View Full Version : Approach Trail "Official" Length
trippclark
12-14-2011, 12:08
I first hiked the Approach Trail back in 2001. My recollection is that at the time, the official length was listed as 8.8 miles. This past spring, I had the opportunity to hike it again, when I re-hiked from Amicalola Falls State Park to Neel Gap. I was especially looking forward to rehiking the Approach Trail since I was aware that it had been re-routed to the staircase adjacent to the falls. With the former route, you scarcely got a glimpse of this impressive feature. I was surprised, however, that in spite of this major re-route that the official distance had remained unchanged at 8.8 miles.
Over the past few years, I have been carrying a GPS device with me. I have found it interesting to keep a track log while I hike, and to record waypoints of interest (campsites, water sources, overlooks, parking areas, etc.). During the hike, it provides a very accurate tool to determine distance to next landmark or destination. Even though "distance to destination" is a fairly useless straight line distance, distance traveled, when a track log is used, is quite accurate, so I can easily see that I have traveled, say, 12.3 miles so far and therefore have 2.7 miles to go, or whatever. It is also interesting after the trip to load the tracklog file into Google Earth and Garmin's Connect and get a different perspective on the trip. I also share the data files with Awol as a resource for "The A.T. Guide."
When I started at the arch at Amicalola Falls, I reset the tracklog as new. I was a bit surprised when I reached the Springer Mountain Summit to find that my tracklog file was well below 8.8 miles. When I got home, and pulled the file into MapSource software, and cleaned up the file (remove stray points where I stepped off the trail for a break or to take a picture, etc.) I ended up with an 8.3 mile measurement for the Approach Trail, from the Arch to the Springer Mountain Summit.
I am fairly certain that this is accurate. Having compared my track log files to AT published mileages from several trips, I have rarely found a variance of more than 0.1 or 0.2 over a 15 - 20 mile day. So, I might could be persuaded that the actual mileage is as much as 8.4 miles, but not 8.8.
In the grand scheme of things, this is no big deal, and I readily admit this. The mileage is what it is, and I am just happy to be able to walk it. But as deliberate and precise as the measurements seem to be of the 2181 miles (year 2011) of the A.T., I find it surprising that the Approach Trail is so far off from the official length, given that it is such a short and much traveled trail. I am curious if anyone else has noticed the "new" Approach Trail being well under 8.8 miles. Regardless, I thought I would share this info.
Rain Man
12-14-2011, 12:18
... as deliberate and precise as the measurements seem to be of the 2181 miles (year 2011) of the A.T., I find it surprising that the Approach Trail is so far off from the official length, given that it is such a short ... trail. ....
If I'm not mistaken, counter-intuitive logic applies here. The larger the quantity, the more the precision. That's because these little discrepancies are balanced out over the long haul and thus "come out in the wash."
Still, I too am curious about your results.
Rain:sunMan
.
trippclark
12-14-2011, 16:31
If I'm not mistaken, counter-intuitive logic applies here. The larger the quantity, the more the precision. That's because these little discrepancies are balanced out over the long haul and thus "come out in the wash."
Still, I too am curious about your results.
Rain:sunMan
.
Good point Rain Man. That does make sense.
I should elaborate, for clarity, about my statement where I say that I sometimes find a variance of 0.1 or 0.2 over a full 15 - 20 mile day. This is almost always that my measurement is very slightly less than the official mileage. I am quite sure that this is accounted for by the cumulative effect of the way the track log is created. The device automatically records a series of many hundreds of points throughout the course of the day. Each of these points is a few feet apart. The track log then connects the dots sequentially to create a linear path. Obviously, this connected series of hundred of tiny straight lines would be expected to be very marginally shorter than the actual curvature of the trail on the ground. What I have found to be true is that over the course of a day's hike (15 - 20 miles typically), that this difference is so small as to either have no measurable impact (less than 0.1 mile) or, at most, result in a difference of 0.1 or maybe 0.2. Again, the difference that I observed on the Approach Trail of 0.5 mile was way too much to be accounted for by this, leading me to believe that the generally accepted length of 8.8 miles is not accurate.
