Bare Soil Surface Moisture Retrieval from Sentinel-1 SAR Data Based on the Calibrated IEM and Dubois Models Using...

Area of Science:

• Remote Sensing
• Geophysics
• Agricultural Science

Background:

• Radar backscattering models are crucial for analyzing Earth's surface.
• Accurate soil moisture estimation is vital for agricultural management and hydrological studies.
• Assessing model performance and retrieval techniques using satellite data is an ongoing research area.

Purpose of the Study:

• To evaluate the performance of the calibrated integral equation model (CIEM) and the modified Dubois model (MDB) for radar backscattering over an agricultural region in Karaj, Iran.
• To estimate soil surface moisture using CIEM and MDB with single-polarization Sentinel-1 images via neural networks.
• To determine the effectiveness of different polarizations for soil moisture retrieval.

Main Methods:

• Field measurements were used to validate the performance of CIEM and MDB.
• Neural networks (NNs) were employed for inverting soil moisture from backscattering coefficients.
• Single-polarization Sentinel-1 Synthetic Aperture Radar (SAR) data were utilized for soil moisture estimation.

Main Results:

• The calibrated integral equation model (CIEM) achieved a lower root mean square error (RMSE) of 0.78 dB compared to the modified Dubois model (MDB) with 1.45 dB.
• Neural network-based inversion demonstrated the efficiency of single-polarized Sentinel-1 data for retrieving soil surface moisture.
• VV polarization proved particularly effective for soil moisture retrieval.

Conclusions:

• The calibrated integral equation model (CIEM) is a more accurate model for radar backscattering over agricultural areas compared to MDB.
• Single-polarized Sentinel-1 data, especially VV polarization, can be effectively used with neural networks for soil moisture estimation.
• The findings support the use of radar remote sensing for agricultural monitoring and water resource management.