If you are studying any aspect of human-environment interactions, or interested in applying complex systems theory with agent-based modelling (ABM) techniques to your research, please join our weekly discussion group. We have been running a student workshop for a few terms, and we will continue meeting this term!
Coupled human and natural systems (CHANS) are an integrated framework from which many urgent questions such as sustainability, climate change, risk management, social innovation, and human well-beings, can be answered. As a complex system it has non-linear dynamics with thresholds, feedback, heterogeneity, and cross-scale effects, which often result in surprising patterns.
ABM is a powerful tool to represent individual behaviours and to form aggregated outcomes that traditional methods often neglect. It provides insights and knowledge of CHANS by investigating the interactions between human actors and the environment.
The previous focus of our working group is how to use ABM in CHANS, and we decide to move beyond ABM implementation to the more general topic of the complexity of CHANS with an emphasis on ABM this term.
Here is some additional information about this group and how to participate:
- Who can participate: Any graduate/undergraduate/postdoc from a wide range of research fields, including geography, biology, urban planning, transportation system design, economics, computer science, and theoretical physics.
- Meeting time for winter 2016: Fridays, 12-1 pm.
- Location for winter 2016: Environment 3 (EV3) room 4268
- Meeting format: 20-minute presentation from one participant followed by 40-minute discussion.
- How to join: Please contact firstname.lastname@example.org and identify your field, academic level, and research interests with CHANS or ABM.
Examples of CHANS and ABM:
- Complexity of Coupled Human and Natural Systems, Jianguo Liu et al. Science 317, 1513 (2007)
- Spatial Land Use Change and Ecological Effect (SLUCE) project
- NetLogo (a popular ABM platform)
- Digging into Data (DiD) research: MIRACLE
- Segregation model: One of the easiest and earliest models to explain “tipping” in macro behaviours in social situations from a large quantity of individuals. Micromotives and Macrobehavior, Schelling 1969.
- John Conway’s Game of Life: A small set of simple rules on a check-board that generates complex pattern.
- Sugarscape model, Growing Artificial Societies: Social Science from the Bottom up, Epstein & Axtell, 1996. One of the best demonstration of agent-based simulation that includes diverse activities as trade, combat, mating, culture, and disease.
ABMs – additional information and useful links and resources
Agent-based models (ABMs) are simulation models that have been used to study complex systems in a wide range of academic fields including biology, computer science, environmental science, economics, geography, planning, and sociology among many other fields. Agent-based models, by directly representing individual actors (agents) and simulating the interactions and feedback among them, provide new level of detail and insight into the non-linear processes that are often neglected by traditional methods.
Railsback and Grimm (2011). Agent-based and Individual-based Modeling: A Practical Introduction. Princeton University Press. This book uses NetLogo for examples.
Journals and working papers
CASA working paper series: Nearly 200 working papers from the Bartlett Centre for Advanced Spatial Analysis, University College London, including many articles on ABM in geo-spatial simulation.
SIMSOC: A mailing list about computer simulation in social sciences.
NetLogo: NetLogo is a very popular ABM platform, using a relatively simple programming language named Logo. The software is free, and comes with lots of demo models on different topics and academic fields.
Repast Simphony: Repast is one of the most popular ABM platforms. It’s based on the Java programming language, and it’s built on the basis of popular Java packages such as Geotools (for geospatial support), JUNG (for network support), Colt (for random number generation), Log4J (for data logging), JOGL (for display) and NASA World Wind (for 3D GIS display). Repast Simphony also comes with many demo models.
Introduction to Complexity by Santa Fe Institute. This is on complexity in general, not specific to ABM models.
Online model library
CoMSES Computational Model Library: Contains source code and data for many ABM models. The website also provides many useful articles and links to ABM software, journals and tutorials.
Multi Agent Systems and Simulation (MASS) Research Group: A collection of ABM videos (including both ABM introduction/lecture videos, and video illustrations of ABM models).