Soil moisture as signal and noise in SAR interferometry
Abstract
The possibility of measuring changes in soil moisture with differential SAR interferometry (DInSAR) was conclusively demonstrated in the first seminal paper on the repeat-pass technique using SeaSat's L-band sensor (Gabriel et al, 1989). There they confirmed that apparent increases and decreases of the soil surfaces of farm fields corresponded exactly with irrigation patterns. Since that time, the DInSAR technique has become a huge success but its application to soil moisture has been largely overlooked. Here we present theory, laboratory work, and satellite measurements that suggest that microwave penetration depth varies smoothly with soil moisture in nearly any soil and can therefore be used as a proxy for it, unlike clay swelling which occurs in a small percentage of soils and is too complex to model adequately. We find from the literature that the relationship between soil moisture and penetration depth is non-linear and indicates that a change of 5% volumetric water content (VWC) should cause between 1 and 50 mm of change in C-band penetration depth depending on initial VWC. Recent experimental measurements of this relationship within an anechoic chamber confirm this non-linear relationship and confirm that the expected magnitudes are sufficient for DInSAR measurement. We also present ERS-2 results that show intriguing spatial variations in millimeter-scale path-length change that correlate with hydrological features such as stream channels and watershed boundaries. Detection of these subtle signals was facilitated using a digital elevation model with high vertical accuracy, which we use to demonstrate that a real, ground-based signal can be found at the 0.5 mm level. Our latest research is focused on using various point scatterer techniques to construct long time-series of soil moisture maps using both ERS and Radarsat data. We conclude that penetration depth may be a viable proxy for measuring soil moisture, and that one reason the signal is often overlooked is that it is usually overwhelmed by errors caused by topographic uncertainties. Soil moisture may also be a significant, and previously unrecognized, source of noise in the measurement of subtle deformation signals or the creation of DEMs using repeat-pass interferometry.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2005
- Bibcode:
- 2005AGUFM.G41D..08N
- Keywords:
-
- 0699 General or miscellaneous;
- 0933 Remote sensing;
- 1240 Satellite geodesy: results (6929;
- 7215;
- 7230;
- 7240);
- 1640 Remote sensing (1855);
- 1866 Soil moisture