
GNSSR Measurement Concept
The Global Navigation Satellite System Reflectometry (GNSSR) aims to retrieve information about the Earth surface by analyzing the signals emitted by GNSS transmitters (such as GPS, GLONASS, GALILEO, future COMPASS...), and captured by an elevated platform after the signal has rebounded off of the Oceans, Land, Lakes, or Ice and Snow. The system has the characteristics of a bistatic radar and scatterometre, at Lband.
The observables of the GNSS and GNSSR in particular, are the waveforms, the crosscorrelation between the signals and replicas (models) of them. Only replicas modelled appropriately (with the GNSS PseudoRandomNoise  PRN, the range and frequency corresponding to the incoming signal) will generate waveforms above the noise. Therefore, tunning the replicas to the right parameters permits to know the range and frequency changes in the range associated to a given satellitereceiver radiolink.
An alternative approach to obtain GNSSR waveforms is by crosscorrelating the reflected signals against the lineofsight (direct radiolink) ones. This technique is the one suggested for the ESA PARIS In Orbit Demonstrator mission (PARISIOD) [MartínNeira, 2011], and it has the advantage of using the entire bandwidth of the transmitted signal, including the encrypted codes. Our experiments have shown that this technique provides altimetric estimates of at least twice precision than the cleanreplica (conventional) approach. [Cardellach et al., 2013].
By comparing the raypath lengths of the reflected and direct radiolinks, the vertical distance between the receiving platform and the surface level can be measured. This is the altimetric application. On the other hand, when the distortion of the reflected signal is analyzed, some geophysical parameters that characterize the reflecting surface can be estimated, such as it roughness and dielectric properties. The roughness essentially acts spreading the signal through the glistening zone, reducing the peak power of the reflected waveform and adding contributions at longer delays. These longer delays are the result of signal raypaths that have been reflected in areas of glistening zone farther away from the specular  the link transmitterspecularreceiver has, by definition, the shortest raypath (see Figure 1).
Furthermore, generating a waveform by crosscorrelating signal and replica during an integration time Ti, filters out any signal arriving with frequencies beyond + 1/Ti Hz from the central one. Because different surface patches induce different Doppler shifts, only certain areas of the glistening zone will contribute to the crosscorrelation. This allows to define Doppler stripes on the surface (areas from which the reflected signal will have the same Doppler shift + 1/Ti). When the reflected signal is correlated along the central frequency solely, it is called Delay Map (DM). Complementary, the correlation can be repeated for different frequencies, mapping different Doppler stripes. The latter measurement is called DelayDopplerMap (DDM).
A review on the large diversity of applications and techniques that potentially can be applied in this GOLDRTR data set are compiled in Cardellach et al., 2011. A summary is presented in Tables 1 to 4. The techniques are identified by a code, used in the paper to assign different potential applications/methods to different campaigns. The table also informs about the type of data required to proceed with each technique: the first part of the datacode might be either C or P, standing for complex or power (raw or integrated data respectively); the second part might be either DM or DDM, for DelayMap or DelayDopplerMaps. Finally, Table 5 lists some of the models involved in GNSS reflectometry.
Figure 1:
GNSSR measurement concept (left): a GNSS satellite transmit Lband signals, that can be received directly by the elevated receiving platform, and also after the signal has rebounded off of the Earth surface. If the surface is smooth, the reflection will be specular (shortest reflected raypath), otherwise, the signal scatters across a wide area called glistening area. (right): if the glistening area is wide enough, their reflections off patches far away from the specular arrive at the receiver with delays with respect to the specular. The surfacepoints corresponding to raypaths of equal delay are called isorange annuli. Similarly, different areas within the glistening will scatter the signal with different Doppler shift than the specular. This defines Doppler stripes. Roughly speaking, a DelayMap maps the power distribution across the range annuli, whereas a DelayDopplerMap also maps the distribution of reflected power across Doppler stripes.
Table 1:
List of the GNSSR altimetric techniques identified in the literature as suitable to be applied in the released data set.
