CRAFTY_NetLogo

CRAFTY_NetLogo is a NetLogo implementation of the agent-based land-use modelling framework CRAFTY, originally developed in Java
https://github.com/CRAFTY-ABM/CRAFTY_v2/tree/master

This version uses a stylised 2D grid landscape and introduces a new behavioural extension that incorporates:

• Environmental attitudes
• Behavioural inertia
• Descriptive social norms
• Optional heterogeneous behavioural parameters from CSV files
• Optional dynamic, time-varying environmental attitudes

The model simulates how land managers (agents) adjust their land-use intensity in response to ecosystem service demand, local and teleconnected social networks, capital levels, and behavioural factors.
It illustrates how spatial land-use patterns emerge from intertwined economic, environmental, and socio-psychological processes.

---

Repository Structure

CRAFTY_NetLogo/
│
├── model/
│   └── CRAFTY.nlogo                 # NetLogo model file
│
├── data/
│   ├── ElevationData.csv            # Required: landscape elevation (101×101)
│   ├── MaxGivingInThresholdData.csv # Optional per-cell behavioural parameters
│   ├── NormWeightData.csv
│   ├── NormSensitivityData.csv
│   ├── InertiaData.csv
│   ├── AttitudeMeanTimeSeries.zip   # Optional dynamic attitude time series (ZIP)
│   └── (unzip to obtain AttitudeMeanTimeSeries.csv)
│
└── README.md

Model Description

This model follows CRAFTY’s core structure.

Agent Functional Types (AFTs)

Each patch represents a land manager belonging to one of three AFTs:

• High-Intensity
• Mid-Intensity
• Conservation

AFTs differ in:

• sensitivity to productive vs. natural capital
• production of material vs. non-material ecosystem services
• colour (dark green = High-Intensity, bright green = Conservation)

Decision-Making

At each tick, a subset of patches engages in competition.
Agents may switch AFT based on:

• Economic competitiveness (residual demand × production capacity)
• Environmental attitude
• Descriptive social norms, based on local + teleconnected peers
• Behavioural inertia
• A giving-in threshold, calculated with a logistic function

Behavioural Extensions

Per-patch behavioural heterogeneity (via optional CSV files)

Each land manager can have unique values for:

• environmental attitude
• norm weight
• norm sensitivity
• inertia
• giving-in threshold

Dynamic environmental attitudes (per-patch × time step CSV)

Environmental attitudes can evolve over time, enabling:

• behavioural trends
• shocks
• scenario pathways

---

CSV File Formats

Per-cell behavioural parameters

All behavioural CSVs must be:

• 101 rows × 101 columns
• same orientation as ElevationData.csv
• values in [0, 1]

Dynamic attitude time series

The dynamic CSV must be:

• 10,201 rows (one per patch, flattened row-major order: top-left → right → next row, etc.)
• T columns, one per time step (e.g., 300)
• values in [-1, 1]

---

Running the Model

1. Configure behavioural and environmental options
   (Choose attitude distribution, CSV loading options, behavioural parameters, production function, etc.)

2. Press SETUP
   Initialises capitals, behavioural parameters, AFTs, and networks.

3. Press GO
   Starts the simulation.

Plots display:

• global supply vs. demand of ecosystem services
• land-use intensity shares
• landscape connectivity metrics

---

Documentation

Additional documentation is available in the Model Info tab inside the .nlogo model, including:

• behavioural equations
• AFT specification
• capital → production mapping
• neighbourhood and teleconnection logic
• clustering & connectivity metrics
• CSV-handling mechanism

---

Related Framework

This model builds on the CRAFTY land-use modelling framework:
https://github.com/CRAFTY-ABM/CRAFTY_v2/tree/master