Quantifying Irrigated Winter Wheat LAI in Argentina Using Multiple Sentinel-1 Incidence Angles.

Area of Science:

• Remote Sensing
• Agricultural Science
• Geospatial Analysis

Background:

• Synthetic Aperture Radar (SAR) offers all-weather Earth surface monitoring capabilities.
• Sentinel-1 (S1) satellite data provides suitable resolution and revisit time for cropland observation.
• Radar backscatter sensitivity to vegetation structure, phenology, soil moisture, and incidence angle is key.

Purpose of the Study:

• To develop and validate a workflow for estimating cropland Leaf Area Index (LAI) using Sentinel-1 SAR data.
• To exploit the dependency of radar backscatter on local incidence angle (LIA) for improved biophysical variable retrieval.
• To generate temporal LAI maps for irrigated winter wheat, supporting agricultural management.

Main Methods:

• A workflow merging multi-date S1 smoothed data acquired at distinct LIAs was developed.
• Gaussian Processes Regression (GPR) combined with a cross-validation (CV) strategy was employed for LAI estimation.
• The GPR-S1-LAI model was validated against in situ data from a 2020 winter wheat campaign.

Main Results:

• Adequate validation results for LAI estimation were achieved, with R^2_CV = 0.67 and RMSE_CV = 0.88 m^2 m^-2.
• The trained model successfully generated temporal LAI maps reflecting the crop growth cycle.
• Processing S1 imagery with distinct acquisition geometries enabled accurate radar-based LAI modeling.

Conclusions:

• The developed GPR-S1-LAI workflow accurately models Leaf Area Index in large irrigated areas.
• This approach supports agricultural management practices, especially in cloud-prone environments.
• Exploiting multi-angle S1 SAR data enhances the retrieval of crop biophysical variables.