Development of a behaviour pattern-based testing approach for coupled socioeconomic and environmental models.

Area of Science:

• Environmental modeling
• Socioeconomic systems analysis
• Sustainable environmental management

Background:

• Integrated models (Dynamically Coupled Socioeconomic system dynamics models integrated with physically-based Environmental Models - DCSEMs) are crucial for understanding human-environment interactions.
• Existing model testing methods focus on point-to-point analysis, which is inadequate for behavior pattern-oriented DCSEMs.
• A lack of behavior pattern-based testing hinders the adaptability and reliability of DCSEMs.

Purpose of the Study:

• To propose and demonstrate a novel behavior pattern-based model testing approach for DCSEMs.
• To enhance the evaluation of integrated models by assessing similarities between model outputs and real-world trends.
• To improve the performance and adaptability of DCSEMs for sustainable environmental management.

Main Methods:

• Global sensitivity analysis (GSA)
• Auto-calibration algorithms
• Evaluation of behavior pattern similarities between model outputs and real-world trends
• Theil inequality statistical analysis and parameter distribution analysis

Main Results:

• The proposed behavior pattern-based approach effectively calibrates and evaluates a DCSEM for water table depth simulation.
• This method better replicates observed system behavior patterns over time compared to conventional point-based approaches.
• The approach successfully captures spatial heterogeneity, improving DCSEM performance in data-limited, spatially-distributed scenarios.

Conclusions:

• The novel behavior pattern-based testing procedure is well-suited for data-limited, spatially-distributed DCSEMs.
• This approach enhances the reliability and adaptability of integrated models for sustainable environmental management.
• The study provides a robust framework for testing and improving complex socioeconomic and environmental models.