GNSS Reflectometry to Complement SMOS-like Measurements
Cardellach, E.¹; Rius, A.¹; Reul, N.^2
1ICE (IEEC-CSIC); ²IFREMER

The GNSS signals reflected on the Ocean surface and captured by a receiver over-flying the area are sensitive to the roughness state of the sea surface, acting similarly to a bi-static radar or scatterometer. Unlike wind scatterometers in Ku and Ka bands, mostly sensible to small ripples at the ocean surface with wavelengths on the order of the cm, which are generated by the wind on top of the larger scale waves, the GNSS reflected signals are sensitive to slightly larger scales of sea roughness, around 20 cm, what we may define as the "L-band roughness". Understanding this effective roughness, and the oceanographic and air-sea interaction processes to which is relevant might boost new applications for the GNSS-Reflectometry (GNSS-R). The development of this understanding runs in parallel to the development of the L-band radiometric Space missions designed to retrieve soil moisture and Ocean salinity (ESA's SMOS and NASA's Aquarius), since the radiometric observations over the Oceans are also affected by the surface (L-band) roughness. This is an important issue in the development of the algorithms to extract Ocean salinity from SMOS-like data.

Two sets of ESA air-borne campaigns have been conducted to address open issues of SMOS Ocean Salinity retrievals: CoSMOS-OS 2006 (12 flights in Norway) and CoSMOS-OS 2007 (3 flights planned for August, in Finland). Two L-band radiometers: EMIRAD (Danish National Space Center, DTU) 2006, and HUT-2D (TKK Helsinki University of Technology) and a GNSS-R receiver: GOLD-RTR (IEEC/CSIC) have been the payload of the flights, together with inertial navigation systems, navigation GPS, and an Infrared Profiler. The GOLD-RTR collected more than 32 hours of data in CoSMOS-OS 2006, in the form of 10 delay or delay-Doppler waveforms, sampled at 64 lags with 15 meter inter-lag space, and using both Left-Hand Circular and Right-Hand Circular polarizations (LHCP/RHCP).

The GNSS-R data has been analyzed following different algorithms, each of them identifying the geophysical information in different parts of the waveform. For instance, one of the algorithms relates the time-delay of the peak power with the roughness state (approach to provide roughness information in near real time). Another algorithm inverts the delay-Doppler maps into a discrete version of sea surface slopes' Probability Density Function (PDF), with no a-priori assumption on the shape of the PDF. The latter approach allows for detection of non-Gaussian features of the sea surface, including the "sense" of the force inducing the surface asymmetry (e.g. up/down-wind sensitivity). The comparisons of the GNSS-R measurements with the L-band radiometric observations is a work under progress to determine the synergies between them and to help understanding both reflectometry and emissivity at L-band. Preliminary results showed how the GNSS-R measurements explained features in the radiometric data that the operational meteorological numerical models could not detect.

The role of the GNSS-R in the ESA CoSMOS campaigns will be presented, including the GNSS-R analysis for retrievals of the sea surface roughness, and the comparison with the L-band radiometric observations.