ON THE APPLICABILITY OF ONE-DIMENSIONAL CRUSTAL STRUCTURES FOR GROUND-MOTION SIMULATION

Ground motion simulation methods, such as the finite-difference method (FDM) or the modal summation technique, require a model of the crustal structure through which the seismic waves pass in terms of density, velocity and attenuation parameters, such as Q. The use of such structural models within simulations means that travel path effects, such as the constructive interference of different phases, can be modelled. Currently one-, two- or three-dimensional models are used without much description of the benefits of using a particular dimension of model. Within FDMs, a 2D or 3D structure can be assumed without a significant increase in computational time. One-dimensional models have the advantage of yielding results that are easier to interpret in terms of phases and also they require only one set of simulations for all considered horizontal source and site locations. In addition, the method introduced by Douglas et al. (2004) for the incorporation of the effect of crustal structure into empirical ground motion estimation equations is only practical for one-dimensional structures. The purpose of this article is to investigate when two-dimensional structures provide significantly different results than using an average one-dimensional model. The difference between the 2D and the derived 1D structure is quantified by a single parameter that seeks to characterise how two-dimensional the structure is. The maximum size this parameter can be before a 2D structure is required for accurate modelling of ground motions is assessed based on a series of FDM simulations. The purpose of this proposed technique is to provide guidance as to when 2D structures should be used or when 1D structures are sufficient, without the requirement to performing many simulations