3D MAGNETOTELLURIC AND DCR INVESTIGATIONS AT SELECTED SITES IN UTTARAKHAND

Abstract

Electrical resistivity structures at four selected regions were deciphered from 3D inversion of Magnetotelluric (MT), Radio Magnetotelluric (RMT) and Direct Current Resistivity (DCR) data recorded in Uttarakhand, India. Earlier, the recorded MT and DCR data from the regions were interpreted assuming a 2D model of Earth. However, the geological structures, in general, are 3D in nature. Hence, the MT, RMT and DCR data, recorded by our group prior to the onset of this study along with the data recorded during the present work, are analyzed using 3D inversion techniques. The entire data comprises four different case studies, (i) MT data recorded in the Garhwal Himalayan Corridor (GHC) along Roorkee-Gangotri (RG) profile, (ii) MT data recorded in and around MCT zone in Chamoli region, (iii) DCR data recorded near a proposed bridge site in Tehri region and (iv) DCR and RMT data recorded in Saliyar Village near Roorkee. The first case study is an MT investigation in GHC where MT data were acquired at 39 sites on RG profile. AP3DMT, a MATLAB based code developed by our group, was used for 3D inversion of this MT dataset. Before performing this 3D inversion several 3D synthetic inversion experiments, on full impedance tensor data, were carried out to estimate the optimal values of important control parameters affecting the inversion results. The synthetic studies established that the off-profile resistivity structures within 20 km from the profile can be deciphered from profile data. The optimal values of some other inversion parameters (e.g. model grids, smoothing, regularization parameter etc.) were also inferred and then used for 3D inversion of the field data. A 3D geoelectric model was obtained after performing several 3D inversion experiments on RG profile MT data using different components of data independently and jointly. The data types used were vertical magnetic field transfer function (VTF), Phase Tensor (PT), full impedance tensor (Z), VTF + PT and VTF + Z. Based on sensitivities of Z and VTF data and on other related tests, the model obtained from the VTF + Z data was chosen as the final model for detailed interpretation. The robustness of important features in the inverted model were validated through sensitivity tests. The model defines the geometry of DHR beneath the IGP, SH and LH in Garhwal Himalayan region. The model also suggests that DHR is a fault bounded structure. The obtained 3D geoelectrical model reveals some new off-profile resistivity structures which are aligned transverse to the main iii Himalayan arc. The off-profile structures are interpreted as conductive (< 10 Wm) fluid saturated fractured zones bounding the highly resistive (> 1000 Wm) Delhi-Haridwar Ridge (DHR). The 3D resistivity model explains the thrust tectonics and flat ramp flat geometry of Main Himalayan Thrust (MHT). The model was also correlated with other geophysical models and with seismicity of the region. It has been observed that the earthquakes are generally located in the resistive zone of the crust with a few exceptions where they are near or in the conductive zone. In second case study, MT data recorded at 28 sites in and around MCT zone in Chamoli region of Uttarakhand, India were inverted using the code AP3DMT. The dimensionality and directionality analysis of the MT data revealed the dominant 3D nature of the impedance tensor. In first step, the dataset was inverted for 28 sites. However, due to some noisy sites data, 23 MT sites were selected for obtaining the final 3D model. The 23 sites were inverted using the full impedance tensor data. Experiments were done to test the consistency, robustness and stability of the obtained 3D geolectrical model. Experiments were also done to identify the depth of investigation upto which the model is valid. The main highlight of the resistivity model is the presence of two doublets comprising low-high resistivity features. One of the doublet (‘AB’) is located in MCT zone and Inner Lesser Himalaya and the second one (‘CD’) in Higher Himalayan region. The low resistivity feature of the doublet represents the fluid saturated fracture zone while the high resistivity feature represents the brittle rigid rocks. The conductive features (‘A’ and ‘C’) are related to the change in porosity, fluid content, pore distribution and possibly high heat flow and these can be explained in terms of the fluid saturated sediments adjacent to the resistive rigid rock matrix (‘B’ and ‘D’). The resistivity model supports the role of fluid in triggering the medium and large size earthquakes in the region. The model features were correlated with the velocity model obtained from seismic tomography studies. The resistivity model also explains the high heat flow and the presence of thermal springs in the area. The third case study was a ERT study where the data were used to characterize the subsurface soil at a bridge foundation site on the banks of Bhagirathi River at Tehri reservoir site, Uttarakhand, India. For this, six Electrical Resistivity Tomography (ERT) profiles were recorded on West and East banks of the river and these profiles were interpreted to determine the electrical resistivity image of the subsurface. The 2D and 3D subsurface resistivity models were obtained after inverting the data using the inversion codes Res2dinv iv and AP3DMT-DC respectively. Based on the 2D and 3D ERT inversions, the resistivity of different lithological units have been defined for the investigated sites. The basement depth has been found on the basis of resistivity variation. The basement depth was found on the basis of resistivity variations. The borehole data and geological inputs were used for lithological correlation and calibration of the resistivity values to the subsurface formation. The Standard Penetration Test (SPT) data (N-values) were correlated with the extracted resistivity values at selected points of the inverted models. The correlation study shows that resistivity is linearly correlated with the N-values. The coefficient of correlation was greater than 0.85, indicating that it is good and consistent. The relationship is site specific but useful for geotechnical investigations in the Himalayan region where undertaking a destructive test is prohibited. The last case study deals with the ERT and RMT data from Roorkee region which were used to study the effect of groundwater contamination due to untreated sewage irrigation near the Saliyar village, Roorkee. Nine RMT and five ERT profiles were recorded in and around the contaminated zone and one ERT and one RMT profile were recorded in the uncontaminated zone, for comparison. The two datasets were inverted independently as well as jointly using the 3D inversion code AP3DMT-DC. Prior to this, a synthetic study was carried out to validate the algorithm capabilities. This study demonstrated that the two methods, ERT and RMT, complement each other and that in the inverted model obtained through joint 3D inversion the resistivity values and geometry of the low and high resistivity features are better resolved than the models obtained through individual inversions. In the present study, the 3D models obtained after inversions of field data were found to be consistent with the published results of 2D inversion. The 3D inverted resistivity model shows an unconfined aquifer of low resistivity which is overlain by a slightly resistive near surface unsaturated soil formation. Moving away from the waste disposal site, an increase in resistivity was observed for the shallow unconfined aquifer. In comparison to the inverted results of an uncontaminated reference site, the inverted results of the contaminated region show a decrease in resistivity of the aquifer layer establishing the influence of contamination. The results of 3D joint inversion of the two datasets were encouraging in terms of accuracy and resolution of the model features and better explains the resistivity variation and geometry of all the features than the models obtained by individual inversions.