Rice Quality and Yield Prediction Based on Multi-Source Indicators at Different Periods.

Area of Science:

• Agricultural Science
• Remote Sensing
• Plant Physiology

Background:

• Modern agriculture demands rapid, non-destructive methods for assessing crop quality and yield.
• Accurate prediction of rice quality indices and yield is crucial for optimizing agricultural practices and ensuring food security.

Purpose of the Study:

• To develop an effective and reliable method for estimating rice quality indices and yield using spectral reflectance.
• To evaluate the predictive accuracy of various spectral indicators and regression models across different rice growth stages.

Main Methods:

• Field experiments were conducted with rice variety Longqingdao 3.
• Measurements included leaf area index (LAI), chlorophyll content (SPAD), leaf nitrogen content (LNC), and spectral reflectance.
• Univariate linear regression models were developed using spectral indicators to predict quality indices and yield.

Main Results:

• Optimal R² values for brown rice rate, moisture content, and taste value were 0.866, 0.913, and 0.651, respectively.
• Optimized models improved R² for brown rice rate to 0.95 and taste value to 0.992.
• The spectral index GM2 during the jointing stage achieved the highest yield prediction accuracy (R² = 0.822).

Conclusions:

• Integrating multiple spectral indicators across different growth periods significantly enhances the accuracy of rice quality and yield predictions.
• The developed spectral-based method offers a robust and intelligent solution for practical agricultural applications.
• This approach supports precision agriculture by enabling timely and accurate crop assessment.