Accuracy of trail distances on signs and in guide books

I am sure everyone has wondered at some point about trail accuracy or felt like they hiked just a bit further than the sign had said.

My main question is this. What historically has been the method for determining the trail distances we see on the signs and in guide books?

Linked below is a good article about calculating trail distances. The author gives simple explanations to various method that can be used and points out that only with traditional surveying equipment (transit, pole, prism) will you get a very accurate reading. Even then you assume that you walk perfectly straight between the two points.

http://www.qcc.mass.edu/brink/general/distance.html

I have been thinking about this quite a bit since I do survey work and from my recent trip out west which had trails that were nothing but endless switchbacks. I got to thinking that there has to be quite a bit of historical and current error, perhaps more than 10% in cases of rugged terrain. I would speculate that it is more common to underestimate distance simply do to the non-linear nature of hiking and mountaineering, slope of the trails, the frequent obstacles and the obvious difficulties pointed out in the article.

As hikers we encounter the 'slope distance' factor quite a bit. It is basically an example of the Pythagorean theorem. The hypotenuse of a right triangle is longer than it's base. The slope, or an uphill trail, is the hypotenuse. The base of the triangle, or map for example, is a level plain. Simply put, walking 100 feet on a perfectly level trail moves you 100 feet on a map according to it's scale; walking 100 feet up a steep Adirondack trail does not move you 100' on the same map. The distance would fluctuate based on the grade and would always be >100 feet.

Surveying a property also assumes a level plain, thus the distances do not account for slope. If you own a hillside, you actually own a greater walking distance than you're legal description calls for just as in my example above.

The author of the article I mentioned makes good points about GPS positions on KEY points (curves, switchbacks, etc.). However, simply mapping a trail with a hand held GPS will not get you that accurate of a reading do to frequent in accurate positions. Too achieve good accuracy would require a lot a careful and methodical work. I frequently use pole-mounted Trimble GPS with an accuracy of +/- 3' after differential correction and I know how difficult it can be to get an accurate reading. You need 1200 one second positions to be averaged and 'corrected' with software in order to get this level of accuracy with equipment that costs in excess of $10,000 dollars. Even then in forested areas and hollows, and just depending on the satellite positions, don't expect the same result twice.

I being a bit technical here, but it is a curious question. Just how accurate are trail signs? Maybe it really did feel like it was a bit longer!

I wish I would have thought of this when I was in college. It would have made a great project. Hopefully one of these days I'll have the time to check out the accuracy of GPS with professional equipment if for nothing else sheer curiosity.

-Dom

I don’t know what current practice is, but in the not too distant past the preferred trail distance measuring tool was a larger diameter measuring wheel. That gauged the actual distance walked with some accuracy, whether the trail was level or otherwise.

Now, let’s take your question about the Pythagorean Theorem, and work a problem, just for fun.

We’ll pick a steep trail, one that rises 1,000 ft in a mile. Using the theorem, we find that the hypotenuse (trail) is 5373.9 ft, just 93.9 ft or about 1.8% longer than a level mile on the map. That rate would add a quarter mile to your hike every 14 miles. You decide how consequential that might be.

G.

I know that in the AMC SNH and WM Guides, nearly all, if not all, the distances are wheeled.

I think the signs disagree with the books so that you will learn not to trust them!



The Catskills have been wheeled as well, in the not too distant past.

I remember having a discussion about trail lengths on the trail. A guy I was with had a GPS, and he said "this HAS to be accurate, it takes 100 points per mile!" I said, "look around, at the last 50 feet and the next 50 feet of the trail. Are they straight?" In most cases the answer is no. So, what's the real distance of that section of trail?

I think the reference page is a source only to that guy's opinion and not to good practice

For instance I think dead reckoning was originally done by elapsed time, which can still be a good method if done carefully

As mentioned many trails are wheeled, but of course there are errors in wheels too - exact circumference may be off, do you measure centerline of trail or take the wandering course among rocks the hiker is likely to choose, experienced wheel users can spend hours debating arcane points

When my sister wants accurate GPS trail measurements, she makes sure to stop at the end of every switchback longer that the interval for log points which can be a pain with short switchbacks but is necessary to get the true length rather than a cutoff. She also plots the log and deletes obvious error points before calculating the distance - a bad point can add a lot of bogus distance. Just running a track log and asking for a GPS distance is not especially accurate.

