A dual-polarimetric SAR processing chain for soil moisture retrieval
According to sustainable agriculture best practices, efficient use of scarce water resources is mandatory for both a marketing objective and an environmental obligation. This implies that in the agricultural production, which is intensive and should at the same time be environmentally friendly, soil moisture is a key parameter to be constantly monitored. In addition, soil moisture plays a crucial role in plant development, human development as well as global cycles of various substances. It serves as an essential input variable for various scientific analyses ranging from hydrological modeling, forecasting of floods and groundwater movement to the modeling of global water fluxes.
Information about soil moisture can be obtained from in field measurements taken, for instance, using point sensors [1] that provide detailed point-like information. An alternative approach to field measurements is to use measurements remotely sensed from satellite-borne instruments. Both optical and microwave radiation exhibit sensitivity to soil moisture, with the optical remote sensing being limited to clear sky conditions and affected by solar illumination [2]. Microwave radiation, on the other side, is largely unaffected by weather conditions and guarantees all-day observations. Among the microwave remote sensing instruments, the Synthetic Aperture Radar (SAR), i.e., a microwave imaging radar, is very promising to soil moisture retrieval on a spatial scale fine enough to be used for sustainable agriculture purposes.
To retrieve soil moisture from microwave remotely sensed data, the key issue is de-coupling surface roughness from dielectric constant. Within this context, two different approaches are widely used: a) physical modelling and b) empirical methods. A promising approach which is both physically sound and computer-time effective was proposed in [3] which consists of using dense time-series of SAR measurements to decouple surface geometric effects (plants growth stage, etc.) from dielectric properties. The underpinning idea is that plant appearance will not change drastically from one image to another if the time series is dense enough, hence variation in the dielectric properties are sorted out. Once the permittivity is estimate, the soil moisture can be retrieved using an empirical approach, e.g., [4].
A mandatory step to design an operational processing chain to retrieve soil moisture using [3] is sorting out built-up areas, vegetation, high-slope terrains, etc. In this study, a polarimetric processing chain is proposed that, starting from dual—polarized SAR measurements, is able:
1. To sort out built-up areas using reflection symmetry, i.e., a property that is satisfied by natural scenes but is not present in man-made targets. This property manifest itself in the inter-channel correlation, i.e., the correlation between co- and cross-polarized channels that is low in case of natural targets and large over built-up areas [5].
2. To sort out vegetated areas using eigenvalue decomposition parameters, i.e., the polarimetric entropy and the mean alpha angle, to partition the polarimetric space to identify vegetated regions according to their peculiar polarimetric response.
3. The digital elevation model (DEM) to identify area calling for steep slopes.
The proposed processing chain will be showcased on actual SAR measurements acquired by Sentinel-1 over two areas of interest, namely the Campania and the Sardinia regions. In the Campania region, the test case includes ground information about soil moisture collected by a ground station provided by Netcom Group S.p.A. First experimental results show the soundness of the proposed processing chain that results in accurate enough estimations with a remarkable computer-time effectiveness.
References
[1] Lekshmi SU S, Singh DN & Shojaei Baghini M 2014. A critical review of soil moisture measurement. Measurement 54, 92-105. doi:10.1016/j.measurement.2014.04.007
[2] Gao, B.-C. 1996. NDWI - A normalized difference water index for remote sensing of vegetation liquid
water from space. Remote Sensing of Environment 58: 257-266.
[3] Balenzano, A., Mattia, F., Satalino, G., Davidson, M.W.J., 2011. Dense temporal series of C- and L-band SAR data for soil moisture retrieval over agricultural crops. IEEE J.Sel. Top. Appl. Earth Obs. Remote Sens. 439–450
[4] Hallikainen, M.T., Ulaby, F.T., Dobson, M.C., El-rayes, M.A., Wu, L., 1985. Microwave dielectric behavior of wet soil-Part II: Dielectric Mixing Models. IEEE Trans. Geosci. Remote Sensing GE-23 ge-23, 35–45
[5] F. Nunziata, M. Migliaccio and C.E. Brown, “Reflection symmetry for polarimetric observation of man-made metallic targets at sea,” IEEE Journal of Oceanic Engineering, vol.37, no.3, pp.384-394, 2012.