Code: 
Technique: 
Bibliography: 
Data: 
GROUPALTIMETRY 
AG.P 
PeakDelay: altimetric range as peaktopeak delay. In our data set, this is obtained by: NominalDelay (reflected) + WavMaxDelay (reflected)  WavMaxDelay (direct) 
MartinNeira et al.,2001 
PDM 
AG.R 
Retracking: Techniques based on fitting a theoretical model to the data. The bestfit model indicates the delay where the specular point lies 
Lowe et al., 2002,Ruffini et al., 2004 
PDM 
AG.D 
PeakDerivative: The derivative of the waveform has a maximum at the delay corresponding to the specular point. In our data set, this is obtained with NominalDelay (reflected) + WavDerDelay (reflected)  WavMaxDelay (direct) 
Hajj and Zuffada, 2003, Rius et al.,2010 
PDM 
PHASEALTIMETRY 
AP.I 
Interferometricbeats: at low altitude or very grazing observations the delay between the reflected and direct signals is short and their correlation functions overlap, producing interferencebeats. These beats are oscillations of the amplitude and phase of the sum of the two signals, and they occur at the frequency 
Cardellach et al., 2004, Helm et al, 2004 
RHCP (low altitude or elevation) CDM 
AP.5P 
5Parameter DM Fit: A more robust fit using the whole complex RHCP (direct+reflected) waveform, to extract five parameters, among them the altimetric range 
Treuhaft et al., 2001 
RHCP (low altitude or elevation angle) CDM 
AP.SC 
Separate Up/Down Channels: The phase between the direct and reflected links is obtained from separate channels, no need of overlap between direct and reflected waveforms, but need to store direct waveforms 
Fabra et al., 2011 
coherent reflected CDM and direct CDM 
Table 2:
List of the GNSSR Oceanapplications and techniques identified in the literature as suitable to be applied in the released data set.
Code: 
Technique: 
Bibliography: 
Data: 
OCEAN ROUGHNESS 
OR.DM 
DMfit: After renormalizing and realigning the delaywaveform, the best fit againts a theoretical model gives the best estimate for the geophysical and instrumentalcorrection parameters. Depending on the model used for the fit, the geophysical parameters can be 10meter altitude wind speed, or sea surface slopes' variance (mean square slopesMSS). 
Garrison et al., 2002, Cardellach et al., 2003, Komjathy et al., 2004 
PDM 
OR.MDM 
Multiplesatellite DMfit: When the same inversion approach is conducted on several simultaneous satellite reflection observations, the anisotropy (wind direction or directional roughness) can be extracted 
Komjathy et al., 2004 
PDM simultaneous PRNs 
OR.DDM 
DDMfit: The fit is performed on delayDoppler waveforms. In this way, anisotropic information can be obtained from a single satellite observation 
Germain et al., 2004 
PDDM 
OR.TE 
Trailingedge: The fit is performed on the slope of the trailing edge, given in dB 
Garrison et al., 2002 
PDM 
OR.SD 
Scatterometricdelay: For a given geometry, the delay between the range of the specular point and the range of the peak of the reflected delaywaveform MaxWavDelay MaxDerDelay is nearly linear with MSS 
NoguesCorreig et al., 2007, Rius et al., 2010 
PDM 
OR.AV 
DDM Area/Volume: Simulation work indicates that the volume and the area of the delayDoppler maps are related to the changes in the contribution to the brightness temperature of the ocean induced by the roughness 
Marchan et al., 2007 
PDDM 
OR.PDF 
DiscretePDF: When the bistatic radar equation for GNSS signals is reorganized in a series of terms, each depending on the surface's slope, the system is linear with the Probability Density Function (PDF) of the slopes. Discrete values of the PDF(s) are therefore obtained. This retrieval does not require an analitical model for the PDF (no particular statistics assumed). In particular, when the technique is applied on delayDopplermaps, is it possible to obtain the directional roughness, together with other nonGaussian features of the PDF (such as up/downwind separation) 
Cardellach and Rius, 2008 
PDM (isotropic) PDDM (anisotropic and assymetric) 
OR.CT 
Coherencetime: When the specular component of the scattering is significative (very low altitude observations, very slant geometries, or relatively calm waters), the coherencetime of the interferometric complex field depends on the sea state. It is then possible to develope the algorithms to retrieve significan wave height 
Soulat et al., 2004 
Direct and Reflected CDM low altitude and/or calm waters 
OCEAN PERMITTIVITY 
OP.PR 
Polarimetricratio: Ratio between co and crosspolar componentes 
 
Reflected RHCP and LHCP PDM 
OP.POPI 
POPI: difference between the carrierphase of the complex co and crosspolarized components of the reflected field (RHCP and LHCP respectively) 
Cardellach et al., 2006 
Reflected RHCP and LHCP CDM 
Table 3:
List of the GNSSR landapplications and techniques identified in the literature as suitable to be applied in the released data set.
