MTB-O
THE MTB-O PAGES SUPPORTED BY WARREN AND TASH KEY FROM THE MELBOURNE BICYCLE CENTRE, CLIFTON HILL
Magic Map-making By Andy Simpson During the middle of 2002 a few Big Feet (from Big Foot Orienteers – NSW) decided that it was about time that the club staged a mountain bike orienteering event. We decided to have a look at the Jenolan State Forest, over the road from the foot orienteering map Billenbenbong. This map had been used for mountain bike orienteering by the club in 1998 when Jock Davis (who didn’t own a mountain bike and ran around the area) set courses that were a little too demanding for most competitors. We rode around the area using the NSW Land Information Centre 1:25000 topographic maps. These seemed to be pretty accurate, although managing three 1:25 000 sheets on a windy day was tricky. The area has a fairly dense network of forestry roads, in various states of maintenance combined with numerous steep-sided valleys to provide plenty of route choice. After lining up a target event (NSW championships 2003) and obtaining permission to use the area from State Forests, we then had to face up to producing a competition map. Our major challenge was to do this with as little field-work as possible, since Map 1: Part of the topographic the area was almost three hours map drive from Sydney and none of us wanted to sacrifice a significant amount of our spare time. We decided that we would try to use data from the NSW Land Information Centre’s Digital Topographic Database to produce a basemap in OCAD. If we were really lucky, having converted the digital data to OCAD symbols we would have a usable map. This type of data is also available from the equivalent authorities in other states and territories. The area that we wanted to map required sections from three different 1:25 000 sheets. For each of those sheets we wanted drainage, contour and transport/cultural data. Data for each of these “themes” is supplied separately for each map. OCAD will happily read and display files in a format called DXF. This is one of the formats available from the mapping agencies. At this point in time, the objects being displayed (contours in the case of map 2) are not mapped to orienteering features, although they do each have an associated type (or “layer”). OCAD can do this transformation from DXF layer to orienteering symbol if you provide it with a key telling Map 2: Contour data imported into it what OCAD feature it should OCAD use for each layer. After importing all of the maps and assigning suitable orienteering symbols to each DXF layer, we had something that looked like an orienteering map, albeit less readable than the map we first started with. Figure 3 shows a sample of the base map after importing all three of the data types (contour, drainage, human) and converting to orienteering symbols. At this stage I converted every forestry track into the symbol for “track (easy riding)”.
Many months had passed by now, and the year of the event was upon us. Jonathan McComb (the planner) and I set off to field check the map, pretty confident that we’d just spend a weekend or maybe two wandering around the area, congratulating each other on what a great job we’d done. Alas it wasn’t to be. We knew that we Map 3: Orienteering symbols would have to classify the track assigned to imported data network according to the IOF MTB-O mapping standards, which would involve walking or driving round about 70km of tracks. We also wanted to check that each track junction was mapped accurately, as this is an important feature for MTB-O competitors. The IOF map symbols changed during the mapping, but a more time-consuming problem was caused by the local trail-bike riders. Every time we were confident that we’d finished a part of the map, we would discover a small trail-bike path, not captured in the base map. Eventually we’d covered over 100 km of track network and spent far too many winter days up on the plateau. The base data from the Land Information Centre was broadly very sound. Most of the forestry tracks were very accurately marked. There were areas where the base data and reality diverged, either because the track network had grown over or just because there was an error in the original data (which also existed on the 1:25 000 sheet). These kind of errors can also be seen in dedicated photogrammetric plots produced for orienteering when there is a thick forest canopy. The additional tracks made by trail bike riders may have shown up if we’d had our own photogrammetry, but they were pretty faint. Producing a usable map from scratch would have been beyond our capability - as volunteers, we simply don’t have that kind of time available. The map we did produce did have some problems, but these related to the efforts and skills of the cartographer rather than to the quality of the base data. By varying some of the colours, future editions of the map should be more Map 4: The map after fieldlegible (even in the middle of a working severe electrical storm). At the NSW Championships the race winner, Anthony Darr, was equipped with a GPS receiver and logger. After the event we were able to plot his course on the map. There were only two points where his GPS track and the track on the map diverged - both of them were on paths added from the ground survey rather than from the base data. I’d recommend anyone producing maps at this level of detail to consider any equivalent data where it is available and affordable. On behalf of the club, we’d like to thank the NSW Land Information Centre for their assistance with this project. Prices for digital data sets vary wildly between states and territories, from a few hundred dollars to thousands of dollars. In theory if the data is held in a single database, it should be possible to provide data for a specific area, rather than on a mapsheet-by-mapsheet basis - this may provide a justification for reducing fees at the higher end of the scale. At the upper end of the charging scale, the costs charged are not competitive with the cost of producing base data from a custom photogrammetric plot. Hopefully as more noncommercial users use these datasets, the higher prices will fall, or they will be placed in the public domain as is the case for the United States. See page 43 for list of contacts.
SEPTEMBER 2004 THE AUSTRALIAN ORIENTEER 45