Following the initial research, Born et al. (1994) present the first T/P calibration results using a GPS buoy at the Harvest platform. The CCAR spar buoy used in Kelecy et al. (1994) was modified with the addition of a Magnerule™ sensor to provide an additional independent measurement of the distance between the waterline and the GPS antenna. The buoy was deployed for a 1 hr period (centred on the T/P overflight time) acquiring data at 1 Hz. Bias estimates for T/P determined from the GPS buoy were comparable to those computed from the in situ instrumentation at Harvest platform (Christensen et al., 1994). The study represented a successful conclusion to the development of the CCAR spar buoy. The second dedicated calibration site for the T/P mission was located at Lampedusa in the Mediterranean (Menard et al., 1994), and operated by the Centre National d’Etudes Spatiales (CNES). The CNES calibration methodology utilised the large ruggedised buoys developed by Hein et al. (1990). The Lampedusa experiments were compromised due to damage of in situ equipment in rough weather, hence the CNES calibration site was moved to the island of Corsica.
Further research conducted by the CCAR group focused on the development of GPS processing algorithms for medium accuracy (~1 m RMS) positioning of ocean buoys as part of the Fast Pegasus experiments (Key et al., 1996 and Key et al., 1999). In comparison with the centimetre level differential kinematic processing of previous experiments, Key et al. (1999) presented a precise point positioning (PPP) algorithm for single receivers. With a reduced accuracy in comparison to short to medium baseline differential processing, the PPP technique eliminated the need for land based reference receivers allowing significantly greater range for the Fast Pegasus application.
By the mid 1990s many groups were investigating GPS buoys for a range of applications. Schutz et al. (1995) developed a wave rider buoy similar to Kelecy et al. (1994) as part of a T/P calibration study in the Galveston Bay region of the Gulf of Mexico. Key et al. (1998) also revisited the wave rider buoy at Harvest platform, with a series of deployments undertaken in 1995. Sixteen deployment locations were chosen with the primary aim of mapping the sea surface over a 10 km diameter circular area surrounding the Harvest platform. The new wave rider buoy was again based on a life preserver as the floatation platform (Figure 2-1), with the receiver operated from a boat within ~6 m (20 ft). The choke ring
Antenna Reference Point (ARP) was positioned approximately 87 mm above the mean water level. The project demonstrated the ability to map the sea surface for regional oceanographic experiments. The Key et al. (1998) experiment also suggested a preference for the simple wave rider style in comparison to more complex designs. Turbo Rogue Choke Ring Antenna Seam Approx. Water Level Antenna Dome Life Preserver Antenna Cable
Figure 2-1 Schematic view of the CCAR wave rider GPS buoy (adapted from Key et
al., 1998).
Parke et al. (1997) provided a review of GPS based water level measurement, with an emphasis on existing and potential applications. In particular, the authors drew attention to the potential value of GPS buoys for regional oceanographic experiments, including investigations into tides, currents, jets, fronts and eddies. Parke et al. (1997) stress that for routine applications, the GPS buoy technique requires demonstration of high accuracy at significant distances from shore. This is a reoccurring statement throughout the relevant literature.
The first in situ measurement campaign at the relocated CNES calibration site in the Corsica area was undertaken in 1996/97. Large geoid gradients in the area lead to a series of experiments using the CCAR wave rider buoy design to map the geoid slope using similar methodology to the Key et al. (1998) study. The first experiment was carried out in May 1998 using two buoys (Figure 2-2a), with different antenna and receiver combinations (Exertier et al., 1998). Continuous deployment and retrieval from a small boat at different locations made the wave rider technique time consuming and logistically difficult. Subsequent investigations included the development of a catamaran style floating platform (Figure 2-2b), equipped with two GPS antennas. The design was first deployed in 1999 (Bonnefond et al., 2003b), towed at a constant speed covering an area approximately 20 km x 5.4 km centred on the ground track of the altimeter. The catamaran represented an innovative approach to mapping the sea surface, using two receivers on the catamaran to provide a redundant check. The differences in the height component between the two GPS antennae throughout the experiment had a standard deviation of 12 mm. Comparisons with a nearby tide gauge showed
a mean bias of 19 mm which (partly) reflects uncertainty in the floatation position of the catamaran when under tow. This bias was not considered to affect the geoid slope estimation due to the assumption of a constant towing speed through the water. This point does however underscore the importance of understanding the dynamics of the floating platform for absolute applications. The geoid gradients derived from the GPS catamaran have been integrated into the determination of the CNES and Observatoire de la Côte d’Azur (CERGA) altimeter calibration values, reducing the variability of T/P bias estimates from 49 mm to 33 mm (Bonnefond et al., 2003b). The episodic deployment (dependent on sea state conditions) of individual life preserver style GPS buoys (Figure 2-2c), remains central to the CNES/CERGA calibration methodology (Bonnefond et al., 2003a). The CNES/CERGA calibration strategy is discussed in further detail in Chapter 4.