ChinMusic
12-14-2011, 17:15
trippclark - I have used a GPS for years. I love gadgets. Unless you have a rogue data point your GPS data will typically show less miles than you actually travelled. This applies to meandering trails and not straight-line travel. The GPS will record data points and assign a straight line between them, when you may have actually done a switchback. Those little cut corners add up over a days hiking. The less often your GPS is set for new points or the harder it is for the GPS to maintain location the more the loss will be. Under heavy tree cover the same trail will record a shorter distance than when the leaves are off. It is not uncommon to come in 2-5% under the true distance.
trippclark
12-14-2011, 17:32
trippclark - I have used a GPS for years. I love gadgets. Unless you have a rogue data point your GPS data will typically show less miles than you actually travelled. This applies to meandering trails and not straight-line travel. The GPS will record data points and assign a straight line between them, when you may have actually done a switchback. Those little cut corners add up over a days hiking. The less often your GPS is set for new points or the harder it is for the GPS to maintain location the more the loss will be. Under heavy tree cover the same trail will record a shorter distance than when the leaves are off. It is not uncommon to come in 2-5% under the true distance.
Yeah, I agree, and that is what I tried to explain in my second post. By the way, I am a gadget fan also.
My GPS has three settings to select from for frequency of points in the tracklog. On my shorter trips (5 days or less), I always use the highest setting (most points) in order to minimize this variance. That is why my measurements tend towards the 2%. 8.8 is more than 6% more than 8.3. I'll concede that my 8.3 measurement might actually be as much as 8.4 on the ground, but I think that 8.8 is a stretch. I would love for others to take the same measurement and see what they come up with.
If you're comparing it to one of the guides, such as the Companion, can't you tell where the data mileage is off against the Garmin map? Also, I was under the impression that GPS don't account for elevation very well.
You may want to share your findings with the GATC.
trippclark
12-14-2011, 18:46
If you're comparing it to one of the guides, such as the Companion, can't you tell where the data mileage is off against the Garmin map? Also, I was under the impression that GPS don't account for elevation very well.
You may want to share your findings with the GATC.
Sly,
Good idea about sharing with the folks in Georgia. I will do so.
Regarding accounting for elevation, all I can go on is my own experiences. I have kept track logs for all of my hikes the past 3 years, which accounts for about 400 miles and roughly 35 - 40 days. Over those trips, the daily mileage on the GPS has always been either spot on or within 0.2 of the official distance. The one exception was the approach trail, which was off 0.5 over a half day.
How do the measure the distance?
I found this interesting...
http://www.nynjtc.org/workshops/training/measure.html
It suggests they use a measurement wheel, and get about 1% accuracy over fair terrain and 2% accuracy over rough terrain. Those are for distances of one mile. If those are random errors, and say the error is 100 feet per mile, the error over 2100 miles would be SQRT(2100x100x100) = 4852 feet, or about a mile. However, I would guess there would be some systematic error as well. Not sure which way. Anyhow, they probable report the distance within 5 miles so it is appropriate that they use all the digits, but it wouldn't make sense if they were to get into tenths of a mile over the full distance. But it would be useful to carry those extra digits in adding up the total.
So, 2181 miles +- 1 mile, would be a fair guess.
Not sure how GPS measures distances along trails, but I would guess they are less accurate, and there might be some systematic errors when compared to measurement by a wheel. I would guess they tend to underestimate wheel distances. Not sure.
The only idea I have to why there is such a large difference, may be the stairs. I'm under the impression that most of the mileage data compiled on the trail is reported via wheel. Are the people walking with the wheel pushing it beside the stairs where possible or on the steps? If pushing on the steps, is the wheel being pushed to measure just the horizontal length or is the vertical included? I know adding the vertical would only be a few inches per stair, but in total it may be the variation. Let us know what you find out, if anything.
trippclark - I have used a GPS for years. I love gadgets. Unless you have a rogue data point your GPS data will typically show less miles than you actually travelled. This applies to meandering trails and not straight-line travel. The GPS will record data points and assign a straight line between them, when you may have actually done a switchback. Those little cut corners add up over a days hiking. The less often your GPS is set for new points or the harder it is for the GPS to maintain location the more the loss will be. Under heavy tree cover the same trail will record a shorter distance than when the leaves are off. It is not uncommon to come in 2-5% under the true distance.
That's my experience also. I've measured trails with a wheel, and the same trails with a Delorme gps. The wheel always finds more miles. The difference I think is that the wheel measures every foot. The GPS adds up the distance between points on the trail, thus skipping all the irregularities between those points. My Delorme allows me to choose the distance between the points. But since I'm still a beginner in GPS matters, I don't recall the details.
