Estimation of regional-scale maize plant nitrogen content based on multi-source remote sensing data

Area of Science:

• Agricultural Science
• Remote Sensing
• Agronomy

Background:

• Maize production in arid regions faces challenges from water scarcity and low nitrogen use efficiency.
• Efficient water and fertilizer management are crucial for sustainable agriculture in these areas.

Purpose of the Study:

• To systematically analyze challenges in maize production in the Hexi Corridor.
• To develop an innovative, data-driven framework for monitoring maize Plant Nitrogen Content (PNC).
• To provide a scientific basis for efficient water and fertilizer management in arid agricultural regions.

Main Methods:

• Integration of multi-source remote sensing data (Sentinel-2A, UAV multispectral) and ground-based observations.
• Development of a comprehensive data fusion framework and a novel band-corrected PNC inversion method.
• Application of Convolutional Neural Networks (CNN) for PNC prediction and comparison with traditional machine learning models (SVM, RF).

Main Results:

• The CNN-based PNC prediction model achieved a high coefficient of determination (R²) of 0.80, outperforming SVM and RF by over 10%.
• Band correction significantly improved Sentinel-2A data performance for PNC retrieval, increasing R² from 0.35-0.45 to 0.70-0.80.
• The model demonstrated high accuracy and reliability across different validation areas and growth stages, confirming its robustness.

Conclusions:

• The proposed method offers an effective approach for rapid and precise monitoring of maize nitrogen status in arid regions.
• Findings provide scientific support for regional-scale crop nitrogen management and precision fertilization decisions.
• The study highlights the potential of integrated remote sensing data and advanced modeling for sustainable agriculture in water-scarce environments.