MASON AfriLand: A Regional Multi-Country Agent-Based Model with Cultural and Environmental Dynamics

Abstract

Agent-based models of regions of the international system composed of several countries are few and not as advanced as other classes of spatial computational models on a comparable scale. Most relevant extant models are of a single country or polity, or they model an entire international system putatively comprising all countries as in a world system. Here we present AfriLand, a new agentbased model for developing a large-scale and more detailed model of the geographic region of contemporary Eastern Africa. AfriLand is part of the Mason-HRAF Joint Project on Eastern Africa funded by a multi-year ONR-MURI grant. We present the motivation and challenges behind the AfriLand agent-based model design and a description of the model architecture and dynamics. AfriLand offers a useful scale for analyzing socio-cultural and environmental dynamics that transcend national boundaries, such as refugee flows, transnational conflict and crime (narcotics, trafficking in persons, smuggling) and natural hazards across national frontiers. AfriLand also presents visualization challenges that call for a range of solutions from software to hardware.