Linking Bayesian and agent-based models to simulate complex social-ecological systems in semi-arid regions

Aloah J. Pope, Randy Gimblett

Research output: Contribution to journalArticlepeer-review

25 Scopus citations

Abstract

Interdependencies of ecologic, hydrologic, and social systems challenge traditional approaches to natural resource management in semi-arid regions. As a complex social-ecological system, water demands in the Sonoran Desert from agricultural and urban users often conflicts with water needs for its ecologically-significant riparian corridors. To explore this system, we developed an agent-based model to simulate complex feedbacks between human decisions and environmental conditions in the Rio Sonora Watershed. Cognitive mapping in conjunction with stakeholder participation produced a Bayesian model of conditional probabilities of local human decision-making processes resulting to changes in water demand. Probabilities created in the Bayesian model were incorporated into the agent-based model, so that each agent had a unique probability to make a positive decision based on its perceived environment at each point in time and space. By using a Bayesian approach, uncertainty in the human decision-making process could be incorporated. The spatially-explicit agent-based model simulated changes in depth-to-groundwater by well pumping based on an agent's water demand. Changes in depth-to-groundwater feedback to influence agent behavior, as well as determine unique vegetation classes within the riparian corridor. Each vegetation class then provides varying stakeholder-defined quality values of ecosystem services. Using this modeling approach allowed us to examine effects on both the ecological and social system of semi-arid riparian corridors under various scenarios. The insight provided by the model contributes to understanding how specific interventions may alter the complex social-ecological system in the future.

Original languageEnglish (US)
Article number55
JournalFrontiers in Environmental Science
Volume3
Issue numberAUG
DOIs
StatePublished - Aug 7 2015
Externally publishedYes

Keywords

  • Agent-based modeling
  • Bayesian cognitive mapping
  • Hybrid modeling
  • Social-ecological systems
  • Sonoran desert

ASJC Scopus subject areas

  • General Environmental Science

Fingerprint

Dive into the research topics of 'Linking Bayesian and agent-based models to simulate complex social-ecological systems in semi-arid regions'. Together they form a unique fingerprint.

Cite this