When I come to a sign I take a picture of it because that's what I do. As to the distance, that's where I'm going so the actual distance isn't all that important. It is whatever it is and I have to walk all of it to get there. I plan my trips pretty well in advance and leave plenty of "Goof Off/Up" time. I have never had my maps, GPS, trail signs, or guide books agree on distance. When my scouts ask how many miles we walked I 'll them "every one".
Bob

dom15931 said:

Linked below is a good article about calculating trail distances. The author gives simple explanations to various method that can be used and points out that only with traditional surveying equipment (transit, pole, prism) will you get a very accurate reading. Even then you assume that you walk perfectly straight between the two points.

http://www.qcc.mass.edu/brink/general/distance.html

Click to expand...

This guy doesn't look very knowledgeable...
* He states that computerized maps measure the slope distance. (They can do either--it is a program analyzing data and can do whatever the author wants.)
* He doesn't seem to know about sampling effects on digital maps.
* He doesn't seem to know much about GPSes and their errors. BTW, GPSes usually compute horizontal distance.
* He doesn't even mention wheel measurement.

The author of the article I mentioned makes good points about GPS positions on KEY points (curves, switchbacks, etc.). However, simply mapping a trail with a hand held GPS will not get you that accurate of a reading do to frequent in accurate positions. Too achieve good accuracy would require a lot a careful and methodical work. I frequently use pole-mounted Trimble GPS with an accuracy of +/- 3' after differential correction and I know how difficult it can be to get an accurate reading. You need 1200 one second positions to be averaged and 'corrected' with software in order to get this level of accuracy with equipment that costs in excess of $10,000 dollars. Even then in forested areas and hollows, and just depending on the satellite positions, don't expect the same result twice.

Click to expand...

Is your accuracy mentioned above and absolute or relative accuracy?
For trail measurement, relative accuracy over periods of, say up to 30 minutes (30 is just a guess), is more important than absolute accuracy.

I being a bit technical here, but it is a curious question. Just how accurate are trail signs?

Click to expand...

Certainly no more accurate than the methods that were used to measure/guesstimate the distances...
[/QUOTE]

As I have stated earlier, trail length is actually a fractal and there is no unique length. (For instance, a wheel distance will depend upon the diameter of the wheel.)

Doug

All that really matters to me is that the same method be used consistently throughout a guide book or map, so that a trail which is labeled 5.0 miles long is, within a few % points, twice as along as a trail which is labeled 2.5 miles. In this way I can make a relatively accurate prediction of time, given the distance, and elevation, because I have previous experience to draw on.

FWIW, I kept track of distance and elevation as reported in the WMG: 27th Ed., combined with TOPO 4.0, for my any season NH 48, and I get:

Peaks: 48
Trips: 26
Miles: 288.8 (11.1 avg)
Feet: 93,060 (3580 avg)
HH:MM: 191:20
MPH: 1.5
FPH: 485

Looking at individual trips, the FPH and MPH varied slightly depending on the terrain, time of year (of the 10 I did in winter, they were slower), etc., but for the most part, I tended to predict, within 15 minutes, how long a given trip would take me. You know, past results are no guarantee of future performance...

It is a frequently-noted observation that TOPO 4.0 is 10-15% shorter on distances than the WMG, for example, highlighting Doug's early point about digital sampling of map data. Elevation appears to be closer to accurate, although it certainly does not consider every single PUD.

Tim

Tom Rankin said:

I remember having a discussion about trail lengths on the trail. A guy I was with had a GPS, and he said "this HAS to be accurate, it takes 100 points per mile!" I said, "look around, at the last 50 feet and the next 50 feet of the trail. Are they straight?" In most cases the answer is no.

Click to expand...

Someone else who doesn't know much about GPSes or measurement in general...

You describe the geometric errors due to a coarse sampling of a continuous track.

Most consumer GPSes take one measurement per second*. (Some special purpose units can take up to ten or twenty per second.) When one records a track, the GPS saves some subset of these points (Garmin GPSes have user-settable criteria). (When you save a track on a Garmin GPS, it further reduces the number of points to 250 or less for the entire saved track.) The distance that you compute depends on which sampling you use. And, of course, each of the points has some amount of error.

