The GNSS-Reflectometry concept was conceived in early 90ies [Martín-Neira, 1993] to densify the Earth observations in a low cost effective way. The GNSS-reflectometry works as a bi-static radar: a system in which the transmitter and the receiver are separated by a significant distance, comparable to the expected distance to the target. This definition can be extended to a system in which a single receiver can simultaneously track a diversity of bi-statically scattered signals, from a diversity of different transmitting sources. Then we call it multi-static. The electromagnetic field at the receiver site has contribution from several GNSS sources (transmitting satellites). Different GNSS transmitters can be identified and separated from the rest of transmitters being received simultaneously by the modulation applied to each GNSS. These contributions correspond to signals that have propagated directly from the source to the receiver, crossing the atmosphere; as well as signals that have propagated down to the Earth surface, scattered off its surface, and up to the receiver coordinates. In principle, these two sort of contributions can be separated using two different antennas, one pointing to the transmitters to gather direct rays, and the other to the surface, to collect Earth-surface scattered signals. However, in some applications the geophysical information is extracted from the interference produced by direct and reflected signals. Then, a single antenna pointing towards the horizon, the Earth limb, or at certain slant orientation is used to collect them both. If the receiver is at air-borne or at higher altitudes, the delay and Doppler information can be used to separate both radio-links. Note that other contributions to the receiver electromagnetic field are also possible, such as those coming from atmospheric ducting (atmospheric multipath), or from reflection off other objects surrounding the receiver or along the propagation path. These other contributions are in general source of noise and systematic effects that need to be corrected or mitigated.
Sketch of the GNSS-R concept as a multi-static system of Earth observations. Figure extracted from [Jin et al., 2014] with permission of the author.
The electromagnetic scattering is a complex process involving surface dielectric properties and topographic features as a whole system. The dielectric properties of the surface have direct impact on the reflected power. Two limit-conditions are typically distinguished and contrasted as topographic features: specular or mirror-like reflection vs. diffuse scattering. In most of the cases, the scattering process contains both types of contribution, that is, the scenarios do not present either specular-only or diffuse-only scattering, but both of them in different proportions.
The specular reflection corresponds to scattering processes in which waves from a single direction are reflected into a single reflected direction. On the opposite side, in diffuse scattering the incoming waves are reflected in a broad range of directions. The specular-to-diffuse regime is determined by the roughness structures of the surface topography, rather than its dielectric properties. Scattering with dominant specular component occurs in smooth surfaces, where the surface topography/roughness has not significant features of spatial scales similar to the electromagnetic wavelength.
The diffuse scattering can be approximated by reflections off surface facets. “Facets” are here defined as surface patches of size and curvature of the order of or higher than a few electromagnetic wavelengths. Because of the roughness, different facets are oriented towards different directions. Incident rays reflect off the facets, each facet producing a mirror like reflection, which forwards the reflected rays towards a direction determined by the facet’s normal vector and the incident ray direction. In bi-static geometric conditions, the receiver only collects those rays reflected off facets with the appropriate tilt. The glistening zone is then defined as the area from where well-oriented facets might exist above a probability threshold. The glistening zone corresponds to the deterioration of the specular image. Note that surface coordinates away from the nominal specular point require higher slopes of the facet to forward the signal towards the receiver. Note also that the rougher the surface the higher the probability of largely tilted facets, meaning higher probability of well-oriented facets at coordinates far away from the nominal specular point. Therefore, the rougher the surface the largest the resulting glistening zone.
Generation of a couple of frequency-slices of a DDM. Contribution to each frequency-slice comes from a Doppler belt (red lines). Contributions to the delay-map (from white iso-delay zones) are indicated by the black arrows. Figure from [Cardellach et al., 2011].
The total optical path traveled by the signals reflected at surface points away from the specular point are longer than the path traveled by the specular one, the farther away from the specular reflection the longer the reflected optical path. Therefore, large glistening zones (rough conditions) result in longer tails in the reflected echo. Similarly, the Doppler effects differ across the reflecting surface, resulting in spread frequency responses as the reflection occurs over large glistening zones (rough surface conditions). The shape and power distribution of the echo along the delay-Doppler domain (called waveform and delay-Doppler map) is thus representative of the reflecting surface conditions: its dielectric properties and roughness state. These are the primary GNSS-R observable.
Further detailed can be found in the State of Art Description document, which includes this introduction and its references.