2D MODELING AND INVERSION OF MAGNETOTELLURIC DATA ACQUIRED IN GARHWAL HIMALAYA

Abstract

The Magnetotelluric (MT) method is used to estimate the conductivity distribution in earth interior by measuring natural time varying orthogonal components of horizontal electric and magnetic fields on the earth's surface. The natural electromagnetic field is generated by the lightning activity in the atmosphere (> 1 Hz), also referred to as atmospherics, and also by the complex interaction of solar wind with earth's magnetosphere (< 1Hz) also referred to as micropulsation. Since the inception of MT method by Cagniard (1953) for simple layered structure, the method has gone through different phases of development and now it can be used to determine geoelectrical image of a complex geological structure. By selecting appropriate frequency range for recording the orthogonal horizontal electric and magnetic field components, the MT method is capable of delineating crustal structure at shallow (<5 km) (Gokam et al., 2002a; Patro et al., 2005; Harinarayana et al., 2006), intermediate (<50 km) (Adam, 1997; Lemonnier et al., 1999; Wu et al., 2002; Pous et al., 2004; Sharma et al., 2005) and deeper levels (>50 km) (Tezkan, 1994a; Ogawa et al., 1996; Gokam et al., 2002b; Ritter, O., 2003; Spratt et al., 2005; Patro et al., 2005; Bologna et al., 2006; Rao et al., 2007). Various authors have used MT method for crustal and upper mental studies in Himalaya (Singh et al., 1992; Gupta et al., 1994; Chen et al., 1996; Park and Mackie, 1997; Lemonnier et al., 1999; Gokarn et al., 2002b; Bai and Meju, 2003; Unsworth et al., 2005; Spratt et al., 2005; Harinarayana et al., 2006; Arora et al., 2007; etc). The theoretical aspects, modeling and inversion of MT data, have been worked out by many authors (Oldenburg, 1979; Weidelt, 1975; Constable et al., 1987; Weaver, 1994; Weidelt and Kaikkonen, 1994; Rodi and Mackie, 2001; Farquharson and Oldenburg, 2004; etc). Present work is devoted to the application of Broad Band Magnetotelluric (BBMT) a method, in the period range from .001 s to 1000 s, to delineate geoelectrical structure in Garhwal Himalaya corridor along the Roorkee-Gangotri profile passing through major Himalayan thrusts: Himalayan Frontal Thrust (HFT), Main Boundary Thrust (MBT) and Main Central Thrust (MCT). During the four field seasons, spread over the period from January 2004 to June 2006, the BBMT data were recorded at the 44 sites along a profile from Roorkee to Gangotri in Garhwal Himalaya corridor. We employed the Metronix ADU06 control unit and the electric and magnetic field sensors. In view of the complex, inaccessible terrain and noise conditions, the sites interval varies from 2 to 10 km. The data were processed to obtain the apparent resistivity and phase curves from the recorded time series. The strike code (Groom and Bailey, 1989; McNeice and Jones, 2001) has been used to estimate telluric distortion, geoelectric strike, and regional 2D impedance from the observed impedance. To arrive at a consistent geoelectric strike from the observed data, the strike code has been used in three different modes: viz (a) Single-site Single-frequency (SS), (b) Single-site Multifrequency (SM) and Multi-site Multi-frequency (MM). The observed responses were rotated along and perpendicular to the geoelectric strike to derive the TE- and TMmode responses. A methodology has also been developed for applying terrain correction to the recorded MT responses. This is based on the Finite Difference Method (FDM) as used in the 2D forward modeling code EM2INV (Rastogi, 1997). The terrain correction results have been verified over the synthetic models available in literature. It has been observed that the terrain correction in the MT data recorded from Garhwal Himalayan corridor is small (below the noise level) and hence not needed for our data set. Finally, 1D and 2D inversion of TE- and TM-mode data has been carried out individually and also jointly to obtain a geoelectrical model along the in profile in Garhwal Himalaya. The geoelectrical model is correlated with the seismicity of the region. This thesis is an effort to systematically present the entire work in the following six chapters. Chapter 1 discusses the basic principles of magnetotelluric method, the relevant differential equations and the constitutive equations. The definition of MT impedance, response function, impedance distortion and Groom-Bailey (1989) decomposition are discussed. The outlines of basic theory of forward and inverse modeling relevant to magnetotellurics are also presented. Chapter 2 presents the basic geological framework of the Himalayan region with special reference to the Garhwal Himalaya