Crop classification by using dual-pol SAR vegetation indices derived from Sentinel-1 SAR-C data

Area of Science:

• Earth Observation
• Agricultural Science
• Remote Sensing

Background:

• Accurate crop classification is vital for agricultural monitoring and food security.
• Time-series Sentinel-1 Synthetic Aperture Radar (SAR) data offers consistent monitoring capabilities, even through cloud cover.

Purpose of the Study:

• To classify rainy season crops using time-series Sentinel-1 SAR data.
• To evaluate the effectiveness of the SVI_DP index for differentiating crop types.

Main Methods:

• Analysis of Sentinel-1 SAR data (dual-pol VV and VH bands) from May to September 2020.
• Utilization of the SVI_DP index, incorporating NRPB, DPDD, IDPDD, and VDDPI ratios.
• Application of Random Forest (RF) and Support Vector Machine (SVM) classifiers for crop classification.

Main Results:

• High classification accuracies were achieved: 93.77% for RF and 93.50% for SVM.
• The IDPDD index demonstrated high sensitivity to crop variations, showing strong correlations with SAR bands and other indices.
• The IDPDD index showed minimal correlation with water bodies, aiding in their differentiation.

Conclusions:

• Time-series Sentinel-1 SAR data combined with the SVI_DP index is effective for rainy season crop classification.
• The IDPDD index is a valuable component for distinguishing between crop types and water bodies.