Crustal deformation studies through GPS measurements

B. Nagarajan
Deputy Director, Geodetic & Research Branch
Survey of India, Dehradun
[email protected]

Introduction
Continuous monitoring of crustal deformation can give an insight into the various phases of the earthquake cycle. This in turn provides the earth scientists with the knowledge required for forecasting earthquakes and provide warnings for impending seismic risk. Space geodetic techniques including Very Long Baseline Interferometry ( VLBI ) , Satellite Laser Ranging ( SLR ) and more recently Global Positioning System (GPS) have been able to provide the direct measurements required for studying complex deformation pattern across plate boundaries all over the world. The strength of GPS over VLBI, SLR and EDM is its ability to deliver three dimensional vector positions easily and inexpensively between sites that are not intervisible. Geodetic and Research Branch of Survey of India, has taken up post earthquake crustal deformation studies in Gujarat through GPS measurements. The results obtained when combined with Gravity and Geomagnetic measurements and High Precision levelling measurements provide an excellent scenario for better understanding of the physics of earthquake processes. This paper is restricted only to the analysis of GPS measurement results in the process of crustal deformation studies carried out in Gujarat after the devastating earthquake on January 26, 2001.

GPS Network
In order to have observations, immediately after the earthquake, which could shed some light into the horizontal movement of the crust, four GPS teams were deployed in first week of February,2001 to carryout GPS observations in Gujarat. Thirty old existing Geodetic Triangulation Stations, spread evenly throughout Gujarat were selected for observations. In addition, three new GPS stations were established to improve the strength of figure. All the thirty-three stations were knit into a network of simple triangles, braced quadrilaterals and central figures as shown at Annexure-I.

Procedure of Observations
After carrying out reconnaissance of the stations of starting figure, GPS observation campaign was started from 19 th February,2001 and completed by 8 th April, 2001. The observations were carried out in static mode, simultaneously at all the sites participating in the measurements for the particular session. Duration of observation sessions were of 24 hours from 05:30 hours I.S.T. to 05:30 hours I.S.T. next day. Recording interval for the data collection was kept at 30 seconds. Observation sessions for the campaign consisted of the following group of vectors

Data Processing
To check the quality of data collected, pre-processing was done in the field using SKI/Bernese Softwares. Final processing of the vectors has been done at the Head Quarter in Dehradun. The raw data was first converted into RINEX (Receiver Independent Exchange) format. IGS precise ephemerides were down-loaded through Internet. Post processing of the vectors was done through Bernese GPS Post Processing Scientific Software Version 4.2, using precise ephemerides. The vectors so derived were further converted into spheroidal arc distances and used for working out new co-ordinatesof the points by Variation of Co-ordinates Method.Old G.T. co-ordinates of SAGBARA H.S. and the G.T. azimuth at SAGBARA H.S. of PILVA H.S. were held fixed and the co-ordinates of remaining 32 points were derived. The fixed station SAGBARA H.S. is situated well away from the earthquake effected zone.

Results and Analysis
Differences in the Horizontal Position of 29 stations ( New Value – Old Value ) are tabulated below :

A Chart showing the horizontal differences in the co-ordinates of 29 points is attached at Annexure-II. Maximum shift is noticed at station Nos. 0095, 0096, 0140, 0141, 0144, 0145, 0147, 0156 and 0158. Remaining points show very moderate shift, taking into account the fact that the G.T. Co-ordinates are more than 140 years old.

The large differences in the old G.T. and new derived positions can not be attributed solely to the shift of station positions. The accuracy of G.T.Co-ordinates may be of the order of 10 ppm. Since the network extends to around 500 km in East-West direction, there could be an uncertainty as large as 5 m. The Maximummovement of 11.1 m in almost Southern direction has been noticed near Lakhpat. Other shift in the positions are in Easterly directions, in the areas lying South-South-East of the epicenter and in SE and SW directions in the areas situated E of the epicentre. The crust appears to have taken an anti-clockwise swing. Lakhpat situated North-West of the epicentre is a case in anomaly, registering Southerly shift as large as 11 m.

Conclusion
GPS observed baselines can yield sub-centimetre accuracies. The present comparison has been made between the classical technique i.e. G.T. Triangulation having limitations of 10 ppm accuracies and the GPS technique which is much more accurate. Although, certain broad indications of the movement of the crust have been noticed in the present study, it is most desirable to have a second set of GPS measurements after a gap of 2 years, with observation sessions having the same group of vectors, as observed during field season 2000-2001. Comparison between the two sets of vectors observed during 2000-2001 and say 2002-2003 will then be more reliable for critically analysing the differences and their relationship with the crustal deformations.