Global Positioning System (GPS) data gives us accurate positions for ground features, which are required for compiling our maps. Photogrammetry, (the technique for making maps from aerial photographs), and tying satellite imagery to its true ground position require coordinates and elevation information for features that can be clearly identified on the images or photos.
We use a highly accurate GPS so that error is kept to a minimum from the start of the project. To do this we use high precision geodetic GPS (dual-frequency Trimble 5700 receivers) and a technique called Differential GPS Surveying to measure these ‘ground control points’. This method involves having one GPS receiver on an already accurately known survey point, the ‘base-station’ collecting GPS data simultaneously with other ‘roving’ receivers measuring new, previously unsurveyed points. It requires the base station position to be measured for over an hour but means the observation time at each point collected by the remote receiver needs to be only 15 minutes. The data is recorded on a compact flashcard inside the receiver and it can then be post-processed using Trimble software. Precise positions are calculated for the base stations (using NASA’s JPL AutoGIPSY processing service) and the baseline between the two receivers is used to calculate the position of the unknown point to sub-centimetre level accuracy.
It shows an extract taken from a typical large-scale topographic map compiled using aerial photographs and GPS points. This survey project was conducted to support the compilation of a new 1:25 000 map of Rothera point, Ryder Bay and Wright Peninsula covering the local and technical travel area and the boating area close to the main base. Within the area shown, we set up the GPS base station on a known survey point on Rothera Point, by attaching the Trimble Zephyr Geodetic GPS Antenna to the tripod positioned carefully over the brass marker survey point and setting the receiver to measure the microwave signals from the satellites at a one-second sampling rate. We then carried out a Differential GPS survey using a roving receiver, by climbing up and walking along Reptile Ridge and Stork Ridge and taking measurements from a number of new locations. Each point that was chosen had to be a permanent feature (i.e. not ice), clearly visible from above so that it could be easily identified on remotely sensed images, and contribute to an even network of points across the area to be mapped. Photographs and a detailed station description were taken at each new point to aid in feature identification at the cartography stage.
Later this data was post-processed and used for orientating the aerial photographs in space for producing a large-scale new map of the area with contours and other topographic information.
An additional method of acquiring this information is by recording GPS on the aircraft as we take the aerial photographs of the area. The precise positions of each of the photograph frame centres reduce the number of positions we need to collect on the ground. For example, with frame-centre GPS positions, the minimum number of ground control points for a block adjustment is only 4 well-placed points.
Other recent GPS survey projects have included taking positions in an area at the southern end of the Antarctic Peninsula in order to georeference satellite images in this previously un-surveyed region. This involved setting up a base camp for a few days and being flown on a Twin Otter aircraft close to each nunatak being surveyed, walking across the ice and climbing up to the summit to set up the roving receiver.
If there are areas that are inaccessible by aeroplane or boat from our stations, then sometimes the only option is to reach the points by helicopter. This was the case in 2005 when we carried out a survey on Pourquoi Pas Island and were assisted by the Royal Navy ship, the HMS Endurance, Lynx helicopters. It is certainly an incredible experience to be able to survey areas that have been visited rarely, if ever, before.