The concept spatial simulation modelling can be better understood by looking at the meaning of its individual words. A model is widely defined as a simplified representation of a real-world system under study, which can be used to explore or to better understand the system it represents. Computer models or simulation models are computer-based implementations of a model to produce outputs based on certain model assumptions. Simulation , therefore, relies on the use of computers for virtual experimentation to gain insight into real-world problems by proposing alternative assumptions that arise from exploring “what if” questions about a dynamic problem of interest over the course of successive simulation experiments.
Simulation modelling is also useful for the study of spatial patterns over time. Spatial simulation models are relevant when the study of spatial elements and their relationships in a system are necessary for a fully understanding of that system. In this sense, spatial simulation modelling approaches include rule-based models, equation-based models, grid-based cellular automata models, discrete event simulation, and agent-based models.
Simulation modelling is often used for prediction, exploration, theory development, or even optimization of conditions to achieve desired outcomes, with the goal of examining how the interconnections and relationships that characterize these systems produces patterns of behavior over time. Across broad areas of the environmental and social sciences, researchers use simulation models as a way to study systems inaccessible to experimental and observational scientific methods, and also as an essential complement of those more conventional approaches to discover or formalize theories about the real world.
Simulation models are a relatively recent addition to the scientific toolbox, and the reasons for their widespread adoption are, on one hand, the impossibility to study in-situ some complex social and environmental systems (e.g. ecosystems, urban systems, social systems, global climate system) and, on the other hand, the availability of High Performance Computing and large amount of data from different sources. Finally, the nonlinear behaviour of many natural systems provides challenges building traditional mathematical models based on linearization.
Simulation modelling is also useful for the study of spatial patterns over time. In this sense, spatial simulation modelling approaches include rule-based models, equation-based models, grid-based cellular automata models, discrete event simulation, and agent-based models.