corridor. The basic tectonic setting of the region and the characteristics of prominent Himalayan Thrusts in the region are discussed in the light of available geological and geophysical literature. Chapter 3 discusses the details of Broad Band Magnetotelluric (BBMT) data acquisition. The field procedure and the acquisition parameters used in data recording are described. Five component MT system deployment and the notions of recording bands, recording time etc. are discussed. The locations of MT sites in Garhwal Himalaya corridor are shown in the geological map of the study area. Single site and remote reference modes of data acquisition are discussed. The characteristics of the recorded time series and of the sources of noise are also discussed. Chapter 4 discusses the details of time series processing. The characteristics of the response curves from Indo-Gangetic Plain (IGP), Lesser Himalaya (LH) and Higher Himalaya (HH) regions are shown. The impedance distortion and decomposition analysis have been carried out. Swift skew and Bahr Phase Sensitive Skew parameters are estimated to analyze the dimensionality of data. Higher values IV of Swift skew (> 0.2) and of Bahr phase sensitive skew (> 0.3) are obtained at six sites in Higher Himalaya region. Telluric distortions due to shallow three-dimensional structures are determined by studying the behavior of skew, and decomposition parameters (shear and twist). High skew and higher values of decomposition parameters indicate presence of strong three-dimensional effects at the six data sites in Higher Himalayan region. Groom-Bailey decomposition has been used to determine regional 2D impedance and geoelectric strike. These parameters have been estimated using the strike code version 5.0 (McNeice and Jones, 2001) in three different modes: SS, SM and MM. Average geoelectric strikes in the IGP, LH and HH regions, obtained using the SS mode, are N15°E, N23°E and N21°E respectively. The MM mode strike values in the three regions are N11°E, N24°E and N26°E. In the SM mode, the average geoelectric strike is N45°E for frequency > 1 Hz and N15°E for frequency < 1Hz. As the estimated geoelectric strike has 90° ambiguity, the average geoelectric strike for the entire profile (Roorkee to Gangotri) line is estimated to be N20°E or N70°W. Using the geological inputs, the geoelectric strike of N70°W is used for further interpretation of MT data. Finally, it has been demonstrated that for the data recorded in Garhwal Himalayan Corridor, the terrain corrections are very small and are not needed for the entire data set. Chapter 5 discusses one-dimensional (1D) and two-dimensional (2D) inversions of MT responses along the profile. Since the TE-mode response is less affected by lateral resistivity discontinuity, 1D inversion of TE-mode data was used to generate the initial model for 2D inversion. 1D inversion has been performed using standard least square, Occam's (Constable, et al., 1987) and Straightforward Inversion Scheme (SIS). The least square and Occam's codes are adopted from the Geosystem's WinGLink interpretation software, while the SIS code was developed by Gupta, et al. (1996). Subsequently, we carried out 2D smooth inversions of individual TE, TM, joint TE, TM modes and TE, TM with tipper data. Several numerical experiments were performed to define the inversion parameters: error floor and smoothness. Inversion was carried out using several initial guess models, with different homogenous half space resistivity values and with initial model based on 1D inversion. To account for any possible static effect in data, a large error floor was assigned to apparent resistivity. WinGLink option for static shift correction was also used. A geoelectrical model along the profile line is presented. Sensitivity tests have been done to validate the broad features of the model. The final geoelectrical model is presented along with the local seismicity of the area. It has been observed that the local seismicity is concentrated near the mid crustal conducting zone near MCT in the Higher Himalaya. Consistent near surface conductive zones in IGP and LH regions represent the loose sediments transported from Himalayan region. Mid crustal conductor (<10Qm)at a depth of about 12 km extending to a depth of more than 25 km is a major feature of the geoelectrical model. Sensitivity tests indicate that the major features of the mid crustal conductor are required by the MTdata. Summary and conclusions are presented in chapter 6 along with the recommendations for future work. The conducting zone also appears to be related with the strain accumulation zone in Himalayan region for future earthquakes (Gahalautand Chander, 1997a, 1997b).