In 1998 a probatory experiment had been conducted at Corsica absolute calibration site (Senetosa) in order to determine the local marine geoid slope using GPS buoys. The very good results obtained during this campaign have encouraged us to extend the covered area (May 1999). This was necessary insure the calibration process by using more 10 Hz altimeter data and above all data far from the shore which are of better quality. However, the GPS buoys technique is too long to cover a large area and we have then decided to build a Catamaran with two GPS antennas (Turbo-rogue/Sercel) on board. Tracked with a boat at a constant speed (between 3 to 3.7 m/s) this have permitted to cover an area of about 20 km long and 7 km wide centered on the TOPEX/Poseidon ground track N° 85.

Results shows a very good consistency between GPS receivers: filtered sea height differences give a mean of -0.2 cm with 1.2 cm standard deviation. Moreover, crossover differences between GPS filtered sea heights have been performed in order to control the quality of the process. No systematism or distortion have been evidenced and statistic results of crossover differences give a mean of 0.2 cm with standard deviation of 2.7 cm. However, comparisons with tide gauges data show a bias of 1.9 cm with a standard deviation of less than 0.5 cm. Even if this bias does not affect the geoid slope determination which is used in the altimeter calibration process, it needs further investigation in the framework of vertical reference studies. Results in term of altimeter calibration show an improvement of more than 1.5 cm rms by using the new geoid slopes.