High-resolution space–time processes for agents at the built–human interface of urban earthquakes

Abstract

Agent-based models (ABMs) have become popular media for simulating the space–time complexity of massively interactive phenomena, but usual approaches in geographical applications tend to emphasize the ability to deploy large numbers of agents in simulation rather than the richness of the behavior that agents can supply to enliven model processes. This paucity in representative detail can dilute the usefulness of ABMs when details are important in exploring what-if dimensions of the phenomena being considered, because abstraction can limit the range of questions that can be asked of a model. This is true, in particular, in multi-hazard earthquake models, where agents are now commonly used for representing humans, but often lack sufficient detail to reconcile them with other parts of the chained modeling pipeline. Increasing the spatial and temporal detail of these models is challenging, for a number of reasons, not least of which is the requirement that realistic process models must be developed for many objects, across many domains. In this paper, I present an innovative methodology for merging physical and human process models at exquisite spatiotemporal detail in a unified simulation meta-system. I demonstrate the usefulness of this approach is supporting broad experimentation with dynamics of building collapse and human response in urban areas.

Keywords:

• 3D visualization
• agent-based modeling
• geographic automata systems
• geosimulation
• spatiotemporal data modeling