GPS in the woods ...

On the road in open terrain (not around high rises), GPS will be pretty accurate since the error has no bias. You will be able to plot your course on a map (using a tool) and it will be pretty darn close.

Out in the woods, it's a different ball game. The foliage will obscure the satellites and you'll get TERRIBLE accuracy. GPS can measure accuracy, but only to the same precision as it can measure latitude and longitude. So if you go up and down a 10 foot hill, It may not be measured at all.

Or you need to get a new GPS. Mine can even get a 2-3D fix in my upstairs bedroom. Readings tend to flucuate a bit but I do get regular fixes. I would consider the fixes I get from my bedroom TERRIBLE but then again I know where I am. I have never had problems in the woods or under a sailboat bimini.

Man with watch knows what time it is.

Man with two watches is never sure.

LOL! Reminds me of a story my father told me. When he was a young man (he's almost 90 now), there was a siren that sounded in town at noon each day. There was also a military academy which fired off a canon at a certain time each day.

He asked the people city officials how they knew when to sound the siren and they said they set their clocks by when the academy fired off their canon. When he asked the people at the academy how they knew the exact time to fire the canon, they said they set their clocks by the siren!

Actually, I think that is what often happens with these measurements for races in a sense. Someone measures it somehow - no one probably even knows how they did it - but from then on that route is consider to be "correct". And if they alter the course slightly - maybe start it a mile earlier, but want to go the same distance, they simply measure a mile from the old ending point. So any errors are still there even with a changed course.

Given that your most precise source for distance measurement is a tape measure and that you are measuring, at best, a few revolutions of your wheel, the error associated with this measurement is probably going to be, at worst, 5mm for a single revolution. 5mm error for a 2100mm circumfrence wheel comes to 0.23% error. On a ride of 100 miles, this equates to an error of +-0.23 miles. So, if you had this uncertainty in your setting, you can simply ride an extra 1/4 mile and you will know that you have ridden at least a full 100 mile century.

However, this does not tell the whole story. There are other error sources that have not been addressed. First off, when you do your rollout test, unless you have a helper push the bike while you are on it, chances are you will not be in your exact riding position. This will result in a different weight above your wheel causing some amount of additional uncertainty in the measurement. Second, even if you are able to get into position, you are unlikely to ride the same position during you entire ride. You might stand up for some climbs. You may sit back some or ride in the drops at times. all these chance your weight distribution and effective wheel circumfrence.

Don't forget that the road has bumps which will bounce your tire and cause momentary changes in effective circumfrence. These may or may not cancel out, I don't know.

Keep in mind that you may not ride the same line as someone else. Depending on how straight you hold your line and what line you take through the curves, your computer will show a bit more, or less than someone else's, and perhaps more or less than the published distances for a ride. When you stop for a break, you may not stop your bike right on the road and add a few hundreths of a mile riding through a prking lot or two.

Finally, most bike rides are informal affairs and any published distances by the organisers should be taken with a grain of salt. The organisers probably either used some mapping software to compute their route distances or drove the route in a car. Either case is highly prone to inaccuracy.

The bottom line? Use your best judgement and don't worry about it. There is no magic to the 100 mile mark. Whether you do 98 or 102 miles, you're going to get the same benefit and feel the same soreness afterward. Remember, your computer will never precisely match mine; and as I stated earlier, my computer is absolutely dead-on perfectly accurate.

> the error associated with this measurement is probably going to be, at worst, 5mm for a single revolution. 5mm error for a 2100mm circumfrence wheel comes to 0.23% error. On a ride of 100 miles, this equates to an error of +-0.23 miles. So, if you had this uncertainty in your setting, you can simply ride an extra 1/4 mile and you will know that you have ridden at least a full 100 mile century.

Oh, I do that - did two extra miles based on my computer and 6 or 8 extra based on the actual century distance as shown on my computer. I'd be kind surprised if I would be 5 mm off since every test seemed to be within a half mm. I can see a potential problem, though it may be meaningless. Sitting on a bike on slowly rolling it may cause a greater compaction of the tire than is the case when spinning fast. But I don't know if that is the case.

> First off, when you do your rollout test, unless you have a helper push the bike while you are on it, chances are you will not be in your exact riding position. <

True. I think I minimized this by being against a wall so I could maintain more stability in this area.

> Second, even if you are able to get into position, you are unlikely to ride the same position during you entire ride. <

Also true, though I suspect with fully inflated tires, the difference is pretty small. It would be interesting to test it though. Maybe someday I will.

> Don't forget that the road has bumps which will bounce your tire and cause momentary changes in effective circumfrence. These may or may not cancel out, I don't know. <

Yeah, but as it happens, since I made my careful measurement, I've had a lot of the road surfaces I ride repaved so they are much smoother. If there is any difference, it is too small for me to tell.

> Keep in mind that you may not ride the same line as someone else. <

True, but like I mentioned before, even deliberately weaving far more than any normal rider would, it didn't make the trip more than 2% longer - maybe just 1% (I can't recall the exact result of this test now). Similarly, I think it would take a HUGE difference in how you take a curve to really matter.

