Agent-based models (ABM) are becoming standard tools to study complex systems especially in ecology and more recently in epidemiology. The developments of models in these domains have highlighted the need for more complex representations of the environment. In this paper, we present an individual-based epidemiological model that has the particularity to make heavy use of geographic data and complex spatial operations. From this example, we review several popular ABM simulation platforms. These platforms do not answer such model's requirements in a whole. In order to answer this problem, we propose a new approach to represent the environment in ABM. This approach along numerous spatial operations tools has been implemented in the GAMA simulation platform, which is described in this paper.