FINITE-DIFFERENCE MODELLING OF 3D BASIN EFFECTS ON GROUND MOTION CHARACTERISTICS

Abstract

The 2D13D basins play a major role in amplification of ground motion level and generation of different seismic phases. The various physical phenomena occurring in a basin during an earthquake are responsible for the significant spatial variation of ground motion characteristics. The presence of very large population-density as well as city-density in the basins worldwide stimulated this research work. The different aspects of basin effects on ground motion characteristics and damage pattern include the physical phenomenon like single and double resonances (Romo and Seed 1986; Stone et al., 1987; Narayan et al., 2002), basin-generated surface (BGS) waves (Kawase. 1996: Graves et al.. 1998: Pitarka et al., 1998). subsurface basement focusing (Gao et al.. 1996: Davis et al.. 2000: Booth et al.. 2004; Narayan and Kumar. 2012: 2014a), complex mode transformation of the basin-transduced surface waves - (Kawase. 1993; Narayan 2010; 2012a: Narayan and Kumar, 2014b) and site-city-interaction (Gueguen et al.. 2002: Kham et al., 2006: Semblat et al., 2008; Sahar et al., 2015). Furthermore, realistic quantification of basin effects on the ground motion characteristics requires an efficient numerical method since analytical solutions are not possible. The incorporation of frequency-dependent damping in the time-domain numerical simulation of basin is very much essential since the BGS-waves are highly affected by the sediment-damping (Day and Minster, 1984; Emmerich and Korn, 1987; Kristek and Moczo, 2003). The quantification of characteristics of the BGS-waves in a basin is very important since the flat layer response are generally inadequate to explain the observed damages in 3D basins (Hatyarna et al., 1995: Kawase, 1996; Graves et al., 1998; Pitarka et al., 1998; Bakir et al., 2002). The BGS-waves develop very large differential ground motion which can adversely affect long-span lifelines such as pipelines, bridges and communication transmission systems (1-lather et al.. 2008). Generally, the damage surveys and numerical simulations reveal that the BGS-waves cause intense damage in a bandwidth parallel to and at some distance frorn the basin-edge (Pitarka et al., 1998; Narayan, 2005). This may be due to 2D nature of most of the basins, where damage survey have been carried out. The extensive literature reviews associated with the basin effects on the characteristics of the BGS-waves revealed that nobody has studied the effects of circular basins like intracratonic basins (e.g. Michigan basin and Congo basin) and impact crater basins (e.g. Sudbury basin and Arizona basin) on the characteristics of BGS-waves. Now, question arise in the mind, what will be the effects of circular basin on the characteristics of BGS-waves propagating towards the centre of circular-basin and associated spatial variation of ground motion? Further, whether there may be focusing of the BGS-wave or not and if it is happening whether focusing will be a frequency dependent or not. Furthermore, it appears that nobody has documented the effects of azimuth of the incident body waves on the characteristics of the BGS-waves for an azimuthal range of 0-360° (Oprsal et al., 2005). In order to fulfil the above identified gaps as well as to find out the answers to the aforesaid questions, the viscoelastic seismic responses of the various considered circular basin models have been computed along different arrays and analysed. The research works have been divided in two parts. The first part of the thesis describes the development of an efficient 3D viscoelastic finite-difference (FD) algorithm for incorporating the frequency-dependent damping in the time-domain simulations. The second part of the thesis describes the effects of various parameters like shape-ratio, geometry of sediment basement interface (GSBI), size and the sediment rheology of the semi-sphericai (SS-) basin on the characteristics of the BGS-waves and associated spatial variations of amplitude amplification. average spectral amplification (ASA) and differential ground motion (DGM). The effects of azimuthal angle of the incident body waves on the characteristics of the BGS-waves are also studied in details. In order to show the importance of 3D basin simulations as well as to help in improving the level of seismic microzonation of an area which has already been carried out based on ID response of sediment deposit, average aggravation factors (AAF) have been computed. Due to the lack of suitable earthquake records across the intracralonic and impact crater circular basin, sediment data as well as the basin basement geometry, validation of the simulated characteristics of the BGS-waves in the circular basin could not become possible - and only parametric studies could be carried out. An excellent correlation between the computed phase-velocities for the S- and P-waves using the FD responses of an unbounded viscoelastic homogeneous medium with the same computed analytically using the Futterman's relations (1962) and the GMB-EK rheological model (Emmerich and Korn. 1987) revealed the accuracy of implementation of frequency-dependent damping in the developed 3D viscoelastic FD program. Furthermore, the almost frequency-independent inferred nature of the computed quality factors also supports the above conclusion (McDonal et. al.,1958: Kristek and Moczo. 2003). The analysis of the computed responses of the semi-circular (SC) and semi-spherical (SS) basins for the different polarizations of the incident plane S-wave front along different arrays revealed that both the Rayleigh and Love waves are generated at each point of a basin-edge. Further, a simple combination of the incident plane S-wave front with a particular polarization and the SS-basin made it possible to study the effects of azimuthal angle of the incident body wave on the characteristics of the BGS-waves in a azimuthal range of 03600. It was inferred that the amplitudes of Rayleigh and Love waves depend on the amplitude of component of the incident body wave within and transverse to a plane which is normal to the basin-edge. Based on the findings of the research work, it may be suggested that there should be parameters like azimuth. focal mechanism and angle of incidence of body wave, shape of basin, location, sediment and rock rheology in an empirical relation for predicting the amplitude of BGS-waves in the components normal and parallel to the basin-edge (Joyner. 2000: Somerville et al.. 2004). The matching of the obtained lowest frequency content in the BGS-waves with the lowest resonance frequency (F0) of the SS-basin corroborates with the finding of Bard and Bouchon (1980a&b) and Narayan (2005). The observed increase of amplitude of the BGS-waves towards to the centre of the SS-basin, even in the presence of sediment-damping and the dispersion of the BGS-waves may be attributed to the focusing of the BGS- waves. Furthermore, it is concluded that the focusing of the BGS-waves in the SS-basin is frequency- - dependent. The obtained level of ASA/DGM is comparable in both the SC- and SS-basins near the basin-edge but the difference is increasing towards the centre of basins. It is also concluded that the focusing and trapping of the BGS-wave in the SS-basin is more sensitive to the impedance contrast as compared to the sediment-damping. The analysis of responses of the SS-basin with different basin shape ratio revealed a decrease of ASA with the decrease of basin shape ratio due to (I) decrease of frequency bandwidth of the BGS-waves and (2) decrease of spectral amplitudes of the BGS-waves due to an increase of phase velocity. It is inferred that the spatial variation of characteristics of the BGS-waves was highly controlled by the geometry of sediment-basement interface. The obtained ID AAF in the Sc- and SS-basins clearly reflects the inadequacy of 1 D response of a 2D/3D basin. It is recommended to compute the AAF using the same components of ground motion as that of the incident body wave to conservatively aggravate the ground motion to incorporate the 3D basin-effects in the seismic microzonation where it has already been carried out using 1 D response of the soil column. The findings of this research work calls for a special attention during the seismic hazard assessment in the intracratonic and impact crater circular basins.