> When you stop for a break, you may not stop your bike right on the road and add a few hundreths of a mile riding through a prking lot or two. <

Yes, and this was something that bothered me about the century here. Like I said, it was 4-6 miles off (can't recall for sure), but that included the distance I put in at the various aid stations which usually were not right along the road. It was more like going up a driveway to a station. To me, a century ride should be based on not taking any such detours at all.

> Finally, most bike rides are informal affairs and any published distances by the organisers should be taken with a grain of salt. The organisers probably either used some mapping software to compute their route distances or drove the route in a car. Either case is highly prone to inaccuracy. <

I still hope to get some real info on the map accuracy. Since the figures are so close to my actual measurements, I tend to think my software is pretty accurate, but that could be coincidence. If someone had the Topo USA software and a GPS unit where they could measure a good distance on a fairly straight course, it would give us some useful comparisons. Don't GPS units have an accuracy to a few meters? Over a distance of 10 miles or so, that isn't much even if the error is the maximum and in opposite directions at each end. Of course, if I was doing it (stickler that I am), I'd actually measure the two points several times and go with the average of each!

> There is no magic to the 100 mile mark. Whether you do 98 or 102 miles, you're going to get the same benefit and feel the same soreness afterward. <

Yeah, but I just like to be able to say confidently that I did a century, or a metric century, or a half-century. I'm constantly pushing to reach at least 63 miles just to be absolutely sure I did a metric century. I've only done 3 centuries in four years, but I have done 13 metric centuries or further just since mid-december.

Hahahahahahahahah!!!!!!!!

I develop airspace and we deal with map accuracy quite often. Here's what I know.

We consider a standard USGS 1:24000 map to have an accuracy of +/- 40' horizontally from any plotted point. The digital map data you get is generally based off something that's based off the same data a printed map is. But for airspace planning where a shift of 5-10 feet can mean a huge difference in the way an airplane arrives to an airport, you don't get any more accurate than the 1:24,000 USGS maps.

USGS digital elevation models, which I imagine support most of the digital data being used, is accurate at the +/- 13m level for airspace planning. All this to say, you're certainly more likely to be more accurate on your own cyclometer or GPS than what you can get from a map.

I use programs called ExpertGPS and GPS Trackmaker to do a lot of my route planning, with fantastic zoom capabilities. They provide me accuracy down to around 1-2km difference from what I bike when I "point and click" a line from start to finish along a properly calibrated digital image of a digital map. I'm doing a new century route tomorrow based on one such "point and clicked" route. I'll post back how close the GPS/cyclometer (which I've gotten within .03km on a measured course 40km) are to the point and click method.

VW

Yes, most modern GPS recievers do calculate altitude, (with a minimum of 4 satellites). However, they don't take this into account in calculating speed and distance. That is exactly why the Garmin Edge 305 GPS reciever, designed specifically for use on a bicycle uses a wheel magnet to attain true speed readings.

You don't have to actually jump off a building to test it yourself. Just watch the speed that is displayed when going down a steep incline. While I usually hike at about 3.5 to 4.5 mph, when going down a steep incline, (and obviously moving faster than usual), the GPS thinks I've slowed down until I hit flat ground again.

Are you certain that your measuring tape is accurate?

What a sight! I like it

This is, essentially, the same thing as a bike computer (I'm sure the irony was intended). The only significant difference is that the bike wheel is semi-deformable. However, the bike wheel is considerably larger and therefore less affected by small uncertainties in radius, and accordingly, circumference (I'm referring to absolute uncertainy, not relative).

As I mentioned previously, for a typical 700c road wheel you need to compress the wheel by ~3mm for an entire revolution to introduce an error of 1%. If you sit on the bike for the rollout, use road tires, and keep your tires properly inflated, it is hard for me to belive you'd exceed this (e.g from riding position, etc) in a systematic manner.

Sorry for being repetitive, but I feel pretty strongly that even with all the possible sources of error, if used properly, your bike computer will give you a very accurate measurement of the actual distance you ride. This isn't hard to do, and you'll know your milage within 1% (less than one mile on your century).

Your measurement procedure was correct. While you can vouch for your own, you certainly don't know how other riders set their computers. I agree that the difference between the computers is beyond what would be expected due to the internal electronics or the computational methods and "rounding" that the different computers might use. I have a riding friend that on many occassions we may ride 60-70 miles. We have different computers and have never tried to intentionally set them to match. On a 60-70 mile ride our odometer difference would be less than 1 mile.

Todays century ride cyclometer vs GPSr measurements:

GPS 1 (eTrex Vista, in the aero bars): 160.14km (99.50 mi)
GPS 2 (GPSMAP 60CS, on the camelbak with external antenna): 160km ( didn't turn the tracklogging on - woops, but the trip odometer shows an even 160km, can't get the decimal point without the tracklog )
Cateye Astrale 8 cyclometer: 162.85 km (101.19 mi)

Considering the cyclometer takes distance based on every rotation of the wheel, while 655 of 2235 trackpoints on the eTrex were over 100 meters away from their previous trackpoint, I'd say the cyclometer was probably spot on for the distance of wheel revolutions. Good lord knows I don't ride any 100 meter stretch in a perfectly straight line - and some of the track points were 270 meters apart.

