TITLE:
Estimating Soil Moisture at Different Depths Using Multiple Satellite-Derived Indices
AUTHORS:
Eniola E. Olakanmi, Lydia E. Ebbuah, Souleymane Fall, Joseph E. Quansah
KEYWORDS:
Soil Moisture, Landsat, Sentinel, NDMI, MSI, Remote Sensing, Soil Sensor
JOURNAL NAME:
Advances in Remote Sensing,
Vol.15 No.2,
June
4,
2026
ABSTRACT: Understanding soil moisture dynamics at different depths is essential for irrigation management. This study evaluated the relationship between two satellite-derived moisture indices—the Normalized Difference Moisture Index (NDMI) and the Moisture Stress Index (MSI) and in-situ soil moisture measurements across the Alabama Black Belt region during the growing season. Soil moisture readings at depths of 15, 20, 25, and 30 cm were collected at five sensor locations between August 2023 and October 2025. Landsat 8/9 and Sentinel-2 imagery were processed to generate the moisture indices, which were then correlated with the sensor measurements. The results showed that, although both satellite data types had similar predictive performance for soil moisture at all depths, there were weak linear correlations at 15 - 20 cm that substantially increased at 25 - 30 cm. Landsat 8/9 soil moisture estimates were most accurate at 30 cm for both NDMI and MSI indices, whereas Sentinel-2 estimates were optimal at depths of between 25 and 30 cm. Both indices produced optimal correlations in the range of r ≈ 0.60 - 0.64. These results demonstrate that satellite-based moisture indices exhibit statistically significant relationships with subsurface soil moisture conditions, particularly at depths greater than 25 cm, necessary for agricultural water management.