VELOCITY STRUCTURE OF THE GARHWAL HIMALAYA FROM INVERSION OF TRAVEL TIME DATA

Abstract

The knowledge of crustal velocity structure of the seismically active regions at local and regional scale is a key requirement to study the relationship between seismicity and geological structure, and to advance understanding of the seismotectonics and geodynamical processes operative in a region. The velocity structure allows studying the variations in physical properties within and along the fault zones to a scale of a few kilometers and provides valuable information on the segmentation of the fault zones, the nature and distribution of fault zone materials, and the locations and extent of possible asperities or nucleation zones (e.g., Michelini and Mcevilly, 1991). Travel time data of seismic phases have been extensively used to estimate the seismic velocity structure of the crust and upper mantle of the earth. For modeling of local velocity structure accurate travel time data are required. Significant improvements have been made in the estimation of velocity models due to advancement in instrumentation, digital data acquisition, and improved capabilities of data processing employing various types of computer softwares. Microearthquakes, because of their high rate of occurrence, provide large amounts of phase data within a short time interval. This phase data is of immense value for estimating the local velocity structure. The Garhwal Himalaya is one of the most seismically active regions of the Himalayan orogen, and has experienced many moderate and large earthquakes. The region has undergone tectonic deformations that resulted in the formation of several thrusts and faults mapped in and around the study region. The Main Central thrust (MCT) and the Main Boundary thrust (MBT) are the two major intracontinental thrusts of Himalayan orogon. The Garhwal Himalaya lies in the postulated central seismic gap between the 1905 Kangra earthquake and 1934 Bihar-Nepal earthquake. The region has a potential threat for great earthquake(s). Since 1991 three gap-filling earthquakes that include, the 1991 Uttarkashi earthquake (Mw 6.8), the 1999 Chamoli earthquake (Mw 6.6), and the 2015 Nepal earthquake (Mw 7.8) have occurred in the central gap. Therefore, the physio chemical mapping of the crustal material beneath the Garhwal Himalaya is useful for evaluating in areas that lie in the seismic gap. The aim of the present study is to estimate 1-D and 3-D velocity models primarily for the region around Tehri in the Garhwal Himalaya using local earthquake data, to iv relocate local earthquakes using these models, and to interpret the distribution of local seismicity and propose a seismotectonic model for the region. This study has a great relevance because a 260.5 m high Tehri dam—the highest earth and rock-fill dam in southeast Asia, lies in the study region. The study area lies between latitudes 29°50̛ N– 31°50̛ N and longitudes 77°E–80°E. The data set used in the study consists of the arrival times of the P and S phases of the local events acquired through the 12-station digital telemetered network deployed around the Tehri region in the Garhwal Himalaya. The network covers an area of about 100 km x 80 km around Tehri dam, with inter-station spacing of 10 -20 km. Each remote station houses a triaxial short-period seismometer (Model: CMG 40T-1(natural period 1 s)) to sense the three components of ground motion. The digital data is collected at a rate of 100 sample/s, the phase picking errors are estimated to be less than 0.05 s and 0.1 s for P and S waves respectively. The data of 2079 local events (0 ≤ ML ≤ 4) occurred in the study area from January 2008 to December 2012 has been used in the study. The SEISAN software (Ottemoller et al., 2011) is used for earthquake analysis and preliminary event location. An optimum 1D velocity model for the study region has been estimated using travel-time data of 145 local earthquakes with azimuthal gap ≤ 180°. Simultaneous inversion of travel time data allowed obtaining 1-D velocity model parameters, together with revised hypocenter coordinates and station corrections (Kissling et al., 1994 and Kissling, 1988). The travel-time curves of crustal phases are used to constrain the Moho depth. The model consists of six layers with P- and S-wave velocities ranging from 4.42 to 6.78 km/s and 2.41 to 3.71 km/s, respectively, up to a depth of 24 km. The Moho discontinuity seems to be at a depth of about 46 km, with P- and S-wave velocities of 8.34 and 4.86 km/s, respectively. A 2-km-thick low-velocity layer has been deciphered between a depth of 12 and 14 km. The existence of the low-velocity layer possibly can be attributed to the fractured basement thrust because of weakening of the crustal material