In this study we investigated the source dependent contribution to the strong ground motion in 3D medium with specific application to a characteristic M7.0 earthquake on the NI fault in the Los Angeles basin. The fault perpendicular component (y) exhibits at least twice larger peak velocity amplitudes than the fault parallel components (x, z) in the area around the right part of the fault, that is, in the direction towards which the rupture is propagating. This phenomenon is also observed in the strong ground motion measurement and is a crucial aspect in earthquake hazard and risk analysis.
Seismic shaking variations due to various rupture processes are also investigated in this study. The static slip (input parameter for kinematic finite source simulations) is impor-
7.5 Discussions and Conclusions 67 tant when calculating seismic motions because the area with large static slip (asperity area) will introduce elevated seismic motions. In combination with directivity, it can increase the complexity of surface seismic shaking.
Comparing the modulus component of PGV with those of the two horizontal components for an M7 strike-slip earthquake, we conclude that the modulus component is dominated by the fault perpendicular component in terms of absolute value while equally by the two horizontal components in terms of relative variation. The different effect due to directivity and slip history, respectively, should be considered when doing ground motion prediction, at least in our specific case, that is, NI fault and LA basin.
The results shown here indicate that to reliably estimate shaking hazard for specific earth- quake scenarios, the source dependent effects should be taken into account. This necessitates calculation of many different slip histories to fully account for the associated uncertainties. In the (long-) period range considered in this paper there is a remarkable difference between the finite-source effects on the various motion components, mixed with effects of the 3D structure. More simulations will be necessary in the future to quantitatively estimate the contributions of small scale structures and source effects at higher frequencies.
Rotational motion is also recorded during the numerical Green’s function generation pro- cess which enable us to investigate the how the ground rotation varies according the variation of different parameters as discussed for the translational part of motion in the forward two chapters. The results are shown, analyzed and compared with the translational motion in the next chapters.
Application: Rotation Rate
ABSTRACT:Recently, the instrument technology (e.g., ring laser) is capable to accurately record the rotational motion observation. However, due to the scarcity of the data, the systematic rotation variation in 3D media is still a challenge. In order to answer that question, we simulate several M7 earthquakes with different source scenarios on the Newport-Inglewood (NI) fault embedded in the 3D Los Angeles Basin and. The recently calculated data base with several hundred numerical Green’s functions for a discretized model of the NI fault allows arbitrary finite-fault scenarios to be synthesized by superposition. We investigate source and basin structure effects on the rotational components of ground motion (e.g., peak ground rotation rates and their variation or horizontal gradients) and compare with the translations. Only the rotations from the seismic source with assumption of linear elasticity are considered.
8.1
Introduction
In conventional earthquake engineering, seismic loads on structures are formulated only in terms of the three translational ground motion components. Nevertheless, each site on the surface can be subjected to six kinds of motion: three translations alongx-,y- and z-axes as well as three rotations about these axes. While the building response shaken by translational motions has been long investigated, building response of the rotational motions is just at the beginning.
The engineering importance of rotational components of seismic strong ground motion was noted during late sixties and early seventies of the last century (e.g., Newmark and Hall,1969; Newmark and Rosenblueth,1971). There are many reports about rotations of tombstones and stone lanterns during large earthquakes (e.g., Yamaguchi and Odaka,1974). Zembaty(2006) also pointed out that the rotation effects become especially important in the case of high buildings. Difference models have also been studied and used to evaluate the effects of torsional excitation (e.g., Gupta and Trifunac,1987,1989,1990) and of rocking excitation (e.g., Gupta and Trifunac,1988,1991). These studies have shown that torsional and rocking excitations can be significant and with combinations of structural and soil properties are important for such excitation. For example, rocking excitation becomes important for tall structures supported by soft soil while torsional excitations can dominate in the response of long and stiff structures supported by soft soils (Zembaty and Boffi,1994). These buildings are very sensitive to angular momentum carried by rotation waves. Moreover, incident rotational waves interact with the geological structure (e.g., granular medium) and may lead to their softening, degradation,
8.1 Introduction 69 damage or fracture. Geological media with rotational degrees of freedom (e.g., the so-called “liquid sands”) are extremely sensitive to any rotational motions or vibrations.
A question whether the rotation motions at seismic source can propagate in a form of wave through geological layers from the seismic source to a distant recording station seems still open. Recent theoretical studies (e.g., Teisseyre, 2004; Boratyński and Teisseyre,2004) and observational results bring a positive answer to this question. An interesting contribution in this field was a simulation study of rotational effects in layered media and near-field excitations byBouchon and Aki(1982). They simulated rotational ground motions near earthquake faults buried in layered media for strike-slip and dip-slip fault models, and obtained a maximum rotational velocity of1.5×10−3rad/s produced by a buried 30 km (seismic moment of8×1018
Nm) long strike-slip fault with slip of 1 m.
Such small rotational motions are difficult to be measured. Deployment of strong motion accelerographs in many seismic areas of the world during the past seventy years has produced data on translational components of motion during many strong earthquakes. This data describes strong motion in three orthogonal directions (two horizontal and one vertical), but because the spacing of the recording sites is much larger than the wavelengths of the recorded motions, little is known today about the accompanying differential and rotational motions.
The ground rotational motions were observed by Nigbor (1994), Spudich et al. (1995), Stedman et al.(1995), McLeod et al. (1998), Spillman et al. (1998), Takeo (1998), and Igel et al.(2005). The measurements carried out byTakeo(1998) were recorded in the near-source region of earthquakes. The measurements bySpillman et al.(1998) were carried out far from an earthquake source (200 km and more).
Many questions related to the rotational motions remain open: up to now, we have no reliable data on propagation properties (velocity and attenuation) of such energy through geological media and on the influence of distance from source to the recording station. Trans- lational and rotational components of strong motion radiated from an earthquake source are modified along the propagation path through interference, focusing, scattering, and diffrac- tion. For the velocities, complex variations are observed in chapter 6 and chapter 7 with respect to the varying hypocentre location and varying slip history. In this chapter, how the ground rotation rates vary with these source and sub-surface parameters are investigated. The results are shown and analyzed.
In recent decades ring laser technology succeeded in detecting small variations of the earth’s rotation rate introduced by earthquakes. At teleseismic distances the seismic wave field can be well approximated with plane waves. Under this assumption, a direct comparison between rotation rate around a vertical axis and transverse acceleration (recorded by a standard seis- mograph) is possible – translations and rotations are in phase and their amplitudes scaled by twice the horizontal phase velocity. Igel et al. (2005) study similarity between the collo- cated translations and rotations by taking the advantage of the new developed instrument – ring laser. Their observations showed over a very broad frequency range (1–150 sec) that the waveforms have the expected similarity and the amplitude ratios allow an accurate estimate of horizontal phase velocities (Igel et al.,2005;Cochard et al.,2006). In the near source region, the fundamental assumption of the above mentioned study – the plane wave propagation – could not hold any more. But what the comparison between the translation and the rotation will reveal is still considered an interesting topic. This is the one of the targets of this chapter, too.
1. What are the peak ground rotation rates to be expected for M7 earthquakes? 2. How does the structure affect the spatial distributions?
3. What are the source-dependent variations?
4. How do these variations (spatial distribution) compare with those of the accelerations? 5. What can we learn from the joint processing of rotation and acceleration?