Zhou et al. (2025) Bayesian-factorial analysis for unveiling multi-factor interactive effect on water demand in Central Asia

Identification

• Journal: Environmental Modelling & Software
• Year: 2025
• Date: 2025-11-30
• Authors: Yanxiao Zhou, Yongping Li, Zhenyao Shen, Yufei Zhang
• DOI: 10.1016/j.envsoft.2025.106806

Research Groups

• State Key Laboratory of Regional Environment and Sustainability, School of Environment, Beijing Normal University, Beijing, 100875, China
• Institute for Energy, Environment and Sustainable Communities, University of Regina, Regina, Sask, S4S 7H9, Canada

Short Summary

This study develops an integrated Bayesian support vector machine-based two-step factorial analysis (BSVM-TFA) method to reveal the individual and interactive effects of human activities on water demand, applying it to Central Asia to project future water demand and identify key influencing factors.

Objective

• To develop a new hybrid method (BSVM-TFA) to identify the key human activity factors influencing water demand and to forecast future water demand in Central Asia, providing scientific evidence for sustainable water management.

Study Configuration

• Spatial Scale: Central Asia
• Temporal Scale: Projections by 2050

Methodology and Data

• Models used: Integrated Bayesian support vector machine-based two-step factorial analysis (BSVM-TFA)
• Data sources: Not explicitly detailed, but implies socio-economic and environmental data related to human activities (e.g., GDP, agricultural irrigation efficiency).

Main Results

• By 2050, water demand in Central Asia is projected to range from 75.66 × 10^9 m^3 to 113.23 × 10^9 m^3 under different scenarios.
• This indicates an uncertainty of approximately 33.18 % in water demand driven by human activities.
• The key factors influencing water demand are Gross Domestic Product (GDP) and agricultural irrigation efficiency (AIE), with a combined contribution of 47.98 %.
• Water demand would be reduced by 16.42 × 10^9 m^3 with low-growth GDP and increasing AIE.

Contributions

• Advances an integrated BSVM-TFA method capable of capturing complex nonlinear relationships and identifying individual and interactive effects of multiple factors on water demand, while preventing overfitting through Bayesian optimization.
• Provides scientific evidence and decision support for reducing water demand, narrowing the supply-demand gap, and alleviating regional water scarcity in Central Asia.

Funding

• Not explicitly mentioned in the provided text.

Citation

@article{Zhou2025Bayesianfactorial,
  author = {Zhou, Yanxiao and Li, Yongping and Huang, Guohe and Shen, Zhenyao and Zhang, Yufei},
  title = {Bayesian-factorial analysis for unveiling multi-factor interactive effect on water demand in Central Asia},
  journal = {Environmental Modelling & Software},
  year = {2025},
  doi = {10.1016/j.envsoft.2025.106806},
  url = {https://doi.org/10.1016/j.envsoft.2025.106806}
}