Signal Power Analysis of GNSS Reflections Recorded at Kongsfjorden, Spitsbergen

Luis Peraza (Sucessful finalization May 2016)

The adapted GNSS receiver, GORS, set up near the village of Ny-Ålesund in the island of Spitsbergen, Svalbard, was used to monitor GNSS reflections on the adjacent Kongsfjorden. Two antennas, one up-looking Right-Hand Circular Polarized (RHCP) and one horizon-tilted Left-Hand Circular Polarized (LHCP), installed at the Zeppelin mountain, have been recording almost continuously since 2013. The field of view from this location allows to capture reflections as far as 14 km from the receiver's position, over the complete fjord and nearby surrounding land areas, using an elevation view mask from 2° to 10° above the horizon. Two reflection events were captured daily for each PRN, a rising and a setting event, which were analyzed throughout a complete year. This data was used to observe the behavior of the Signal-to-Noise Ratio (SNR) of the reflected signal due to seasonal changes on the fjord's water surface considering the presence of sea-ice. A new method was tested to retrieve the signal power for the observations from the recorded signal amplitude. The in-phase (I) component of the signal was used to derive direct and reflected power, and the quadrature-phase (Q) provided the background noise. For this study, only samples recorded by the up-looking antenna (Master samples) were used. Generally, very strong SNR was observed for reflections occurring over areas of the fjord that remained ice-free during the sea-ice season, with values above 40 dB for most observations. Due to the significantly low sea-ice extent during the observation year, compared to previous years, only a few of the captured reflections took place on sea-ice. The SNR values retrieved for those observations over sea-ice covered areas were relatively stronger than for those over water surfaces, which corresponds to the Fresnel reflection coefficients for RHCP L-band signals over such media. However, and according to the same coefficients, stronger signal reflections were expected for the observations over sea-ice since they occurred at lower elevations than those off water. As a result, a definite conclusion about SNR changes due to sea-ice reflection could not be arrived at. Nevertheless, this investigation shows the usability of the GNSS-R technique to monitor Arctic coastal regions or other environments that are, otherwise, very difficult to access and/or to be modelled accurately with established remote sensing techniques.