* GPSes use smoothing across points to reduce the absolute error. Thus each measurement is not independent of the adjacent measurements. This will tend to reduce random errors, but will also tend to overshoot corners, etc.

FWIW, in GPS bakeoff #1 (hiking on a trail), we used too coarse a measurement interval. http://www.vftt.org/forums/showthread.php?t=14406. In bakeoff #2 (walking/biking on relatively straight paved paths), I used 1 sample every 5 sec for walking and 1 sample per sec for biking (both ~30ft of travel) to minimize sampling effects. http://www.vftt.org/forums/showthread.php?t=15025

Doug

All I care about is that the distance is correct to within 20% or so. What matters to me, is time, and distance (and elevation) only show a part of that. If all trails were hard packed, and all hills had equal sized steps, maybe I'd care more about distance.

I agree that consistency is more important than absolute accuracy, particularly when there is no fundamentally correct length*. IMO, if several reasonable methods come out within 5 or 10% of each other, that is good enough for me.

* Since there is no fundamentally correct length one cannot compute an absolute accuracy, so the general concept will have to do...

Unfortunately, the NG TOPO! distances fairly consistently come out different enough from the WMG to be annoying or disturbing.

Doug

Several of you guys have mentioned that their is no mention of using a measuring wheel in the article. That is a good point. I was thinking about that after I first posted, but was off to bed.

"Dead Reckoning" would in my opinion be a waste of time. In rugged terrain maintaining an even stride, let alone counting it, would become impossible. That method is generally only used, in surveying at least, for shot 'guestimates' when attempting to locate or measure something quickly to get a rough idea of where something should be or how far away from another point you are.

Doug makes some excellent points on GPS and consistency. One thing about GPS is when moving it becomes less accurate, when standing still it is still measuring if the satellite configuration changes and have a 'route' function (at least the one's i have used). I know when using the professional equipment we have at work you are to sit the unit still for 5 minutes to acquire a good satellite configuration.

A good way to notice this is with GPS elevation, which is always less accurate than the horizontal measurement. By letting your GPS sit on top of a stump or large rock I have seen handhelds change by more than 150' over the course of a few minutes.

As for how knowledgeable the person who wrote that article is, probably not the most being that he left out the use of a measuring wheel. I also didn't see anything about using a measuring rope, which if you are gonna include dead reckoning, you might as well include that as a method.

So it looks like using a wheel has been the historical practice? Would make sense as it would account for slope distance and curves. This method would only get difficult in steep rugged terrain or very rocky surfaces in which the wheel would not function quite as well.

DougPaul said:

Unfortunately, the NG TOPO! distances fairly consistently come out different enough from the WMG to be annoying or disturbing.

Click to expand...

Every time this comes up though the general consensus is that it undershoots distances by 10-20%. Thus I tend to add back that 15% or so if I am relying on TOPO! for distance. Since the use of TOPO! requires hand tracing the route, a certain amount of error will be introduced by the mouse movements anyway. At least for me, I am probably consistent with myself in how I trace out the routes, and so trip-to-trip, the relative accuracy (or error if you prefer) remains fairly constant.

As an experiment, I traced the Wilderness Trail from Lincoln Woods to Bondcliff Trail which, as we all know, is pretty flat, straight, even, etc. There should be minimal errors in the digital sampling, and minimal errors in my tracing.

TOPO 4.5 comes up with 4.74 miles and +448 feet. The WMG, 27th ed says 4.7 miles, +450 feet. That's pretty accurate.

I also traced the Wildcat Ridge Trail, notorious for steeps and ups and downs, and came out with 3.93 miles and +3110 feet, while the WMG says 4.2 miles and +3150 feet. The elevation is pretty close (40 feet), but the mileage comes up 3/10th short by 6% or so.

Both were traced at full zoom (level 5) and 300% magnification, which is my standard method for tracing routes.

Tim

dom15931 said:

"Dead Reckoning" would in my opinion be a waste of time. In rugged terrain maintaining an even stride, let alone counting it, would become impossible. That method is generally only used, in surveying at least, for shot 'guestimates' when attempting to locate or measure something quickly to get a rough idea of where something should be or how far away from another point you are.

Click to expand...

