Garmin Forerunner 205 / Google Earth

Given that the Garmin Forerunner 205 is a GPS reciever it only made sense when I learned you can view your workouts overlaid on spacially referenced maps such as Google Earth. Here's what you need:

1) Garmin GPS unit with data logging capabilities
2) Garmin Training Center software
3) Google Earth Pro
4) ForeConv (freeware)

Once you've completed a workout and downloaded the data into Training Center you need to export your history to a TCX file. Then use ForeConv to convert the file to a KML file. From here you select File>Open in Google Earth Pro. Under Places you should notice a folder called History-Data Garmin. All you need to do now is select the workout you want to view. You'll end up with an image like this one of the Noland Trail in Newport News, VA.


At first glance the accuracy of the data looks quite impressive. But a closer look may discourage you from relying on a GPS receiver all together. I noticed substantial deviations or 20 feet or so from the start. As the run continued the margin of error only increased. By the conclusion of my second mile I was around 50 feet off the path, which can be seen below.


When I crossed the final long bridge a little after 4 1/2 miles I'd estimate the receiver had me 200+ feet off course.


It's tough to say at this point if the Garmin Forerunner's poor data collection is the culprit for the errors or if Google Earth's aerials are true orthophotography.

In one of my earlier posts (February 2008) I stated I had difficulties duplicating distances on clear days during the winter while tree cover was minimal. Based on the randomly weaving data points on straight, open sections of the course it is pretty evident something is up with the Garmin GPS receiver. I still maintain the unit is good for letting you get a decent idea of how far you've run on easy and long run days. The Noland Trail is five miles long and the Garmin Forerunner 205 has always given me a similar mileage total. With a tool like this you can run aimlessly through town, into a park, and down a street you’ve never been on with reasonable accuracy.

Garmin Forerunner 205

For the last month and a half I’ve been using the Garmin Forerunner 205 GPS receiver watch on all of my easy and long runs. At first I borrowed it from a friend for the trip to Europe. Once I got back she stated she wasn’t using it due to her pregnancy and that I could try it out on a few local commonly used run/walk courses. Both of us had been curious to see how the watch faired against a specific course in particular, that being the 5-mile loop of the Noland Trail in Newport News, VA.

The course is a packed dirt and sand trail that winds along Lake Maury. Along the way the trail includes fourteen bridges, scenic lookouts, benches, water fowl, a million squirrels, turtles sunning, and importantly 1/2-mile markers. The trail is heavily wooded and has some relief but nothing I consider challenging. I’m guessing the largest “hill” has an elevation of +25 feet. However a few locations could be considered steep, especially if you do all your running in southeast Virginia.

After two measured runs on the Noland Trail the GPS unit indicated the course to be approximately 0.2 miles shorter than published. If you’re running a 10 min/mile pace this means you’d need to run an extra two minutes on the course to get your distance in. I believe 0.2 miles is a substantial difference if you want to gauge future performances off trail times using the existing markers.

I wondered where the mileage discrepancy occurred so I began to keep a lap count for each 1/2-mile. This is when the watch began to give some oddball answers. Over two consecutive sunny days this winter (no leaves on the trees) I ran 18 miles combined on the trail. By doing so I was able to record at least three pieces of data for each segment of the trail. The results showed 9 of 10 markers to be shorter than 0.5 miles. I averaged each segment length and summed the total distance, which was surprisingly 4.6 miles, not 4.8 as previously measured on continuous runs. The fact that the data points for each segment showed a variation of approximately 0.05 miles was troubling. If you’re running a 10 min/mile pace that’s a difference of 30 seconds per half mile.

The point I’m getting at is that the Garmin Forerunner 205 is not as accurate as most people would like, at least not on the Noland Trail. I am concerned it is difficult to consistently replicate a distance over a known course with this watch. Yet I do believe it is a useful tool for 1/2 marathon and up training.

What the watch is good for is letting you get a decent idea of how far you've run on easy and long run days. It would appear that the discrepancies even themselves out over larger distances. With a tool like this you can run aimlessly through town, into a park, and down a street you’ve never been on.

I think one feature that could be removed from the watch is the pace function. It consistently indicates I'm running 2-3 minutes per mile slower than I end up running for the whole workout. It has never been a reliable source for pace information. You can only deduct that information from the watch by looking at lap times on key intervals and multiplying in your head. But even that's not accurate because the mileage can easily be off 0.1 or 0.2 miles. But it’s something, which in some cases is better than nothing.