccan trap region of western India generally considered as a seismically stable continental region (SCR), has been experiencing significant level of seismicity in the past two decades including the devastating Latur earthquake ( M 6.3) of September 29, 1993. This occurrence in the Central Indian shield has led to serious introspection among geo-scientists and motivated the Department of Science and Technology, Government of India, to sponsor the programme ‘Seismological upgradation and related studies in peninsular Shield’. Under this programme, to estimate the crustal deformation, a 73 site GPS geodetic network, distributed in an area of 200 x 350 km2 (shaded region in the inset of Fig. 1) has been designed. The first campaign started during January, 1995. The results presented here are based on GPS data collected for 3 campaigns during 1997-99 at 21 sites.

Velocity vectors ( in ITRF96) obtained from 3 GPS campaigns during 1997-99, indicate that the magnitude of the horizontal velocity of Deccan trap region is in the range of 40-60 mm/yr with an average of 51 mm/yr in N47o E (Fig.1). The estimated dilatational strain (Fig.2 ) is about 0.4 micro strain/yr in average. 
An extensional strain regime is observed along the west coast and south of Koyna and Warna reservoirs, transcending into a region of compressive strain towards the interior of the shield area. The extensional strain regime coincides with the West Coast Geothermal Province and intersecting fault system south of Koyna-Warna reservoirs. And the compressional strain regime could possibly be correlated with the India- Eurasia collision forces in the NS to NNE-SSW direction. Reddy et al.(2000) gives more details of this study.

Bhuj region in Gujarat
The Bhuj earthquake on 26th January 2001 with magnitude Mw 7.7 (Ms 8) is considered most devastating in last 50 years. About 25,000 people died and 400, 000 houses were destroyed. This region had experienced many high magnitude and damaging earthquakes in the past too. The earthquake during 1819 in the northwestern part of the Great Rann of Kutch and the Anjar earthquake of 1956 are two major earthquakes. Apart from these major earthquakes, the region has experienced several earthquakes in the magnitude range of 4-5. Considering the geo-tectonics and seismic history of the, occurrence of this Bhuj earthquake, though not predicted, it is not surprising. For studying the seismo-tectonics of the Bhuj earthquake affected region, many geophysical investigations are being carried out as a part of DST sponsored project. GPS measurements for monitoring post seismic deformation in this region considered one of the important studies. This zone demarcated as active zone V and extends approximately 250 km in east-west direction and 150 Km in the north-south direction. 

Fig.3 shows GPS network of 14 sites. Amongst these sites, 11 sites are located in the Kutch region. Both Trimble and Leica GPS receivers were used in GPS data collection. Fig.4 shows changes in East-West, North-south and vertical components of the baseline Lodai-Ratanpar ( which passes through the epicentral area) during 21-28 February,2001. During this time, though there were some after shock occurred, no displacement is seen in these components. However, a few days of GPS data analysis will not lead to any conclusion that there is no significant deformation taking place. Once enough data is collected at all the 14 sites, the data analysis will yield velocity distribution from which strain distribution can be estimated. This information can be supplemented by Interferometric Synthetic Aperture Radar (InSAR) (Reddy et al., 2000) results for better reliability on crustal deformation taking place the region

Fig. 4 East-West, North-South and vertical components of baseline Lodai-Ratanpar passing through the epicentral. The estimates are based on analysis of 6 hours data segments during 21-28 February 2001. The lower right corner graph shows the variations in baseline vector. Large errors are associated with the vertical component.

Niijima-Kozushima islands in Japan
Niijima – Kozushima islands in Izu peninsula, Japan started experiencing swarm type seismic activity starting from June 25, 2000. The Geographical Survey Institute (GSI), Japan has dense network of GPS sites around this region. 100 day of GPS data has been analyzed ( using GAMIT software ) at seven sites viz. Ohima, Toshima, Niiijima, Shikene, Kozushima, Miyake and Mikura. Fig. 5 and Fig.6 show the day to day variations of North-South, East-West and Up-down components starting from June 25, 2000 (corresponds to julian day 177) at Niijima the site which is close to the seismic activity and Ohima the site which is away from the activity. From the variations in the components of point position, it is clear that the variations are quite significant at Niijima and quite at Ohima (which is northern site much away from the activity). It is also clear from Fig.5 that, at Niijima there is about 70 mm displacement in the North-south component. This displacement occurred following an earthquake on July 1, 2000 (julian day 197). It is also clear that, it is not necessary that every earthquake cause displacement. The change in Niijima and Kozushima baseline (22 km) is about 80 cm and the linear strain is 37 microstrain. The large strain suggests that the source causing deformation is close to these Islands and the depth to the source is shallow.

Fig. 5 The North-South, East-West and Up-Down components of point position of the site Niijima island during 177-277 julian days of year 2000. The small open triangles on the lower most x-axis indicates the days on which the earthquakes occurred. Note the displacement of about 70 mm in North-South component following the earthquake on julian day 197.

Fig. 6 The North-South, East-West and Up-Down components of point position of the site ohima island during 177-277 julian days of year 2000. The small open triangles on the lower most x-axis indicates the days on which the earthquakes occurred. This site is away from the activity and no significant changes it the displacements.


  • Boucher, C., Z.Altamimi, and P. Sillard, Results and analysis of the ITRF96, IERS technical note 24, Observatory de Paris, 1998.
  • Feigl, K.L., R.W. King, T.A. Herring and M. Rothacher, Space geodetic measurement of crustal deformation in central and southern California, 1984-1992, J.Geophys. Res., 98, 21677-21712, 1993.
  • Gunter Seeber 1993 In: Satellite Geodesy Foundations, Methods, and Applications. Berlin, New York.
  • Herring, T.A., Davis, J.L. and Shapiro, Geodesy by radio interferometry: The application of Kalman filtering to the analysis of very long baseline interferometry data, J. Geophys. Res., 95, 12561-12581, 1990.
  • Hofmann-Wellenhop,B., Lichtenegger, H. and Collins, J. 1994 Global Positioning System: Theory and Practice. Springer-Verlag Wien, New York.
  • Kaplan, Elliott D.ed. 1996 Understanding GPS: Principles and Applications. Boston: Artech House Publishers.
  • King, R.W. and Y.Bock, Documentation for the MIT GPS analysis software, Mass. Inst. Of Technol., Cambridge. 1991.
  • Leick, A. 1995 GPS satellite surveying. John Wiley, New York.
  • Parkinson, B.W. and Spilder Jr.J.J.,(ed) 1996 Global Positioning System: Theory and Applications. Published by the American Institute of Aeronautics and Astronautics, Inc. Washington, v. I and II.
  • Reddy, C.D. and Teruyuki Kato, K.S. Rao and K.S. Rajan, 2000. Feasibiity of InSAR studies in Koyna region, western India. Spatial Information Technology (Remote sensing and Geographical Information Systems), BS Publications, Hyderabad, 243-253.
  • Reddy, C.D., Gamal El-Fiky, Teruyuki Kato, Seiichi Shimada and Vijay Kumar, 2000. Crustal strain field in the Deccan trap region, western India, derived from GPS measurements. Earth Planet Space, 52, 965-969.
  • Reddy,C.D. 2001. Error contributors and accuracy in GPS measurements. Special publication of GSI (in press).