پیش بینی حرکت زمین مبتنی بر فیزیک
Abstract
1. Introduction
2. Ground motion prediction of the 2010 Darfield, 2011 Christchurch, and 2016 Kaikoura earthquakes
3. On-going challenges in ground motion simulation
4. Discussion and conclusions
Acknowledgements
References
Abstract
This paper presents on-going challenges in the present paradigm shift of earthquake-induced ground motion prediction from empirical to physics-based simulation methods. The 2010–۲۰۱۱ Canterbury and 2016 Kaikoura, New Zealand earthquakes are used to illustrate the predictive potential of the different methods. On-going efforts in simulation validation and theoretical developments are then presented, as well as the demands associated with the need for explicit consideration of modelling uncertainties. Finally, discussion is also given to the tools and databases needed for the efficient utilisation of simulated ground motions both in specific engineering projects as well as for near-real-time impact assessment.
Introduction
Earthquake-induced ground motion prediction is presently undergoing a paradigm shift from the empirical prediction of ground motion intensity measures (IMs, e.g. PGA, SA), based on regression analysis of observed IMs from past earthquakes, toward the use of physics-based simulation methods that directly predict the ground motion time series (i.e. multi-component acceleration as a function of time). This paradigm shift is presently occurring as a result of three key factors. Firstly, the diminishing returns offered from the continual efforts in empirical ground motion modelling, most evident in terms of the lack of any appreciable reduction in the standard deviation of IM prediction over four decades [13,44]. Secondly, recent well-recorded earthquakes (such as those discussed herein, among others) illustrate that, even now, physics-based simulation methods provide predictions that are comparable to, or even superior than, those from empiricallybased predictions [4,10,15,18,22]. Thirdly, the physics-based nature of such simulations provides a natural framework within which a substantially greater volume of data from seismological observations can be synthesised, enabling the incorporation of region and site-specific features, thus promising appreciable improvements in the ability to reduce prediction uncertainties in the coming years, and realising the flow-on benefits in the seismic design and assessment of built infrastructure [49]. It is important to appreciate that this empirical to physics-based modelling paradigm shift is akin to that which occurred in weather forecasting in the 1950’s – although the seismic problem is complicated relative to the weather problem because the salient phenomena occur beneath the earth’s surface, making direct observation challenging as compared to direct atmospheric observations.