Dead reckoning may be of limited use in surveying, but is still very useful in both land and nautical navigation. (I'm sure GPS has largely replaced it at sea.) In both bushwacking and boating, one often goes off on some approximate heading for some approximate distance or until one reaches something.

So it looks like using a wheel has been the historical practice? Would make sense as it would account for slope distance and curves. This method would only get difficult in steep rugged terrain or very rocky surfaces in which the wheel would not function quite as well.

Click to expand...

I believe a wheel has been the historical practice. The pre-electronic alternatives would be surveying or tape measure and compass + inclinometer. (I believe tape measure and compass + inclinometer are still used in caves.) A wheel would simply be quicker and easier.

Doug

bikehikeskifish said:

Every time this comes up though the general consensus is that it undershoots distances by 10-20%. Thus I tend to add back that 15% or so if I am relying on TOPO! for distance. Since the use of TOPO! requires hand tracing the route, a certain amount of error will be introduced by the mouse movements anyway. At least for me, I am probably consistent with myself in how I trace out the routes, and so trip-to-trip, the relative accuracy (or error if you prefer) remains fairly constant.

Click to expand...

My experience is similar.

http://tchester.org/sgm/analysis/trails/chantry/topo.html observed underestimates of up to ~30%.

Sampling effects alone can either increase or decrease a distance. A lot depends on the details of the unpublished algorithms used in such programs.


Seems to me we have had this conversation more than once in the past...

Doug

I also use TOPO, tracing the route beforehand, and let the GPS follow me on the route. For me, the readings from all sources agree enough so I am not concerned.
In situations where the GPS has had bad reception areas, of course, it undereads. For a GPS check on accuracy, try this one:Mark your car's waypoint before you leave. When you return place the GPS in the same spot on the car and check it out.
Very interesting

dom15931 said:

I am sure everyone has wondered at some point about trail accuracy or felt like they hiked just a bit further than the sign had said.

My main question is this. What historically has been the method for determining the trail distances we see on the signs and in guide books?

I frequently use pole-mounted Trimble GPS with an accuracy of +/- 3' after differential correction and I know how difficult it can be to get an accurate reading. You need 1200 one second positions to be averaged and 'corrected' with software in order to get this level of accuracy with equipment that costs in excess of $10,000 dollars. Even then in forested areas and hollows, and just depending on the satellite positions, don't expect the same result twice.

I being a bit technical here, but it is a curious question. Just how accurate are trail signs? Maybe it really did feel like it was a bit longer!

I wish I would have thought of this when I was in college. It would have made a great project. Hopefully one of these days I'll have the time to check out the accuracy of GPS with professional equipment if for nothing else sheer curiosity.

-Dom

Click to expand...

Larry Garland and his assistants used a Trimble navigation system with differential correction when he located trail junctions and other points of interest in the latest versions of the AMC topographic maps. I know that Steve Smith and guidebook editors before him have used a measuring wheel for a lot if not most of the trails in the Whites. That said, the signs located at trail junctions are not always correct, as apparently information is commonly lost in the process of sign making.

Dr. Dasypodidae said:

Larry Garland and his assistants used a Trimble navigation system with differential correction when he located trail junctions and other points of interest in the latest versions of the AMC topographic maps. I know that Steve Smith and guidebook editors before him have used a measuring wheel for a lot if not most of the trails in the Whites. That said, the signs located at trail junctions are not always correct, as apparently information is commonly lost in the process of sign making.

Click to expand...

A few years ago, I ran into someone (on the top of South Twin) who saw my GPS and told me that some of the trails in the Waterville? area were being GPS mapped. Do you know if trails in the Whites have been GPS mapped and if so, which ones (in a general sense--I'm not looking for an exact list)?

Doug

DougPaul said:

A few years ago, I ran into someone (on the top of South Twin) who saw my GPS and told me that some of the trails in the Waterville? area were being GPS mapped. Do you know if trails in the Whites have been GPS mapped and if so, which ones (in a general sense--I'm not looking for an exact list)?

Doug

Click to expand...

I believe all of them.

Dr. Dasypodidae said:

I believe all of them.

Click to expand...

I thought so too. It was my understanding that (at least for the last two editions) the AMC trail maps were based on all of Larry's GPS work. At the very least Larry Garland has the distinction of being one of those who has hiked every trail in the WMNF.

Brian