Enhancing Leaf Area Index Estimation for Maize with Tower-Based Multi-Angular Spectral Observations

Area of Science:

• Earth and Environmental Sciences
• Remote Sensing
• Agricultural Science

Background:

• Leaf Area Index (LAI) is critical for ecological, hydrological, and climate models.
• Normalized Difference Vegetation Index (NDVI) is commonly used for LAI estimation but suffers from saturation and soil background interference.
• Existing vegetation indices (VIs) have limitations in accurately estimating LAI under dense or sparse vegetation conditions.

Purpose of the Study:

• To develop and evaluate a Multi-Angular NDVI (MAVI) for enhanced LAI estimation.
• To minimize soil background effects and saturation issues inherent in traditional NDVI.
• To leverage tower-based multi-angular spectral observations for improved LAI retrieval.

Main Methods:

• Collected continuous tower-based multi-angular reflectance and LAI data over three years in maize cropland.
• Developed the MAVI by analyzing canopy reflectance variations with solar zenith angle (SZA).
• Quantitatively evaluated MAVI performance against eight other VIs using statistical tests.

Main Results:

• MAVI showed an improved curvilinear relationship with LAI after correcting NDVI with multi-angular data (R² = 0.945, RMSE = 0.345, rRMSE = 0.147).
• The MAVI-based model effectively mitigated soil background effects in sparse vegetation (R² = 0.934, RMSE = 0.155, rRMSE = 0.157).
• MAVI demonstrated superior performance compared to eight other vegetation indices.

Conclusions:

• Tower-based multi-angular spectral observations are valuable for accurate LAI retrieval.
• MAVI offers a robust method for overcoming limitations of traditional VIs in LAI estimation.
• This research enhances the application of multi-angular observations and provides data for validating space-borne LAI products.