Measurement of Trail Races
By Andy Milroy, Ken Young and Ross Zimmerman.
Until recently the difficulty of measuring a trail course to any degree of accuracy has proved a major problem. The advent of Global Positioning System (GPS) has enabled runners in the Tucson Trail Run Series (TTRS) to obtain the distance of specific trail runs, even those which are on very rocky and mountainous trails. This essentially uses a runner instead of a bicycle as the measuring vehicle. A serious runner will tend to run the shortest available route.
TTTRS was started in the late 1970s by Ken Young. Ken laid out the original routes using the series of maps created by the Southern Arizona Hiking Club with the trails marked in red, including distance estimates. Ken used those estimates and hand-calculated climbs from the contour lines on the topographic maps. Most of the distances were obtained by the traditional methods that National Forest and National Park trail builders used for decades. When runners began to use handheld GPS units in the 1990s they could compare their estimates to well-known routes on excellent maps. This provided a “calibration standard” to help evaluate the accuracy of the new technology. In addition to one of the authors (Ross Zimmerman), Wayne Coates has done extensive GPS mapping with trails and Duane Arter used GPS tools to lay out the Old Pueblo 50 mile course. On the web check out , , and , The latter weblink will provide a route to further information for interested readers. Ross would happy to correspond with those interested and to share his knowledge and experience. Contact him at ross.zimmerman – he uses Google’s Gmail
GPS operation depends on the handheld unit triangulating its position relative to the constellation of GPS satellites orbiting the earth. Typically the unit needs to lock onto at least 4 satellites to locate itself with decent accuracy. With previous generations of GPSes, this was often a problem when trees or cliffs blocked direct line-of-sight to the satellites. Since new radio circuits have come out with much better signal lock, GPS technology has become much more effective. Ross Zimmerman, TTRS co-administrator makes most of the maps that are provided to participants. He has used several makes and models of GPSes over the past 10+ years. Currently he uses a Garmin 60CSx, which has the newer radio and has a stubby external antennae. It almost never loses signal. In his study at home it finds 8-10 satellites.
A GPS can determine its position at the user’s request, or it can automatically keep a “track log” wherein the GPS takes a reading at intervals which the user can specify and then keeps recalculating locations, distances, speed, etc. Typically, one uses a GPS in conjunction with mapping software such as the mapping tools from Delorme, Maptech, TopoFusion (All Windows only), National Geographic (Windows and Mac), and MacGPS Pro (Mac only). GPS manufacturers also may have software, like Garmin’s Mapsource. In all cases, one attaches the GPS unit to the USB or serial port of the computer and transfers data into the mapping software.
Until recently the 60CSx was one of Garmin’s top of the line consumer models. . Garmin is the handheld GPS market leader, with a reputation for accuracy and ease of use. Other well-regarded manufacturers include Magellan and Lowrance. (Newer Garmin GPS models with more features came out a few weeks ago, but they appear to have the same radio receiver.) The 60CSx is oriented for the hardcore GPS user; Garmin has other models that are smaller and oriented toward runners (wrist wearable Forerunner series) and cyclists (Edge). The model numbers in those lines ending in x05 have the newer radios. The 60CSx probably still holds signal better. The sport-specific Garmins come with PC software for analysing your GPS data in ways that are useful to your sport.
When Ross undertakes a GPS run or ride, he clears the counters and the track logs on his GPS unit. Typically he keeps his on the counters displays. This tells him things like how far, how fast, how high, how much time, etc. He uses the GPS for road and mountain bike rides as well as runs. This has provided another useful set of comparisons, since his bikes have their own “bike computers” which count wheel rotations. For example, at the Death Ride in the Sierra Nevada of California, the event is claimed to be 129 miles long. Both Ross’ bike computer and GPS calculate the distance at 125 miles.
As mentioned, the accuracy of your GPS unit depends on the number of satellites you have a lock on. 3 is bare minimum for 2D, 4 for 3D. With a good set of satellites, accuracy can be within 10-20 feet of true location. When Ross finishes a run, he turns off the track logging straight away. (He found when he tried to save a track and use a “Saved Track” for his maps that Garmins only save a digest of the full track log.)
He has found the estimates directly from the GPS are can’t always be trusted for reasons he is still trying to work out. The calculation produced by the GPS unit assumes you’re moving on the smooth curve of sphere, not a mountain.
It is likely that in the older units which had more data jitter, with variable readings, Garmin was using a smoothing filter to compensate for aberrant data points. This could affect the readings seriously on switchbacks. Newer, more sophisticated units can now also calculate total feet of climb, however probably with some extra feet due to jitter. It’s very important to have a GPS with a barometric altimeter if you want accurate elevation reading. GPS-only elevation estimates are notoriously inaccurate. Even with the barometric altimeter, there is some scatter to the GPS altitude data, resulting in overestimates of climb and descent. Ross is working with Larry James, the author of MacGPS Pro, on this issue.
After completing a run or ride, Ross notes the GPS tripmeter value, but downloads the GPS track log into the computer. Newer GPSes have USB ports. Maptech Terrain Navigator and MacGPS Pro typically compute slightly different, usually shorter, trip distances than the built-in tripmeter of the GPS. Both are simply adding up all the track point to track point distances.
So in addition to an array of choices for the GPS itself, one has lots of computer mapping software choices. That could be an entire series of articles. One differentiator is the ability to edit your individual data points. If that’s important consider Topofusion. If you’re a Mac user, MacGPS Pro lets you supply your own maps from multiple sources.
One issue that concerns some people is the loss of signal through tree cover.
Both tree canopy and canyon walls can block satellites. When your GPS loses signal, usually the unit will bleep. If it is swiftly removed from its holster and held over head it will often lock back on with a further beep. Switchbacks can compound the issue, especially if they are under trees or between cliffs. In Ross’ experience, this problem is vastly improved with the new model GPSes with the new radio receivers. But if you’d like hear about Dave Patten’s tests with an older model eTrex go to this link—
Until we get to the next generation of GPS satellites and associated handheld receivers, the technology is still some way from the precision of calibrated bike and steel tape measurements.
If the short course prevention factor currently used for road courses were doubled, [from one metre per kilometre of distance of the race course, to two metres per kilometre, over the marathon distance from 42 m to 84 m,] that should be sufficient to insure the course to be at least the advertised distance.
If several runners in a particular race used GPS technology, particularly combined with good computer mapping software that could give a very good indication of the actual distance of the course. Such ongoing re-evaluation would take care of any year-to-year variations in the course. In the discussion list TTR_AZ used for the Tucson Trail Run Series, past run comparisons of GPS readings are a frequent topic.
The GPS gives total distance, splits, times, elevations, total climb and descent, and all sorts of other figures. From this a detailed profile of the course can be produced, and much of the uncertainty of trail running can be resolved. This would greatly aid in determining whether great performances have been set, or if the course was just short