EARTHQUAKE SOURCE PARAMETERS AND ATTENUATION CHARACTERISTICS OF ARUNACHAL REGION

Abstract

The seismic waves produced by the earthquakes are used to study the structure and composition of the earth’s interior, and the properties of earthquake source. However, to understand how the wave field is radiated from the earthquake source it is important to deduce the properties of the earthquake source as the resolution of the earth’s structure requires precise knowledge of the parameters of earthquake source and path travel by seismic waves (Shearer, 1999). The parameters characterizing the source of an earthquake are called the earthquake source parameters. These parameters are computed in the time domain from the measurements of arrival times of seismic phases and in the frequency domain from the spectra of the seismic waves. In the early 1930’s Richter introduced the magnitude parameter to measure the size of an earthquake using the amplitude of seismic waves. The similar work of Aki (1967) and Brune (1970) laid a strong foundation to develop the scaling laws and to compute the earthquake source parameters using the spectra of seismic waves. The earthquake source parameters can be broadly classified into two types: the kinematic parameters and the dynamic parameters (Duda, 1978). These parameters provide a great deal of information about the properties of the earthquake source, and find a large number of applications in Seismology and Earthquake Engineering. Study of seismic wave attenuation provides the information regarding the medium through which they travel and this information can be interpreted in terms of both physical properties of the geological formation as well as level of inhomogeneities present in the medium. This is important for understanding the seismotectonics and also play significance role in estimation of seismic hazard for a given region (Aki 1969, Aki and Chouet, 1975). Himalaya mountain range is formed by the collision of Indian plate with Eurasian plate about 40 to 50 million years ago. It leads to the building of regional and local scale tectonic features. The main boundary thrust (MBT) and the main central thrust (MCT) are few of them.. The most of earthquake occurred in the Himalaya are along these tectonic features. It is important to understand the earthquake source process and predict the strong ground motion. Two main factors that plays important role in understanding the tectonics and medium properties of a region are earthquake source properties and seismic wave iii attenuation. So the understating of these attributes of earthquakes is important for various purposes such as river valley projects and other construction works in the Himalayan region. The Himalayas can be divided structurally (tectonically) from north to south as: Tibetan Himalaya (Northwest part of Arunachal Himalaya bordering Bhutan and Tibet, trending NE– SW), Higher Himalaya (limits between Tibetan Himalaya and MCT, ENE–WSW trend adjacent to Bhutan and changes to NE– SW eastward), Lesser Himalaya (limits between Higher Himalaya and sub-Himalaya, trending E–W in western part, swinging NNE–SSE till the syntaxial then NW–SE), sub- Himalaya (trending E–W near Bhutan, swings ENE–WSW towards east) and the division from east to west: The Eastern Himalaya, Central Himalaya and western Himalaya. (e.g. Gansser 1964; Le Fort 1975). In the northeastern part of India, is Arunachal Pradesh which mostly occupied by the high mountain range of eastern Himalayas. The Himalayan range enters in Arunachal Pradesh from Bhutan at the west of Kameng district, and the altitude in this region varies from 800–7,000 m above mean sea level. It runs through the northwards region over the Kangto region before ending at the easternmost part of Arunachal Himalaya, i.e. The Namcha Barwa Massif. This part of the Himalayas includes extensive faulting and over-folding as the major structural element. The regional structural trend of the eastern Himalayas is mostly E–W to ENE–WSW from Bhutan to the northeastern Arunachal Pradesh, which changes gradually to NE–SW near Siang valley and terminates against Siang fracture (e.g. Nandy 1976). Geotechnically, the Arunachal Pradesh can be divided into four geotechnical blocks: the Himalaya, the Mishimi hills, Nagapatkoi ranges of the Arkon Yomo Mountain and Brahmaputra Basin; each of these have experienced intense stages of tectonic development in response to collision of plates and uplift of the Himalayas (Kumar 1997). From the broad description above, the study region, lower Siang of Siang valley is in Arunachal Himalaya surrounded by Main Frontal Thrust, Main Boundary Thrust in the south, Mishmi Thrust, Tidding Suture and Lohit Thrust in the east and Bame-Tuting Fault in the west. This region is in seismic zone V as per IS Code (IS 1893 (Part 1): 2002) of India. This region shows polyphase deformation, and three stages could be recognized in the region. Six major lithotectonic belts with different litho-stratigraphic settings and deformation patterns iv separated from one another by regional thrust planes are present in the valley (e.g. Singh and Chowdhary 1990). Like the other parts of