EFForTS-LGraf: A landscape generator for creating smallholder-driven land-use mosaics

Area of Science:

• Ecological modeling
• Socio-economic systems analysis
• Geospatial data science

Background:

• Spatially-explicit simulation models require detailed input data, like land-cover maps, for ecological and socio-economic research.
• Existing landscape generators often lack the human dimension crucial for understanding land-use change in smallholder systems.

Purpose of the Study:

• To introduce EFForTS-LGraf, a process-based landscape generator for creating artificial land-use maps of agricultural landscapes.
• To incorporate the human dimension of land-use change by simulating smallholder farming households, their strategies, and crop distributions.
• To validate the generated maps against real-world data and demonstrate the tool's utility in simulation modeling and landscape analysis.

Main Methods:

• EFForTS-LGraf generates roads, villages, and fields based on smallholder farming household strategies.
• Crop types are distributed using crop fractions and specialization levels, with parameters derived from surveys or geospatial data.
• Model performance was validated using Indonesian oil palm and rubber farming landscapes, comparing generated maps with classified satellite imagery.

Main Results:

• EFForTS-LGraf successfully generates realistic land-cover maps with properties consistent with classified satellite images.
• Simulations showed that larger households and higher crop specialization lead to more homogeneous landscapes with reduced edge areas.
• The model's parameters can be informed by empirical data, offering an advantage over abstract neutral landscape generators.

Conclusions:

• EFForTS-LGraf provides a valuable tool for generating realistic input data for simulation models in ecological-economic studies.
• The generator can also serve as a standalone tool for landscape-scale analyses of smallholder farming systems.
• The model explicitly integrates human decision-making into land-use change simulations, enhancing ecological and socio-economic research.