Meanwhile, I checked my turn around point on my "point and click" route to my actual turn around point (when I hit the 80km mark on the GPSr), and the two points were roughly .2 km difference. The actual turn around was .2km shorter than the turn around that I "point and clicked" on the map.

> Considering the cyclometer takes distance based on every rotation of the wheel, while 655 of 2235 trackpoints on the eTrex were over 100 meters away from their previous trackpoint, I'd say the cyclometer was probably spot on for the distance of wheel revolutions. Good lord knows I don't ride any 100 meter stretch in a perfectly straight line - and some of the track points were 270 meters apart. <

Thanks for the info.

I never had a GPS, but I wonder about somethings. You mention the track points which I assume are the specific points a new reading is made?

If so, then I assume that the distance is based on a straight line between those two track points.

If you are going around curves, this would obviously give the wrong distance and if you are doing sharp turns like switchbacks on a mountain, it would be way off. If changing altititude, that would give erroneous figures too, but I think it takes a pretty steep hill to really make a notable difference.

But if you were riding a straight and level road (granted, not likely for a century!), then it would not matter how often the reading were made, would it? The only factor creating error would be the built in error. So, if the built-in error was plus or minus 50 ft, then the maximum error would be 100 feet (if one end of the ride was in error by +50 and the other end was off by -50).

Now, if you rode a straight course for 10 miles (52,800 feet), you would be off no more than 2/10 of 1 percent! If it was just 5 miles, the error could be 4/10 of 1 percent. Heck, if you can get even a two mile straight course, your would still be within 1 percent. Doing the test periodically and averaging the distances would increase accuracy considerably.

I don't know how accurate your GPS or other ones actually are, but I thought 50 feet was likely a worse case scenario for a decent unit.

Granted, none of us holds a perfectly straight line when riding, but the GPS info would help narrow down the error. I mean, if the GPS is, at worst, off by 100 feet on either the high or low end, then you know, wobbles or not, you can't have a result lower than the lowest possible GPS result. If you do, then you calibrated too low. The high end is a bit trickier since you can't tell how much you will wobble from a straight line.

Bob

Exactly. Like measuring the circumference of a circle with a straight edge ruler.

And that's if you get a good signal around a mountain switchback at all. Riding along a mountain side will generally give you a bad reception as it is, adding more error to the equation.

If you could find a straight line for say, 1km. You could take most good GPSrs and mark a waypoint at each end of the distance and "average" the start and end point markers. With most GPSrs, they'll allow you to get a good avg of 1-2' of accuracy if the GPSr sits in place for 15 minutes or so. Then you can get a good solid reading for that 1 km distance, hop on the bike and ride from one painted line on the ground to the next and you'll know how close your cyclometer is. Once you put the GPSr on the bike with you and you start moving, it's going to take readings every couple seconds and see how "straight" it was from the last reading and determine whether or not it should plot another point on its route. A two point line can be accurate to the low accuracy described above. Once you start moving, every point along the route will be "placed" with the accuracy at its location - likely anywhere from 10-50feet. So on a route for 2300 track points, each of the 2300 trackpoints will have a 10-50 foot (sometimes more, very rarely less) accuracy.

The assumption is that the accuracy is always off in the same direction, so that the accuracy readings would offset each other, which is an assumption that just can't be real. But at the end of the day, I'm pretty pleased with a 1.6km difference between my cyclometer and GPSr over 160km. It at least attests that all of the "inaccuracies" in measurements of our simple wheel rotation measurement or high-speed GPSrs all equal out to very little - however they are doing it.

> With most GPSrs, they'll allow you to get a good avg of 1-2' of accuracy if the GPSr sits in place for 15 minutes or so. Then you can get a good solid reading for that 1 km distance, hop on the bike and ride from one painted line on the ground to the next and you'll know how close your cyclometer is. <

Sure wish I had a GPS! But hard to justify since my use would be rather limited - mostly just get a few really accurate readings as you describe. Once that was done, I wouldn't have a lot of use for it. Similarly, I'd like to have some serious altitude checks. I think the maps would likely be more prone to error on this than on distance measurements. Likewise, I think the GPS units have a wider margin of error for this, but again, by averaging multiple times, I could get a much better feel for how close my topo map software is. But once I've checked some enough, I wouldn't need it anymore.

Thanks for the info.

I didn't say it wouldn't work in the woods. I said it will work POORLY.

So it all depends on what you want it for. If you want a "ballpark estimate" of your location so you can find yourself on a national park map, than it's an EXCELLENT tool. If you want to use it to calculate your distance, than it's not nearly as good as a $50 cyclometer.

For a fairly straight trail on a VERY, VERY large property (like on national parkland) GPS can be a great mapping tool that will give you an excellent approximation. If you are dealing with a much smaller park where the trail tends to be twistier to fit more in, a standard GPS receiver really isn't very useful beyond getting a "starting point".

MINIMUM error for a GPS (unaided by WAAS) is 3m. Add foliage and that error starts shooting up.