Typically if you are going to use a GPS to map a trail, you don't use a handheld model you found at an outfitter. There are more purpose built units for that that log far more points then a typical unit does and they usually are more sensitive to Satellite signals. The whole problem with GPS tracking can be summed up by: 1) the distance between datapoints is assumed to be a straight line and may miss direction changes such as switchbacks or some elevation changes that occured inbetween points. 2) Your position accuracy. The GPS is only accurate to several feet at its best but that accuracy varies over time and can be more then 30ft off. Its' affected by the number of satellites locked on and the signal quality which is affected by terrain. This accuracy error will introduce some small error in your position. I'd like to think this error would average out over a long enough distance, but I don't know the type of statistical error it is to say for sure. 3) And as mentioned, GPS don't always account for elevation changes on a trail properly which can change the distance measured.
That said, if the GPS unit is constantly receiving good signals from enough satellites , the trail is pretty straight, and you are logging enough points to account for most trail direction changes, it is possible to get a fairly accurate track. But I wouldn't trust it enough to make a guild book out of it. I could tell you the story about a guy who did just that for the PCT and how much variance there was verses the official distance, but I'll spare you. His overall distance was pretty close, but there was alot of variance in the day to day distances inbetween. I believe I heard he was trying to redo sections of the track with a more professional GPS tracker due to feedback about the issue.
ChinMusic
12-14-2011, 22:59
I just checked my GPS data from 2008. I got 8.6 miles which makes me believe the 8.8 miles as listed is accurate. I would expect my 8.6 to be slightly under reality.
trippclark
12-14-2011, 23:51
I just checked my GPS data from 2008. I got 8.6 miles which makes me believe the 8.8 miles as listed is accurate. I would expect my 8.6 to be slightly under reality.
Thanks. That is the sort of info I was hoping to uncover. Additional measurements to either confirm or correct the measurement that I found.
When I first hiked the trail back in '78, I think the advertised length of the approach trail was 7 miles. I'm pretty sure they were off by a good bit. It took me three days, so it must be at least 45 miles.
Az
MuddyWaters
12-17-2011, 00:14
A wheel will roll up and down every ft-sized imperfection in the trail, creating a longer distance than a persons stride-length which skips over a distance of about 2.5 ft with every step.
Consider that the smaller the wheel, the longer the measured length will be. If you had a wheel the size of a grain of sand, the distance might become infinite. What is the true length depends on the use of the measurement, there is no absolute.
Since we actually dont walk those small imperfections, a wheel likely is more accurate than needed. However, a gps may be less so.
smithjerd
12-17-2011, 12:09
fractals I'm hoping (but not believing) that the elevation gains listed in the AT itinerary spreadsheet were measured with a wheel the size of a marshmallow. :)
A wheel will roll up and down every ft-sized imperfection in the trail, creating a longer distance than a persons stride-length which skips over a distance of about 2.5 ft with every step.
So if you hop, skip and jump the trail, it's only 1500 miles? :-?
The gps sattelite is looking down from orbit and sees the trail as a 2 dimmentional object when in fact is 3 dimmentional.
Panzer
smithjerd
12-17-2011, 20:03
I respectfully disagree. The GPS sattelites (usually 4 or 5 of them are communicating with your GPS device) 'triangulate' on your position and read altitude as well. My inexpensive GPS device could normally see changes in height in the range of 10 feet. I'm no GPS expert, but I agree with previous posts in this thread that attribute differences to linear interpolation of distances between points, rather than the curved (longer) true path the trail takes.
my 2 cents worth.
smithjerd
12-17-2011, 20:19
ok before someone corrects me. I looked it up on wikipedia and GPS actually uses trilateration rather than triangulation. But the same idea about being able to detect changes in altitude applies.
learn something every day.
When I first hiked the trail back in '78, I think the advertised length of the approach trail was 7 miles. I'm pretty sure they were off by a good bit. It took me three days, so it must be at least 45 miles.
Az
I hiked it in 1977... I seem to remember 7 miles. Didn't matter, I was so excited to get going on my trip I think I did it in just over 2hrs. I've never been one who obsessed over 1/10th's of a mile. Thos that do usually don't hike very far IMO.