Explicit Modeling of Spatial Growth Patterns in Shama, Ghana: an Agent-Based Approach

The integration of geographic information systems (GIS) and agent-based models (ABMs) can be an efficient tool to improve spatial planning practices. This paper utilizes GIS and ABM approaches to simulate spatial growth patterns of settlement structures in Shama. A preliminary household survey on residential location decision-making choice served as the behavioural rule for the household agent while physical environment properties extracted from a 2005 image provided the inputs for the model. The resulting growth pattern model implemented in NetLogo was compared with empirical growth patterns to ascertain model accuracy. The paper establishes that the development of unplanned structures and its evolving structural pattern is a function of land price, proximity to economic centres, household economic status, and location decision-making patterns. The application of the proposed model underlines its potential for integration into urban planning policies and practices and for extending the understanding of residential decision-making processes in emerging cities in developing countries.