QUANTIFICATION OF SEISMIC REGIMES IN KANGRACHAMBA REGION OF HIMACHAL HIMALAYA

Abstract

The Kangra-Chamba region of NW Himalaya is seismo-tectonically active and geologically one of the most complex region of the world. Fresh compilation of seismic catalogue of NW Himalaya reveals that though Kangra-Chamba region witnessed only one great earthquake (M8.0 Kangra earthquake of 1905) the region has been a zone of intense seismicity evidenced by clustering of epicenters and energy release irrespective of the duration and period of observation. A distinctive aspect of this part of the Himalaya is that the region is transgressed by the extended Chamba Nappe (CN). The weakly metamorphosed strata of the CN are similar in facies to the Tethys Himalaya (TH) and are considered to have resulted from southwestward sliding of the TH sediments from the north over the metamorphic Higher Himalayan Crystalline (HHC) along the south dipping Chenab Normal Fault (CNF) that separates the CN from the HHC. The Kistwar (KW) and Rampur (RW) windows are large antiformal folds to the NW and to the SE of the present study area. The present thesis describes the design, processing, results and interpretation of a special local seismic network undertaken to improve our understanding of the structures and seismotectonics of the region. The high quality local earthquake data is first used to investigate the ID and 3D variation of crustal velocity structure. The inversion of travel-time paths could resolve velocity structure for the upper 20 km or so as > 85% earthquakes are confined to depth less than 20 km of the upper crust. Therefore, the determination of the thickness and velocity up to Moho and upper mantle were supplemented independently by traveltime- distance plot using data of regional seismic networks and scanning epicentral distances up to 550 km. The minimum 1-D velocity model divides the average 44 km thick crust into four layers. The top -10 km thick layer represents the metamorphosed sediments of the CN that dominates the surface geology of the study area in the central part. A thin low velocity layer at 15 km depth, after allowing for the station corrections, portrays the geometry of the detachment zone separating the down-going Indian plate from the overriding wedge. In addition to the expected relative high station corrections over the Siwalik, the single most anomalous feature indicates unambiguous presence of a structural discontinuity in the southeast corner striking NE-SW. While inverting the data for 3D crustal structure investigation using Local Earthquake Tomography (LET), a systematic and significant variation up to 14% in Vp and Vs along with up to 6% in the Vp/Vs ratio has been observed across the major iii tectonic units. The low velocity perturbations (LVP) largely overlap with known exposure of the Siwalik and Lesser Himalaya, while relative high velocity area correspond to metamorphosed sedimentary sequence of TH and thus helps to map the spatial extent of the CN at different depths. It is clearly seen that the southward extent of the CN on the western part of the study area is limited by the Chamba Thrust (CT) whereas on the eastern part the CNextends right up to the Panjal Thrust/Main Boundary Thrust (PT/MBT). This varying limit of southward extension of the CN helps to resolve the long standing structural puzzle that the CT bordering the Chail formation to the north is well marked west of Chamba whereas this contact is not traced east of Chamba. Another distinctive feature of velocity tomograms is a narrow NE-SW aligned LVP in the eastern part, confined to depth interval of 2 to 5 km and is underlain by a high velocity zone in tomograms at 10 km depth. The presence of a structural discontinuity striking NE-SW is coincident with the sharp terminus of positive station delays. The distribution of hypocenters in specific depth intervals clearly depict that most of the hypocenters are concentrated at the boundary between low-high velocity couplets correlating with crustal heterogeneity. Since such linkages are better developed and outlined in Vs tomograms, part velocity variability may reflect rheological changes associated with varying degree of saturation. As an independent mode to constrain the seismotectonic model, the fault plane solutions (FPS) of 42 well located events (RMS value of <0.1s and M>2.5) are computed and analysed. The complex tectonic setting of the sector is immediately evident from the fact that three dominant focal mechanisms noted are: (1) thrust fault mechanism with strike slip component, (2) strike slip movement with thrust component and (3) normal fault motion with strike slip component. The FPSs of the past earthquakes located at detachment highlight the dominance of the thrust environment of the region on a broad tectonic scale. The earthquakes occurring along the shallow sections of the MBT and PT are characterized by thrust mechanisms with varying component of strike slip movement. The decreasing dip of the nodal planes with increasing depths on these thrusts favor the tectonic hypothesis that steeply dipping thrusts near the surface flatten out at depth to merge with the detachment plane zone marking top of the down-going Indian Plate. The normal fault mechanism along the plane which is seen as a subsurface extension of the CNF is consistent with the tectonic model which postulates NE-SW directed extensional tectonics responsible for the southward displacement of the CN along the normal fault. Earthquakes located beneath iv the detachment zone in a localised cluster northeast of the epicenter of the Kangra earthquake are dominated by normal fault mechanism or the NE-SW directed structural discontinuity. The source parameters of the earthquakes are obtained using Brune's circular model through the corrected amplitude spectra based on geometrical spreading and site effect. For these earthquakes the determined corner frequency of the spectrum is in the range 1 to 10 Hz. The seismic moment and source radius of these events is 7.0x1017<Mq<7.0x1022 dyne-cm and between 0.1 to 1.2 km respectively with larger values for higher magnitude earthquakes. The specified relation of Brune is used to convert the seismic moment into moment magnitude (Mw). The linear regression analysis is performed to obtain relations between Mw and the local magnitude ML as well as coda magnitude Mc. The empirical linear relation shows that the variation in ML is equal to 0.70 times change in the value of Mw whereas, this change in Mc is 0.86 times relative to Mw. The obtained stress drop has a different characteristic for microearthquakes compared to bigger size events. The maximum stress drop is 27 bar with a majority in the range 0.1 bar to 10 bar showing increase in value with the earthquake size. The low stress drop near the plate boundary is related to high levels of earthquake activity with typically weak fault rheology. This work is among the first studies to provide both spatial and temporal analysis of seismic data for quantification of seismic regime in Kangra-Chamba region of Himachal Himalaya. Given the shallow character of seismicity, the limitation of not imaging the velocity structures for full thickness of the crust can be circumvented by the application of global tomography using receiver function approach. Further, given the significance of imaging the true character and geometry of the BTF, both in the tectonic evolution and seismicity in the Himalayan collision zone, active seismic and high resolution broadband magneto-telluric surveys would be timely and rewarding.