Closeup of the GPS antenna mounting. 

Context in source publication

Context 1

... past. This in- cludes geological estimates of crustal rebound in the Vestfold Hills near Davis (Zwartz et al. , 1997, 1998); age-height re- lationships of marginal moraines elevated above the present ice surface in the Lambert Glacier (Stone et al. , 1998) and global considerations of ice-ocean mass balance (Nakada and Lambeck, 1988; Yokoyama et al. , 2000). However, debate remains about the amount of ice reduction that has occurred since the time of the LGM (c.f. for example, the estimates by Denton and Hughes (1981) and Huybrechts (1990)). Debate also remains about the recent and present mass balance of Antarctica. Some indirect indicators suggest that the reduction in the Antarctic ice sheet continued up to the present (as for the ice sheet model calculations of Huybrechts (1990) or the Late Holocene sea-level inferences of Fleming et al . (1998)) but direct measurements remain few and uncertain. By combining rates of vertical movement derived from GPS observations with geological and glaciological indicators of rebound and ice movements, it becomes possible to separate some of the factors contributing to the sea level signals, isostatic rebound and ice-ocean balance issues. The Lambert Glacier represents one of the largest drainage basins of East Antarctica and is a particularly appropriate study area because substantial fluctuations in glacier volume may have occurred there (e.g. Stone et al . (1998)). Other geological and glaciological studies have been conducted there and, because of the existence of outcropping rock surfaces, a GPS transect can be established well into the ice sheet interior. It has been shown that GPS can be used to detect postglacial rebound (Scherneck et al. , 1998). However, detec- tion of these vertical motions requires long-period observations with high frequency sampling at permanent stations. Tregoning et al. (1999) have shown that remote GPS sites can be operated successfully during the summer seasons using the available solar power. Additional information on similar projects in the Trans-Antarctic Mountains and Marie Byrd Land is available at . gov/antarctica . The major difficulty of conducting GPS observations in remote locations in Antarctica is providing power during winter months when there is insufficient solar power available to run the equipment. We have installed a network of GPS sites in the Lambert Glacier region (Fig. 1) to establish whether postglacial rebound is significant there. A solar-powered site has been operating at Beaver Lake since January 1998 and in January 2000 we will be installing a further two solar-powered sites—one at the coast (Landing Bluff) and one ∼ 450 km inland (Dalton Corner). In addition, we will be installing a hydrogen fuel cell at the Beaver Lake site in order to provide power throughout the winter period when solar power is not available. In this paper, we review the predicted rates of uplift in the region, describe the existing solar-powered systems and the new power management system to be installed in January 2000 and present preliminary results from the data recorded to date. Zwartz et al. (1999) computed predictions of glacio- isostatic uplift rates along a transect across the Lambert Glacier from the coast to the southernmost rock outcrops, about 700 km inland. Such predictions are more strongly dependent on the choice of ice models than on Earth models (see Zwartz et al. (1999) for details of parameters used) as can be seen in Fig. 2 for three different models of the changes in Antarctic ice during the last glacial cycle. The three models — ANT3 (Nakada and Lambeck, 1988), ICE- 3G (Tushingham and Peltier, 1991) and HUY (Huybrechts, 1990) — are based on different assumptions and approaches. The important feature here is that they lead to quite different rebound predictions such that observations of this rebound may help discriminate between models. For example, a relative change in height of + 5 mm/yr between Mawson and Beaver Lake would agree with the HUY model but would not be compatible with the ICE-3G or ANT-3 models. In January 1998, a simple solar-powered GPS system was installed at Beaver Lake as a pilot phase of the main project. The system comprised 4 × 53 W solar panels, two 28 Ah gelcell batteries, a laptop computer, Ashtech Z-XII GPS receiver and a small datalogger for recording system temperatures and voltages (Tregoning et al. , 1999). The system operated from 12 January until 25 March at which time there was insuf fi cient solar power available to maintain operation. A total of 70 days of GPS data were recovered, along with diagnostic data which showed that the temperature maintained inside the insulated suitcase was typically above 30 ◦ C. Bat- tery voltages never fell below 12 V until the system actually failed in late March (Fig. 3). The GPS antenna was mounted on a plate connected to bedrock via a central rod (driven 200 mm into the rock) and three supporting