Simulating Human Agropastoral Activities
Using Hybrid Agent-Landscape Modeling: A Multidisciplinary Approach
H. Sarjoughian, G. Mayer, E. Allen
Mediterranean Landscape Dynamics Project (MEDLAND)

Objectives:

• Studies the effects of agropastoral activities on landscapes and societies from the
Neolithic through the Bronze Age in the Mediterranean Basin.

Investigate three primary themes of human landuse:

• International, interdisciplinary team: researchers from the United States, Spain, and
Jordan in the disciplines of Anthropology, Geology, Geography, and Computer Science
and Engineering are collaboratively studying the long-term dynamics of human land use
in the Mediterranean Basin.

Human Model
• A human household is represented as an agent
• Households are grouped into villages
• Households have goals, requirements based upon
population, and the ability to manage some resources
• Household actions impact the surrounding landscape
(e.g., deforesting to plant crop increases soil erosion)

Interaction Model
• Relations between agents and landscape dynamics
are represented in a rigorous and flexible fashion
• Connects agent and landscape models from both the
formalism and software viewpoints
• Unified user interface for simulation control and data
display

1.The effects of growth in agropastoral economies on biodiversity,
2.Subsequent land use intensification and diversification and its impacts
on landscape vulnerability and resilience, and
3.Studying the sustainability of human-maintained ecosystems.

Approach: Model and simulate agent and landscape
dynamics separately and synthesize them to
understand their complex interactions.

agent

Challenges

interactions

•Correctness

Agent DEVS Models

•Resolution vs Scale
•Computing Resources
•Synchronization

Environment Model
• Landscape and climate elements (e.g., soil, slope, and
precipitation data)
• Landscape / climate dynamics
• Changing environment impacts the agents (e.g.,
decreased soil quality produces a reduced crop yield)

landscape

•Visualization
Data Transformation +
Control Models

Cellular GRASS Models

Supported by the National Science Foundation ERE Biocomplexity in the Environment program

http://www.asu.edu/clas/shesc/projects/medland/
Collaborating Schools at ASU: Earth and Space Exploration, Geographical Sciences, and Human Evolution & Social Change