Development and validation of a physics-based root-zone soil moisture retrieval algorithm for multi-frequency dual-polarization (MFDP) signals of opportunity (SoOp) observations

Signals of opportunity reflectometry (SoOp-R) is a promising new approach for sensing subsurface soil moisture profiles through re-utilizing existing satellite transmissions within bands allocated for communication and navigation. Accessibility to broadcast signals at multiple frequencies enable sensing at different depths in the soil. The large range of frequencies used by existing communication satellite signals is not limited to bands allocated or protected for science. SoOp-R, therefore, has a great potential for remote sensing of soil moisture profile moisture and vegetation (above-ground biomass). The Signals of Opportunity P-band Investigation (SNOOPI) cubesat mission will be the first in-space demonstration of SoOp-R at 255 and 370 MHz. Modeling and retrieval algorithm development is necessary to apply this new technology to generate useful RZSM and vegetation products.

In support of this need, a new field experiment is being conducted at an agricultural research site near Purdue University in Indiana. A new receiver has been installed on a tower over a field planted with Soybeans. The receiver uses multiple internal noise sources for radiometric calibration and is installed in a temperature-controlled enclosure for stability. It is capable of collecting polarimetric observations from four (4) signals of opportunity: (1) ORBCOMM (137-138 MHz), a commercial communication satellite constellation in Low Earth orbit (2) Mobile User Objective System (MUOS), a government-operating communication satellite in geostationary orbit (360-380 MHz), Global Navigation Satellite System (GNSS) (1575.42 MHz) and satellite digital audio radio service (SDARS) in geostationary orbit (2.3 GHz).

Several retrieval algorithms are under consideration using a forward model based upon the SCoBi-Veg electromagnetic scattering model. Soil moisture distribution will be first modeled empirically using the principle of maximum entropy (POME). Later, a one-dimensional hydrological model of the sub-surface distribution of water will be studied.

In situ data, including soil moisture from multiple depths, terrain roughness, and vegetation water content from destructive sampling of the soybean crop, will be collected and used to validate the retrievals.