rods drilled into the rock (150 mm deep) (Fig. 4). The pre-amp base of the antenna sits ∼ 50 mm above the ground. While the close proximity of the ground will affect the electrical properties of the antenna, it was thought that this type of mark provided a stable, long-lasting monument that was quick to install and required very few ma- terials to construct. The mark was designed by the Geodesy Section of the Australian Surveying and Land Information Group and took less than 1 hour to install. It was not known whether the original system would sur- vive the Antarctic winter. When revisited in January 1999, the system was found to be intact with no apparent damage to any of the solar panels, frame or GPS antenna. All electronic equipment functioned normally once warmed up; however, the batteries were no longer capable of holding charge or supplying power. We do not have a record of the external air temperature while the equipment was operating; however, the temperature dropped to at least − 40 ◦ C in April 1998 (Fig. 3). Without knowing that the fi rst system had survived, we designed a more rugged system, increasing the number of solar panels from 4 to 6 and increasing the battery capacity to 3 × 73 Ah gellcell batteries (Fig. 5). This system was left operating and will be revisited in January 2000 when the data recorded since 2 February 1999 will be recovered. It is possible that the GPS antenna and radome may become covered with snow. All visits to the Beaver Lake site have been in mid-summer and no one has ever visited the site during or immediately after a blizzard or snowfall to know how much snow accumulates at the site. However, at the time of the visits in the past three years the ground was snow-free even though snow had accumulated in surrounding areas (see Fig. 5) The site is windy and the antenna has been placed to minimize snow accumulation (Tregoning et al. , 1999). In January 2000 we will be installing integrated power systems designed to operate throughout the Antarctic winter. Power will be supplied from a combination of solar and power generated by a hydrogen Proton-Exchange-Membrane (PEM) fuel cell. The fuel cell is a 22 cell PEM stack with no moving parts and generates electricity along with byprod- ucts of water and heat; hence, it is an environmentally clean power supply which will not adversely affect the Antarctic environment. The fuel cell will be running at all times; however, it will only be consuming signi fi cant quantities of hydrogen when a load is connected to it, that is, when it is required to recharge the batteries. We power the 22 W parasitic loads of the fuel cell (fans and air pumps) from the batteries, thereby reducing the amount of hydrogen consumed in the summer months when solar power can be used to maintain the battery charge. We have developed a low-powered controller (PCON, < 0.8 W) which continuously monitors the operation of the system and makes decisions about the power and heating needs of the equipment. If the battery voltages reach criti- cally low levels the PCON will check whether the solar panels are capable of recharging the batteries. If not, it will connect the fuel cell to the batteries in order to maintain the power supply for the GPS receiver. The fuel cell will provide ∼ 3 Amperes to slowly recharge the batteries. At all times, the batteries will be supplying power to the GPS receiver, the PCON itself and the parasitic loads for the fuel cell. There is an absolute minimum voltage level set at which point the PCON will connect the fuel cell, irrespective of the availabil- ity of solar power, in order to prevent a total power failure of the system. Every 20 minutes, the PCON reads and records temperature and voltage sensors and the status of the heaters and batteries. It also communicates with the fuel cell controller to receive diagnostic data on temperatures and voltages within the fuel cell stack. These data are transferred to the computer once per day (see below). In addition to monitoring battery voltages, the PCON measures and records the temperature at three locations within the equipment housing as well as the external air temperature and pressure. If the internal temperature falls below 9 ◦ C, the PCON will begin turning on up to three 14 W heaters to increase the temperature of the box. Also included in the system are a low-powered (2.3 W) PC- 104 card computer with 440 Mb solid-state disk, a Saturn-B satellite phone (capable of transmitting data back to Australia at 9600 bits/second) and an Ashtech Z-XII GPS receiver. The computer (2.3 W) and satellite phone (28 W standby, 110 W transmit) will only be powered up when required while the GPS receiver (12.5 W) will operate continuously. The equipment has been housed in an alumimium-clad wooden box lined with Styrofoam. If the equipment is to operate throughout the year, the thermal design of the insulation must be capable of adapting to outside air temperatures ranging from − 45 ◦ C to − 5 ◦ C and internal equipment power ...