Himalaya, the eastern most Himalaya of Arunachal Pradesh exhibits quiet high seismicity and lies within the seismic zone V as per IS Code (IS 1893 (Part 1): 2002). Several earthquakes of smaller to moderate magnitudes have occurred in this region (Kayal, 1987). The two great earthquakes namely Shillong (1897) and Assam (1950) earthquakes having magnitudes 8.1 and 8.6, respectively, fell in close proximity to the study region. Shillong earthquake on June 12, 1897 (Mw, −8.1) located near the northern edge of Shillong Plateau while Assam earthquake on August 15, 1950, located in Mishmi hills. On September 18, 2011, the Sikkim The present study is based on observational data and is devoted to the estimation and interpretation of source parameters and wave attenuation characteristics in the Lower Siang region of Arunachal Pradesh using the data of microearthquakes of small and moderate earthquakes. This study has been carried out with the following objectives: • To estimate the source parameters of local earthquakes for Lower Siang region of Arunachal Himalaya and to develop the scaling relations between various source parameters. • To estimate the focal mechanism of local earthquakes using the waveform inversion method. • To determine the seismic wave attenuation characteristics of Lower Siang region of Arunachal Himalaya by estimating the quality factor of P-wave (Qα), S-wave (Qβ) and coda waves (Qc) • To investigate the lapse time dependence of Q of the local earthquakes data and to develop frequency dependence attenuation relationships for the region. c • To separate out the effect of the intrinsic and scattering attenuation from the total attenuation. . The P-wave and S-wave arrival time data of 104 local earthquakes has been measured from the digital seismograms obtained during the period from July 2011 to May 2012 from v five seismological stations.The hypocentre parameters of local earthquakes have been estimated using HYPOINVERSE computer program (Klein, 1978) integrated in SEISAN software used for the estimation of hypocentre parameters. The velocity model given by Khattri et al. (1983) used for the estimation of hypocentre parameters. The standard errors in the estimation of hypocentre parameters for these events are ≤ 0.50 sec in origin time (RMS), ≤ 5.0 km in epicentre (ERH), and ≤ 5.0 km in focal depth (ERZ). The results on the source parameters and fmax have been obtained for the Lower Siang region of Arunachal Himalaya from the analysis of 104 local earthquakes The results showed that the values of seismic moment, Moment magnitude, the source radii and stress drops varies from 1.6x1018 to 3.1x1023 dyn-cm, 1.4 to 5.0, 157.8 to 417.1 m and 0.1 to 74 bars respectively. Except a few events, the stress drop remains constant and does not vary with focal depth. Hence no dependency of stress drops with depth is obtained for the study region that indicates that the stress drop is only the function seismic moment and fracture radius. Based on results on seismic moment and corner frequency, the scaling relation M0 α fc-3.27 has been obtained for the study region. Δσ relationships with M0, log (Δσ) = 0.1804 log(M0) -2.945 is also obtained for the study region. Satoh et al. (2000) obtained a relationship log (Δσ) = -2.91 log(M0) +5.8 using deep borehole data of Japan. The results obtained in present study are seems to be more realistic than Satoh et al. (2000). Satoh et al. (2000) relationship suggests stress drop increases as the seismic moment decreases. But in real scenario it is not possible because lower magnitude event can never have stress drop greater than high magnitude event. The results have been obtained based on the comparative study of fc and fmax showed that both fc and fmax behaves in a similar manner with respect to source, focal depth and epicenteral distance. From the various plots of both fc and fmax with seismic moment, focal depth, epicentral distance at different recording site showed the same amount of scatters and trends in the distribution of data. From this, it is observed that fmax is having almost similar behavior to seismic moment as obtained for fc to the seismic moment. For this it is brought out that fmax is due to source process because it is affected by source as well as site in similar way as fc and it is well known that fc is the source property so fmax is also a source effect. Both fc and fmax are found to be independent of epicentral distance and depth of occurrence. Relationship of fmax with seismic moment and stress drop has also been obtained. vi In the present study efforts have been made to manifest the source of non-DC (CLVD) component of MTs. from the moment tensor analysis of 104 local earthquakes. The moment tensor analysis of 41 earthquakes out of 104 earthquakes having magnitude greater than 2.5 and signal to noise ratio (SNR) greater than 6 with small epicentral distances brought out more clarity in the identification of cause of CLVD. It has also been observed that variation in CLVD caused by various source such as noise in the data, focal depth and magnitude of the event. Source depth acts a controlling factor of CLVD. For selected 41 events the