Work

Validation of the NISAR Multi-Scale Soil Moisture Retrieval Algorithm across Various Spatial Resolutions and Landcovers Using the ALOS-2 SAR Data

Year: 2025

Type: article

Abstract: High-resolution soil moisture data are essential for numerous geophysical applications, enabling improved decision-making in environmental and resource management. However, current satellite-derived global soil moisture products suffer from coarse spatial resolution, limiting their utility. The upcoming NASA-ISRO SAR (NISAR) mission, set for launch in July 2025, aims to overcome this limitation by providing high-resolution soil moisture data at 200 [m]. One of the mission’s key approaches is the multi-scale algorithm, which enhances coarse-resolution data by incorporating fine-scale synthetic aperture radar (SAR) observations. While initial validation of this algorithm has been conducted over cropland, a broader evaluation is needed across various land covers and climates to ensure its robustness. This study investigates the performance of soil moisture retrieval across 5 diverse test sites, covering forest, shrubland, cropland, and grassland environments, as well as hydrometeorological conditions ranging from arid to polar. The algorithm was assessed at 100 [m] and 200 [m] resolutions, revealing consistent moisture patterns, with the finer resolution offering greater detail. Validation using in situ measurements showed that the unbiased root mean square error was less than 0.06 [m 3 /m 3 ] for most sites, matching NISAR’s accuracy requirements. A wet bias was observed, and challenges emerged at a polar site due to organic soil. A minimum performance test was conducted to evaluate the impact of SAR backscatter measurements. The results demonstrate that these measurements contribute to improving the accuracy of high-resolution soil moisture retrieval using a multi-scale algorithm. Overall, the study highlights the algorithm’s capability to retrieve soil moisture at high resolution, reinforcing its suitability for the NISAR mission.

Source: Journal of Remote Sensing

DOI: 10.34133/remotesensing.0729

Authors: Preet Lal, Gurjeet Singh, Narendra N. Das, Rowena B. Lohman

Institutions: Michigan State University, Cornell University

Language: English

Topic: Soil Moisture and Remote Sensing

Subfield: Environmental Engineering

Field: Environmental Science

Domain: Physical Sciences

SDG: Life on land

Open Access status: diamond

Funder: National Aeronautics and Space Administration