The GPS Tide Gauge Problem Revisited

Conference poster, 2011

It is well-known that GPS instruments can be used to measure local sea level. In most experiments, two antennas are deployed at a coastal site. A geodetic antenna - optimized for RHCP signals - is used in the traditional orientation and tracks the direct signal. The second antenna is optimized for reflected signals - which are primarily LHCP - and is pointed towards the ocean. The sea surface can then be estimated by analyzing the carrier phase data. While the data from the "up" antenna are dominated by the direct signal, the effects of signals reflected from the ocean are also present in its data. Thus in principle, one might be able to estimate sea level using only data from the "up" antenna. This is similar in concept to recent multipath studies where geodetic GPS installations are being used to measure soil moisture variations and snow depth. We have analyzed GPS data for a three-month period from a GPS tide gauge installation at the Onsala Space Observatory. It is located on the western coast of Sweden. We used the SNR data from the "up" antenna only. The data were windowed by azimuth for ocean-reflections and elevation angles from 18-40 degrees. This provides hourly sea level measurements. Comparisons were made to an average for tide gauge records 18 km south and 33 km north of Onsala. The standard deviation of the residual between our solutions and the tide gauges is 4.9 cm. This is less precise than the combined up-down antenna system of 2.6 cm. These precision values include errors associated with real tidal motion at the GPS site. While the "down" antenna performs poorly in high-wind conditions (> 8 m/s), we found that the "up" antenna performs significantly better at these times.