Global Positioning Satellites have been with us for quite as while, and their use in orienteering mapping has long been suggested. The problem has always been the desired accuracy needed for orienteering maps - an error of 5m is about the limit acceptable, with half of that being more desirable. To find out if technology has advanced enough to allow sufficient accuracy for their use in O maps, 13 of us met for a weekend of experimenting at the Lilliput Adventure Centre, Jonathan Swift Park, Co. West Meath.

GPS has its foundations in the US military, being developed to guide missiles accurately to their destination. It consists of nearly 30 satellites orbiting the earth at a height of about 20km. A ground based receiver needs at least 3 (ideally 5 or more)of these to be 'visible' at any one time, each satellite emits a signal, which via triangulation, the ground unit can work out its position and give it in grid co-ordinates. Hand held units are readily available for a reasonable price and are a popular companion for outdoor activities such as hill walking.

The problem for mapping has always been the accuracy, while 10m might not matter to the casual walker, it can make a significant difference to the location of a point feature on an O map. To make matters more difficult the US government have inserted a dither into the equation (which increases the positioning error), to try and prevent accurate civilian use (or rather, to try and prevent it being used against them). This option had eventually been turned off at the start of the decade, but September 11th saw it being reactivated.

To overcome the problem, the shipping industry installed static ground based stations (we were receiving a signal from Holyhead), which record the satellites position, work out the error and transmit this on as a separate signal, allowing ships (and any unit that can pick up their signal), to subtract the error from their co-ordinates, and therefore give an accurate position.

While a 200 Euro GPS will give you your position, it can't record any information, which is necessary for mapping. For the weekend, two 20,000 Euro Lecia surveying units were rented. These could be programmed to include a complete list of orienteering symbols, enabling the operator to select the necessary feature off a menu. With the antenna and receiver in a back-pack, the mapper simply had to go to each feature, select the symbol and push a button, or in the case of a line feature, walk along it, with the receiver updating your position every second.

In the predominately open area of our test survey, the system seemed to work quite well, with the same feature recorded by different people at different times being within an acceptable error. As the satellites are consistently moving, coming in and out of view, recording the same position at different times of the day will give slightly different co-ordinates. However as soon as any tree coverage was encountered, the receiver started to lose both the Holyhead signal and the satellites, and once the figure gets below 4, accuracy can not be guaranteed (X Y co-ordinate can be recorded with as few as 3, but not Z (height)).

Data was downloaded in the evening, converted into a dxf file and imported into Ocad (a programme specifically designed for drawing orienteering maps). Unfortunately in our test, symbols weren't transferred on different layers, meaning that the cartographer had to remember what each individual feature was as it was only represented as a point or a line - a small problem which should be easy enough to get around.

The overall conclusion seemed to be that although it's possible to map via this method, there are still too many limiting factors. The accuracy of the positioning was sufficient, but its price and its lack of ability in covered areas means that the system is not yet suitable for orienteering mapping.

A further experiment carried out, was to hook Ocad up with a hand held GPS unit, although this worked well with point features, Ocad doesn't yet have the capability to be able to track its position while following a linear object. On the plus side this meant that features were being recorded directly into Ocad (so no further cartography is necessary), but it also means carrying a notebook computer around with you, not an ideal situation. With a bit of doctoring to Ocad and a palm top computer, this method seems like the most likely way forward.

GPS should have a future in O mapping, but for the present its use is primarily helpful only in open areas. Although a contour map can be generated from the Z co-ordinates recorded, it is basic and would only be usable in an area with little contour detail, meaning that photogrammetry would still be necessary. It also doesn't help with perhaps the two most important aspects of mapping, the mappers interpretation of the terrain and the manipulation of contours. For the foreseeable future the pencil and compass are likely to remain, but we now know what's possible and what needs sorting out.