Towards an Automated Workflow for the Creation of Complex 3D Fault Models.

Abstract

Fault geometry and the connectivity between faults at depth are both important controls on earthquake behaviour, so modelling these parameters accurately is essential to models of the earthquake cycle. However, the effects of uncertainties in fault geometry and connectivity are often not fully explored, partly because using current techniques building a fault model is difficult and time consuming. For example, the current 3D Community Fault Model — which results from several hundred person hours of work — represents only one of many hypothetical fault models that would be consistent with available constraints.

We present a preliminary automated workflow for the creation of 3D models of faults in Aotearoa New Zealand and worldwide, using python and the mesh-cutting capabilities of Leapfrog Software. This workflow creates a 3D fault model from: (1) GIS fault traces; (2) dip estimates; and (3) a text file containing information on which faults are thought to terminate against each other. Our approach is faster, more internally consistent and much less labour-intensive than previous (mainly manual) methods of fault model creation; it will allow a thorough exploration of the sensitivity of models to fault geometry and will therefore be of use in several diverse areas of earthquake science.