An InSAR-based Soil Moisture Product for Arid Regions

High-resolutions soil moisture products from remote sensing are very valuable but are not free from limitations. Products based on back-scatter change can give conflicting results over dry areas, when the penetration in dry soils becomes significant.[1]

The interferometric phase of Synthetic Aperture Radar acquisitions (in short, the InSAR phase) contains information about the soil moisture variations of the observed target.

This work shows the characteristics of a novel In-SAR-based soil moisture product derived from Sentinel-1 radar observations. The algorithm is based on phase closure inversion, an observable which is immune from atmosphere and deformation contributions to the phase.[2]

The proposed soil moisture product has a resolution of about 200 m and a good coverage in arid and semi-arid regions. It has the potential of filling the gaps of existing high-resolution products based on backscatter change.

An example of the InSAR-based soil moisture product is given in the following figure, which shows a moisture pattern over a rare rain event in the Namibian gravel plain in 2021. Notice the fine structure of channels present in the product. The colorscale units are m³/m³.

 

Validation

The figure below presents a comparison with the ERA5 weather model and a radiometer-based product (C3S passive), which, despite the low resolution, seem to be reliable also over dry areas. The InSAR soil moisture time series corresponds to one year of Sentinel-1 acquisitions over eastern Spain from the ascending orbit direction. The standard deviation of the difference between the InSAR soil moisture and ERA5 surface soil moisture is just under 3% (m³/m³). Notably, the errors are concentrated on a few dates. As expected, products derived from scatterometry (C3S active) are rather unreliable over this site.

Comparisons with weather radar precipitation data validate the high resolution patterns seen in the InSAR product. The match between the two is typically very good.

So far, all the results indicate that the InSAR-based soil moisture can be a reliable product on arid regions and can complement back-scatter change methods in areas with significant penetration. It is expected that InSAR-based soil moisture product will be able to cover larger portions of the land areas with sensors operating at longer wavelengths.

---

References

[1] F. De Zan and G. Gomba, “Vegetation and soil moisture inversion from SAR closure phases: First experiments and results,” Remote Sensing of Environment, vol. 217, pp. 562–572, 2018

[2 ] W. Wagner, R. Lindorfer, T. Melzer, S. Hahn, B. Bauer-Marschallinger, K. Morrison, J.-C. Calvet, S. Hobbs, R. Quast, I. Greimeister-Pfeil, and M. Vreugdenhil, “Widespread occurrence of anomalous C-band backscatter signals in arid environments caused by subsurface scattering,” Remote Sensing of Environment, vol. 276, 2022