(c)
(a) (b)
Figure 2-2 CNES/CERGA buoy designs. (a) CCAR buoy design used in the Exertier
et al. (1998) study. (b) GPS catamaran used for the Bonnefond et al. (2003b) sea surface mapping experiment. (c) Individual buoy used at the CNES/CERGA calibration site
Bonnefond et al. (2003a). Images courtesy P. Bonnefond.
With increased significance placed on in situ verification of satellite altimeters in the lead up to the launch of the Jason-1 mission (December 2001), various groups developed GPS buoy based calibration experiments. Following an evaluation of the performance of T/P over the Great Lakes of the USA (Morris and Gill, 1994), researchers at Ohio State University (OSU) developed a series of calibration experiments undertaken in 1999-2001 utilising CCAR life preserver style buoys in Lake Michigan and Lake Erie (see Shum et al., 2003 and summaries in Cheng, 2004a and Cheng, 2004b).
Cardellach et al. (1999) and Cardellach et al. (2000) from the Institut d’Estudis Espacials de Catalunya (IEEC, Barcelona, Spain) presented results from a range of experiments for the calibration of ERS-1, ERS-2 and T/P off the Catalonian Coast and around the Balearic Islands. The IEEC work differs to previous studies due to
the extended baseline length (~80 km) and the development of a tri-buoy system (Figure 2-3a). The design was based on the connection of three life preserver style GPS buoys, each separated by a flexible ~2.8 m connection. Two of the buoys housed a choke ring antenna, connected to a separate receiver operated in a boat approximately 50 m away. The third buoy was simply a dummy of the same design and weight, included to ensure the entire system was symmetrical about the central axis. This innovative set-up was designed to enable a check between two solutions and provide a redundant data set to fill possible gaps. The second phase of the research was directed towards absolute calibration of the EnviSat altimeter (Torrobella and the GRAC-II team, 2003). The research refined the design, developing the ‘bi-buoy’ system (Figure 2-3b). The bi-buoy incorporates two wave rider buoys which are fixed 1.6 m apart using a rigid connection. Despite altered dynamics, the bi-buoy system allows additional quality control through comparison of observed versus known separation distance between the two buoys. Constraining this known distance in a combined multi-rover, single reference station GPS solution also has the potential to assist long baseline kinematic solutions.
(a) (b)
Figure 2-3 Multiple wave rider designs from the Institut d'Estudis Espacials de Catalunya (IEEC), CSIC Research Unit, Barcelona, Spain. (a) Tri-buoy design (Cardellach et al., 2000), Image courtesy J. Font. (b) ‘Bi-buoy’ design (Torrobella and the
GRAC-II team, 2003), Image courtesy J. Torobella.
A second Spanish team (Universitat Politecnica de Catalunya (UPC) in Barcelona, Spain) have undertaken a range of GPS buoy experiments, primarily for Jason-1 calibration in the area surrounding Ibiza Island in the NW Mediterranean Sea. The buoy design is once again based on the CCAR life preserver, however the antenna is situated within the life preserver ring, reducing the antenna height with respect to the mean water level by ~100 mm (Figure 2-4a). The lower position of the antenna was found to be an improvement to the initial design (Figure 2-1),
lowering the centre of mass and hence stabilising the buoy and reducing buoy tilt (Martinez-Garcia et al., 2004).
Calibration activities in the eastern Mediterranean Sea are undertaken in the area surrounding the Greek island of Gavdos (Mertikas and the Gavdos team, 2002). The Gavdos collaborative project has incorporated GPS buoys developed at the Geodesy and Geodynamics Laboratory (GGL), at ETH Zürich. The buoys have aided in the calibration of airborne altimetry in the region of the calibration site (Geiger et al., 2003). The ETH design has moved away from the life preserver style, adopting a 0.4 m diameter spherical design (Figure 2-4b). The buoy is capable of autonomous operation as it houses the antenna, receiver and battery. Using a Novatel antenna and receiver the buoy is able to operate for approximately 20 hours.
(a) (b)
Figure 2-4 Wave rider ‘life preserver’ buoy designs. (a) Universitat Politecnica de Catalunya (UPC) design with choke ring antenna. Image courtesy M. Martinez-Garcia. (b)
Geodesy and Geodynamics Laboratory (GGL), ETH Zürich design with a Novatel antenna. Image courtesy H-G. Kahle.
The move towards light weight, wave rider style buoys capable of autonomous operation over short periods, was also adopted for the UTAS Mk II buoy presented in §2.3 of this Chapter.