Remotely Estimating Aerial N Uptake in Winter Wheat Using Red-Edge Area Index From Multi-Angular Hyperspectral Data.

Area of Science:

• Agricultural remote sensing
• Plant nutrient monitoring
• Spectroscopy and spectral indices

Background:

• Non-destructive, rapid detection of wheat nitrogen (N) status is crucial for efficient agricultural management.
• Existing vegetation indices (VIs) show variable stability in estimating N uptake under different viewing conditions.
• Multi-angular canopy data offers potential for improved nutrient status assessment.

Purpose of the Study:

• To evaluate the stability and accuracy of selected VIs and develop a novel index for estimating winter wheat aerial N uptake using multi-angular remote sensing data.
• To investigate the influence of view zenith angles (VZAs) on the estimation of N status.
• To develop a simple, applicable model for guiding N management.

Main Methods:

• Collected multi-angular canopy spectral data across diverse growing seasons, locations, years, wheat varieties, and N application rates.
• Assessed 17 VIs and developed a novel modified right-side peak area index (mRPA) based on spectral area calculation and red-edge features.
• Analyzed relationships between VIs, mRPA, and aerial N uptake at various VZAs, focusing on back-scatter angles and developing predictive models.

Main Results:

• Back-scatter angles generally provided better VI performance than forward-scatter angles, with correlations improving at decreasing VZAs.
• The novel mRPA index demonstrated superior predictive accuracy for aerial N uptake, achieving R² = 0.804 at -10° VZA.
• Models using mRPA, DIDA, and DDn showed good performance across a wide angle range (-20° to +10°), with mRPA exhibiting the lowest relative error (12.6%).

Conclusions:

• The novel mRPA index offers a more accurate and reliable assessment of winter wheat N status compared to existing indices.
• Multi-angular remote sensing, particularly at specific VZAs, combined with optimized indices like mRPA, is effective for N management.
• The developed models enhance the simplicity and applicability of remote sensing for guiding in-field N application.