|
Being able to accurately predict the level and variability of near-source ground motions for future earthquakes is a key ingredient to assess the seismic hazard and to mitigate and manage the resulting risks for people, buildings, and infrastructure. A particularly promising avenue for improved ground-shaking estimation comprises ground-motion simulations using physics-based earthquake-rupture modeling. We therefore seek financial support to expand previous work related to the pseudo-dynamic earthquake characterization (Guatteri et al., 2003, 2004) by considering the latest insights from rupture dynamics on the scaling of fracture energy and dynamically consistent slip-velocity functions (Tinti et al., 2005a; Rice et al., 2005; Mai et al., 2006). We will develop, implement and test an improved numerical method for constructing physically self-consistent source-rupture models for ground-motion simulations. This pseudo-dynamic approach poses no restrictions on the frequency range which simulated rupture models may radiate; however, ground-motion calculation techniques (frequency-wavenumber integration; finite-difference or spectral-element methods) have strong limitations when computing seismograms up to frequencies of engineering interests (5 - 20 Hz). Moreover, earth structure remains unknown at these short spatial scales, and hence alternative, stochastic techniques are needed to compute the high-frequency contributions of seismic radiation. Based on work by Zeng et al (1993, 1995), Mai and Olsen (2005) proposed a hybrid technique to calculate broadband near-fault seismograms which includes effects of scattering in inhomogeneous media. We propose to further develop this method by conducting an extensive parameter-space study and by establishing a platform-independent modular implementation. The goal is to provide a well-tested and calibrated toolbox that can be installed in other large-scale ground-motion simulation efforts (e.g. SCEC) in order to adequately represent the high-frequency components of the seismic wave-field. The above approaches require considerable efforts for implementing and testing the numerical methods that capture the rupture physics and near-source scattering; moreover, the new methods need to be validated against well-recorded ground-motions of past earthquakes and the engineering applications. To that end, we have started to develop innovative ground-motion intensity measures for application-specific purposes. We investigate “efficient” and “sufficient” ground-motion intensity measures (IM) for structural engineering, where the demand measure is residual displacement, and for geotechnical work, where the demand measure is related to the liquefaction potential of soils. In this context we are building upon and continuing the work started in the SNF-funded MERCI project (Managing Earthquake Risk using Condition Indicators, in original 200021-104027, renewed for 1 year in 2006 under 200020-112326/1). This proposal seeks funding for one PhD student for a period of three years. The work builds upon (partially) existing numerical codes and ground-motion simulation tools that need to be combined, tested, calibrated and validated, before they can be applied for engineering applications and released to the earthquake-engineering / seismology community. We are in the fortunate position that SNF has already funded the first year of a graduate student, Ms. B. Mena Cabrera-Sanli, who has proven with her initial work to be highly qualified for this project. We thus request financial support for Ms. Cabrera-Sanli to complete the project tasks (detailed below), to work towards her PhD-degree, and to expand her interaction with the earthquake-engineering community. This PhD-project will thus allow her to gain a broad insight into all aspects of earthquake physics, seismology, and the engineering application of advanced numerical simulations.
|