Temporal and Spatial Associations of L-Band Fully Polarimetric SAR Metrics with Soil Moisture over a Wet Volcanic Ash Soil Field
L-band synthetic aperture radar (SAR) offers repeated observations of soil moisture, but its ability to track field-averaged changes and within-field differences under wet conditions requires separate assessment. We compared 21 fully polarimetric ALOS-2/PALSAR-2 acquisitions with time-domain reflectometry (TDR) observations at 12 locations in a short-grass volcanic ash soil field. Field-averaged volumetric water content (VWC) ranged from 0.23 to 0.54 cm3 cm−3 and exceeded 0.30 cm3 cm−3 on 17 acquisition dates. On the 19 dates common to all metrics, VV backscatter and the Freeman–Durden surface component showed comparable temporal associations with VWC (Pearson r = 0.72 and 0.71; Spearman ρ = 0.71 and 0.69, respectively). Their dependent correlations did not differ detectably (p = 0.923). In contrast, within-date spatial correlations varied substantially across dates, and reducing or omitting the backscatter and Pauli filtering windows did not produce consistent correspondence with the observed within-field VWC differences. The Improved Integral Equation Model (IIEM), driven by TDR-derived permittivity, reproduced VV more closely than HH and HV: baseline model–observation RMSEs were 1.41, 4.40, and 25.5 dB, respectively. This ordering persisted across the evaluated roughness range, while varying the assumed loss tangent had negligible influence. These results support VV as a comparatively simple indicator of field-averaged temporal soil moisture variation under predominantly wet conditions, with Freeman-S providing complementary component-level information. The IIEM comparison supports a dielectric surface-scattering contribution to VV, whereas the weak VWC association and large model discrepancy of HV indicate limited usefulness for temporal moisture monitoring at this site. Field-averaged temporal sensitivity should be distinguished from the ability to represent fine-scale spatial moisture patterns.
Authors
- Kosuke Noborio (ORCID: https://orcid.org/0000-0002-8749-0318)
- Shinsuke Aoki (ORCID: https://orcid.org/0000-0003-0265-3470)
- Daiki Kobayashi (ORCID: https://orcid.org/0009-0003-0782-6315)
- Naoto Sato
- Maiko Kawaguchi (ORCID: https://orcid.org/0009-0006-3033-7562)
Institutions
- Meiji University (JP)
- Kagawa University (JP)
- NTT (Japan) (JP)
- Hokkai Gakuen University (JP)
Publication Details
- Journal
- Remote Sensing
- Published
- 2026-09-25
- DOI
- https://doi.org/10.3390/rs18193316
- Primary Topic
- Soil Moisture and Remote Sensing
- Type
- article
- Field-Weighted Citation Impact
- 0.00