at the interface between the overriding Himalayan block and the upper part of the underthrusting Indian plate. This is also supported by the large number of local events occurring in the vicinity of the MCT in the Garhwal Inner Lesser Himalaya. To study the local seismicity, more than 1400 events are relocated using the optimum 1-D velocity model, with the method of joint hypocenter determination technique (JHD). The spatial distribution of the relocated events shows that about 70% of the seismic activity occurred in the Inner Lesser Himalaya between the MCT and the Srinagar thrust (SNT). v The majority of the locatable events defined about 300-km-long northwest–southeasttrending seismicity zone that follows the trace of the MCT. The depth section drawn along the MCT delineates the geometry of the seismically active Main Himalayan thrust (MHT) below a 300-km-long segment of the MCT. It is found that the MHT is composed of two shallow-dipping fracture zones that seem to represent seismically active thrust zones dipping in opposite directions. The depth section of seismicity across the MCT showed two seismicity zones, at 10 and 15 km depth with a 5 km vertical separation. These zones clearly defined a flat-ramp-flat structure of the MHT in the vicinity of the MCT. The postulated front of the underthrusting Indian plate lies at a depth of about 15–18 km. The lower-flat seismicity zone bifurcates into two, indicating further slicing of the lower-flat zone. The postulated thickness of the brittle part of the underthrusting Indian crust is about 20 km in the vicinity of the MCT. Local Earthquake Tomography has been carried out using the algorithm LOTOS. The method of simultaneous inversion of P- and S-velocity structure with local earthquake data has been used. A total of 25,030 phase readings (12,794 P phase and 12,236 S phases) of the 1368 local earthquakes recorded from 2008 to 2012 have been used for the inversion. Horizontal tomographic images of the study area are shown as slices at depths of 1, 3, 5, 10, 12, 15 and 18 km. The horizontal sections of the P- and S- velocity anomalies include several interesting features. The near-surface and sub-surface images of high velocity regions identified in the Sub Himalaya and the Lesser Himalaya seems to indicate the trend and configuration of the postulated Delhi-Haridwar-ridge (DHR) beneath the IGP, the Sub Himalaya and the Lesser Himalaya. From the tomographic images we infer that the DHR continues up to the MCT with a lateral offset in the vicinity of the SNT. In the region around Chamoli in GHH, a high S-wave anomaly with a low P-wave anomaly was found. This area is located near the aftershock zone of the moderate magnitude Chamoli earthquake, and we infer that the subsurface area is highly fractured due to the after effects of the Chamoli earthquake. Subsurface geological structure defined by distributed velocity anomalies indicates the depth of sedimentary rocks is up to about 4 km. This sedimentary layer has relatively low VP and Vs velocity, low VP/Vs value and the Poisson's ratio range from 0.14- 0.18. The Relationship between seismicity distribution and 3D velocity structure has been studied from the revised hypocenters of events, obtained simultaneously with 3D inversion. From the pattern of distributed seismicity in the tomograms, it is found that most of the seismicity in the investigated area is concentrated on the transition boundaries of high and vi low velocity anomalies. The transition boundaries define the contact zones of the high velocity material and low velocity material. Seismicity along such zones occur because of stress amplification. These zones of high rigidity contrast, are the areas of high energy accumulation and prone to fracture because of stress failure and probable cause of the seismic events. The flat-ramp-flat type subsurface geometry of the underthrusting Indian plate below the Garhwal Himalaya has been outlined from the velocity distribution and depth distribution of more than 800 relocated hypocenters. This subsurface structure of the Indian plate is in conformity with the current understanding of the Indian crust. The subsurface low velocity anomalies indicated the thickness of sedimentary rocks is about 4 km. An area of the geometrical asperity has been found in the northwestern part, which is located on the basement thrust in the vicinity of MCT. The asperity region is defined by the high strength material with capacity to accumulate high strain energy that can be released in the future in the form of moderate earthquake. The narrow zone of concentrated seismic activity between the Uttarkashi thrust and the MCT seems to be the nucleation zone of future earthquake. This earthquake might nucleate at a depth of about 15 km.