Vol. 13, Issue 9 (2024)

Author(s):
Selvaprakash Ramalingam, Jagadeeswaran Ramasamy and PV Naga Sindhuja

Abstract:
The present study was conducted in Perambalur district, Tamil Nadu from September 2018 to January 2019 and soil moisture mapping was accomplished using Sentinel-1A Synthetic Aperture Radar (SAR) data. Ground truth measurements were carried out to estimate the soil moisture employing a gravimetric method concurrently with satellite passes. Standard SAR data processing techniques were applied to derive backscattering coefficients (σ0) for both VV and VH polarizations. These coefficients were then used, along with local incidence angle, to empirically estimate soil moisture. The results revealed that low backscattering coefficients in September 2018 and January 2019 indicating drier conditions, whereas higher values were observed during the northeast monsoon season in October, November, and December 2018, indicating increased soil moisture content. Multiple linear regression analysis was performed to establish relationships between soil moisture, backscattering coefficients, and local incidence angle for both VV and VH polarizations. In September 2018, VV-derived soil moisture ranged from 2.27% to 7.14%, while VH-derived moisture was higher at 26.74%. October 2018 exhibited a wider range of soil moisture, with VV-derived values ranging from 4.85% to 29.17%, and VH-derived values ranging from 5.22% to 26.73%. Multiple linear Regression is highly related with observed and predicted soil moisture on both VV and VH polarization. This study demonstrates that SAR data, particularly VV polarization, can be effectively utilized to map and monitor soil moisture with a high degree of reliability. These findings have significant implications for assessing agricultural drought and monitoring soil moisture conditions in near real-time.