high CLVD component confined to shallow depth. For high SNR the CLVD% is quite low.CLVD is also dependent on the magnitude of event. In the higher magnitude range greater than 3.5, DC% is greater than 60%. Hence most of high magnitude earthquakes are having double couple (DC) source mechanism. The appearance of CLVD in low magnitude may also be attributed to the directivity effect (Adamova and Síleny, 2010). The seismic wave attenuation has been studied for Lower Siang region by estimating the quality factor of P-wave (Qα) and S-waves (Qβ) in the frequency range 1.5 to 24 Hz adopting the extended coda normalization method. The estimated value of Qα and Qβ are found to be strongly frequency dependence in the study region. Their mean values vary from 49±4 at 1.5 Hz to 1421±6 at 24 Hz for Qα and from from118±6 at 1.5 Hz to 2335±5 at 24Hz for Qβ. The frequency dependence Qα and Qβ relationships are obtained as, Qα = (25±1)f(1.24±0.04) for P-wave and Qβ = (58±1)f(1.16±0.04) for S-wave. The comparison of Qα and Qβ has brought out that P wave attenuates more rapidly as compared to S-wave at all frequencies. The results obtained in the present study are found to be comparable with the other seismically active regions of the India as well as world. The comparison of Qα and Qβ with the other seismically active regions of India showed that the region around the Lower Siang is more attenuating among almost all Indian regions except the Chamoli region which showed higher attenuation, whereas the comparison among some regions of World showed similarly increasing pattern with increasing frequency. The higher frequency dependence of attenuation describes the region is high seismically active. Also low value of Qα and Qβ in the Lower Siang region as compared to other Indian regions indicate the high tectonic activity. Higher Qβ than Qα for the entire frequency range indicate the crust of the Lower Siang region of Arunachal Himalaya is highly heterogeneous. vii The coda waves of 104 local earthquakes have been analyzed for three lapse time windows (30, 40 and 50sec) employing the single backscattering model at seven frequency bands with a central frequency in the range of 1.5 Hz to 24.0 Hz. Obtained Results show the average variation in Qc is from 109±33, 138±42 and 162±46 at 1.5 Hz to 3149±923, 3439±944, and 3889±1165 at 24 Hz for lapse time windows of 30, 40 and 50 sec, respectively. The frequency dependence relationships; Qc=(52±1)f1.22±0.03, Qc=(83±1)f1.18±0.02 and Qc =(105±1)f1.16±0.02 are obtained for lapse time windows of 30 , 40 and 50 sec respectively. So in the study region Qc is found to be function of frequency and lapse time window and Qc increase with increase of frequency as well as lapse time window. The increase in Qc value with the time window is attributed to the increase in Qc with depth because heterogeneities of the medium decreases with depth. in the frequency dependent relation given above Q0 (Qc at 1 Hz) value increases with the lapse time window while there is a nominal decrease in the degree of frequency dependence (η) with increasing window length. This may be due to decrease in scattering effect and hence the decrease heterogeneities of the medium with depth. The comparison of Qc value obtained in the Lower Siang region with that obtained in other part of India and World showed that Siang region is tectonically active. The separation of scattering (Qs) and intrinsic (Qi) attenuation from the S-wave attenuation (Qβ) and coda wave attenuation (Qc) has been carried out employing the Wennerberg (1993) method and developed the frequency dependent Qs and Qi relations. The Qi and Qs show the frequency dependent character in the frequency range 1.5 to 24 Hz. The average scattering and intrinsic relationships are obtained for the region as Qs=(31±1)f1.04±0.02, Qs=(48±1)f1.05±0.02and Qs =(61±1)f1.05±0.02 and Qi=(68±1)f0.95.±0.06, Qi=(134±1)f1.01±0.05 and Qi =(167±1)f0.96±0.03 for lapse time windows of 30 , 40 and 50 sec, respectively. Both Qc and Qi at a given frequency increases with lapse time duration. This can be attributed to the decreases in the heterogeneities and the inter-grain fraction with increase in depth as larger the lapse time represents the characteristics of deeper depths. Qi and Qs comparison and values of seismic albedo shows that scattering attenuation is more prominent over intrinsic attenuation in the region over entire frequency range 1.5 to 24 Hz. Various studies of Q in various regions of the world show high Q-value in seismically stable regions and relatively low Q value in the seismically active regions. The Q-values and viii their frequency dependent relationships estimated in the present study are well correlate with highly seismically active and heterogeneous regions. The frequency dependent attenuation relations developed in present study would be useful in various scientific and engineering applications including earthquake hazard assessment, earthquake source parameter estimation and understanding the physical phenomenon related earthquake elastic energy propagation of Lower Siang region as well as other regions of Arunachal Himalaya and NE region as a whole.