Agent-Based Modeling and GIS

Increasingly we see interests in the sciences for understanding bottom-up driven social, ecological, and social-ecological processes utilizing concepts of complexity and complex adaptive systems.[1] The need to understand emergent phenomenon in a variety of fields has led to not only greater use of agent-based models (ABMs), but we are increasingly seeing tools that integrate GIS with ABMs. One such tool is the Repast Simphony suite of tools, which integrate open source GIS, specifically GeoTools, and WorldWind 3D visualizer. The Java platform also allows one to integrate with other potential tools, including GRASS and QGIS. The use of ABMs within GIS is now allowing geospatial analysis of not only emerged phenomenon, but it allows the integration of spatial processes. Recent papers on ecology, social upheaval, traffic patterns, and others are just some examples of how GIS is being integrated with ABMs.[2] The new area of advanced applications increasingly includes the utilization of high performance computing (HPC) with GIS and ABMs. This now allows very large geospatial problems to be addressed using more realistic population and numbers of individuals with a complexity perspective. Within GIS and ABMs we see network analysis, spatial interaction modeling, and least cost pathway analysis used among many other methods. Post-simulation runs from ABMs are also increasingly analyzed using spatial statistical techniques such as spatial autocorrelation.

With the merger of GIS with ABMs, we now have also better way to deal with issues of time, as many simulations are event driven simulations that respondent to changes in systems rather than even time steps. Overall, this gives a way for GIS to realistically address time and events shaping how systems evolve or emerge from given states. Looking to the future, the open source platforms are likely to have great benefit to ABM tools, as these often more easily integrate with existing mainstream ABM platforms. This should mean we should see a substantial increase in research use.

References 

[1] For more information on the topic of GIS and ABMs, see:  Arifin, S. M. N., Madey, G. R., & Collins, F. H. (2016). Spatial agent-based simulation modeling in public health: design, implementation, and applications for malaria epidemiology. Hoboken, New Jersey: John Wiley & Sons Inc.

[2] For examples of recent research papers on GIS and ABMs, see: Bone, C., & Altaweel, M. (2014). Modeling micro-scale ecological processes and emergent patterns of mountain pine beetle epidemics. Ecological Modelling, 289, 45–58. http://doi.org/10.1016/j.ecolmodel.2014.06.018. Also, see: Altaweel, M., Sallach, D., & Macal, C. (2013). Mobilizing for Change: Simulating Political Movements in Armed Conflicts. Social Science Computer Review, 31(2), 143–164.