Code: 
Technique: 
Bibliography: 
Data: 
LAND 
L.SMC 
Soilmoisture crosspolar: LHCP SNR is the observable used to extract the surface reflectivity. It can be normalized by the direct power level or even calibrated with observations over smooth water bodies 
Masters et al., 2004, Manandhar et al., 2006, Katzberg et al., 2005, Cardellach et al., 2009 
PDM and direct PDM if calibration wanted 
L.SMP 
Soilmoisture polarimetricratio: Method based on the assumption that the received signal power is proportional to the product of two factors: a polarization sensitive factor dependent on the soil dielectric properties and a polarization insensitive factor that depends on the surface roughness. Therefore, the ratio of the two orthogonal polarizations excludes the roughness term and retains the dielectric effects. The same references note that real data did not support this hypothesis. Some of the assumptions might be too crude, and better modelling is required 
Zavorotny and Voronovich 2000, Zavorotny et al., 2003 
reflected RHCP and LHCP PDM 
L.OI 
Objectidentification: A combination of computing the GNSSR derived total reflectivity together with the carrierphase positioning of both up and downlooking antennas 
LieChung et al., 2009 
RHCP and LHCP CDM 
Table 4:
List of the GNSSR Ice and Snow applications and techniques identified in the literature as suitable to be applied in the released data set.
Code: 
Technique: 
Bibliography: 
Data: 
SEAICE 

I.PP 
Permittivity by peakpower: effective dielectric constant empirically related to the peakpower 
Komjathy et al., 2000, Belmonte 2007 
PDM 
I.VHP 
1styear thickness VHphase: inferred from the phase difference between the vertical and the horizontal polarized components 
Zavorotny and Zuffada, 2002 
reflected RHCP and LHCP CDM 
I.POPI 
Permittivity POPI: from the phase difference between the co and crosspolar circular polarized components 
Cardellach et al., 2006, Cardellach et al., 2009 
reflected RHCP and LHCP CDM 
I.R 
SeaIce roughness: fitting waveform shape 
Belmonte 2007 
reflected PDM 
SNOW 
S.V 
Volumetricscattering: resulting from internal reflections between firn layers (substructure) 
Wiehl et al., 2003 
PDM or PDDM 
Table 5:
Examples of waveform and waveform required models.
GNSS BiStatic Radar Equation 
General DM, DDM 
Zavorotny and Voronovich, 2000 
Convolution forms 
Garrison et al., 2002, Marchan et al., 2007 
Linear in surface slopes PDF 
Cardellach and Rius, 2008 
EM Scattering Models 
KGO 
e.g. Zavorotny and Voronovich, 2000 
KA 
e.g. Ulaby et al., 1990 
SSA 
e.g. Voronovich 1994 
SPM 
e.g. Rice 1951 
Volumetric Scattering 
General 
Ulaby et al., 1990 
Ice/Snow 
Wiehl et al., 2003 
Dielectric Properties at Lband 
General 
Ulaby et al., 1990 
Soil 
Vallllossera et al., 2005 
SeaIce 
AGU M. Series 68, 1992, Winebrenner et al., 1989 
Sea Surface 
Wave spectrum 
Apel 1994, Elfouhaily et al., 1997 
Slopes' distribution 
Gaussian, binormal, GramCharlier (Cox and Munk, 1954) 
