Keywords

social-ecological interaction; equation-free; steady state analysis; regime shifts; shallow lake restoration

Location

Session B3: Methods for Visualization and Analysis of High-Dimensional Simulation Model Outputs

Start Date

12-7-2016 5:10 PM

End Date

12-7-2016 5:30 PM

Abstract

Conventional tools to analyze regime shifts in ecosystems, e.g. stable state and bifurcation analyses, are increasingly adapted to investigate social-ecological systems. The challenge lies in combining the analysis of system dynamics with the analysis of transient dynamics emerging from adaptive and heterogeneous agents. A typical question is, to which extent do micro level interactions contribute to a macro level outcome? We demonstrate the application of a new tool, equation-free analysis, for evaluating the state space from an agent-based model (ABM) on lake restoration. By sidestepping the requirement of an equation describing a systems macroscopic behavior, equation free analysis enables system level tasks (such as bifurcation analysis) to be performed on micro level models. This is particularly beneficial for agent-based models that aim to explain macro level patterns and include more parameters than one can feasibly analyze by simulation experiments alone. In our example on lake restoration, the macro scale pattern is the ecological regime shift between the clear and turbid water state. This regime shift can be influenced by micro level decisions from lake users affecting the main driver through alternative strategies. To find out which strategies are more effective to restore the lake into the desired clear state, we linked the resulting bifurcation diagram from the equation-free analysis to simulation experiments along the alternative trajectories. Concluding, we describe suitable steps for integrating the equation-free analysis with more traditional agent-based model analyses and discuss the difficulties and advantages therefrom.

COinS
 
Jul 12th, 5:10 PM Jul 12th, 5:30 PM

Analyzing regime shifts in agent-based models with equation-free analysis

Session B3: Methods for Visualization and Analysis of High-Dimensional Simulation Model Outputs

Conventional tools to analyze regime shifts in ecosystems, e.g. stable state and bifurcation analyses, are increasingly adapted to investigate social-ecological systems. The challenge lies in combining the analysis of system dynamics with the analysis of transient dynamics emerging from adaptive and heterogeneous agents. A typical question is, to which extent do micro level interactions contribute to a macro level outcome? We demonstrate the application of a new tool, equation-free analysis, for evaluating the state space from an agent-based model (ABM) on lake restoration. By sidestepping the requirement of an equation describing a systems macroscopic behavior, equation free analysis enables system level tasks (such as bifurcation analysis) to be performed on micro level models. This is particularly beneficial for agent-based models that aim to explain macro level patterns and include more parameters than one can feasibly analyze by simulation experiments alone. In our example on lake restoration, the macro scale pattern is the ecological regime shift between the clear and turbid water state. This regime shift can be influenced by micro level decisions from lake users affecting the main driver through alternative strategies. To find out which strategies are more effective to restore the lake into the desired clear state, we linked the resulting bifurcation diagram from the equation-free analysis to simulation experiments along the alternative trajectories. Concluding, we describe suitable steps for integrating the equation-free analysis with more traditional agent-based model analyses and discuss the difficulties and advantages therefrom.