Assessing temporal-spatial land use simulation effects with CLUE-S and Markov-CA models in Beijing

Area of Science:

• Geographic Information Systems (GIS)
• Spatial Analysis
• Land Use Change Modeling

Background:

• Extensive land use in rural residential areas poses challenges to farmland protection and sustainable rural development.
• Understanding the temporal and spatial evolution of rural settlements is vital for effective land management.

Purpose of the Study:

• To analyze and compare the efficiencies of the Conversion of Land Use and its Effects at Small regional extent (CLUE-S) model and the Markov-Cellular Automata (Markov-CA) model in simulating rural residential land use.
• To evaluate the predictive accuracy of both models for rural residential land use in Beijing.

Main Methods:

• Utilized 14 driving factors to simulate and predict land use in Beijing for the years 2000 and 2005 using both CLUE-S and Markov-CA models.
• Employed a grid size of 220m x 220m for spatial analysis.
• Assessed simulation and prediction accuracy using accuracy rates and Kappa coefficients.

Main Results:

• The CLUE-S model, based on logistic regression, showed a higher accuracy in capturing discrete land use changes.
• The Markov-CA model, considering neighborhood effects, is more suited for land uses with nearest-neighbor diffusion, like urban land.
• CLUE-S model achieved correct prediction rates of 60.99% (2000) and 81.35% (2005), outperforming Markov-CA's 51.33% (2000) and 73.68% (2005) for rural residential areas.

Conclusions:

• The CLUE-S model provides more accurate simulation and prediction results for rural residential land use evolution compared to the Markov-CA model.
• The findings support the use of the CLUE-S model for better land use planning in rural areas.
• Effective modeling of rural residential land use is essential for promoting new socialist countryside construction and farmland protection.