[Comparison of precision in retrieving soybean leaf area index based on multi-source remote sensing data].

Area of Science:

• Agricultural remote sensing
• Plant science
• Geospatial analysis

Background:

• Remote sensing technology offers abundant data sources for agricultural monitoring.
• Accurate retrieval of crop biophysical parameters like Leaf Area Index (LAI) is crucial for precision agriculture.

Purpose of the Study:

• To analyze the retrieval accuracy of soybean LAI using multi-source remote sensing data.
• To compare the performance of ground hyperspectral, UAV multispectral, and Gaofen-1 (GF-1) WFV data for LAI estimation.
• To assess the influence of sensor characteristics on LAI retrieval accuracy.

Main Methods:

• Utilized ground hyperspectral, UAV multispectral, and GF-1 WFV data.
• Developed LAI retrieval models using various vegetation indices (RVI, NDVI, SAVI, DVI, TVI).
• Validated models using calibration accuracy and assessed estimation accuracy (R², RMSE).

Main Results:

• Models using ground hyperspectral and UAV multispectral data showed higher accuracy (R² > 0.69, RMSE < 0.4).
• Ground hyperspectral data slightly outperformed UAV multispectral data.
• GF-1 WFV data yielded the lowest accuracy (R² < 0.30, RMSE > 0.70), insufficient for field-scale monitoring.

Conclusions:

• UAV remote sensing provides a highly precise and efficient approach for soybean LAI retrieval.
• Ground hyperspectral data offers advantages but is not significantly superior to multispectral data.
• Agricultural UAV remote sensing is a key information resource for precision agriculture and field-scale crop management.