Crop water productivity assessment and planting structure optimization in typical arid irrigation district using...

Area of Science:

• Agricultural Science
• Remote Sensing
• Artificial Intelligence

Background:

• Enhancing crop water productivity is vital for agricultural sustainability and water resource management, especially in arid environments.
• Data limitations pose challenges in assessing the spatial and temporal variability of crop water productivity.

Purpose of the Study:

• To develop and validate a framework for assessing and optimizing crop water productivity under water scarcity.
• To determine optimal crop planting structures for the Qira oasis, balancing productivity, food security, and ecological benefits.

Main Methods:

• Integration of remote sensing data with a time series generative adversarial network (TimeGAN) and a dynamic Bayesian network (DBN).
• Development of an optimization model to guide crop planting structure adjustments.
• Analysis of spatial crop suitability and temporal prediction accuracy.

Main Results:

• The TimeGAN-DBN model achieved high accuracy (R² > 0.8) for dynamic crop water productivity prediction, with optimal short-term prediction at a 4-year timescale.
• Spatial analysis identified suitable cultivation areas for wheat, corn, and cotton, with specific regions unsuitable for certain crops like cotton and walnuts.
• Optimized planting structures increased crop water productivity by 14.97% and ecological benefits by 3.61% without a proportional rise in irrigation water consumption.

Conclusions:

• The proposed framework effectively enhances crop water productivity and ecological benefits through optimized planting structures in water-limited arid regions.
• The integrated approach provides a valuable decision-support tool for sustainable agricultural planning and water resource management.
• The findings highlight the potential for significant improvements in agricultural sustainability through advanced data integration and AI techniques.