Postholer.Com


A Place to discuss hiking along your favorite trail
Home Journals Maps Planner Postholer

Google trail maps WITH mileages!

Anything concerning the Postholer Google Maps. This may be feature requests, marker information updates or bug reports.

Moderator: postholer

Google trail maps WITH mileages!

Postby postholer » Wed Sep 23, 2009 1:20 pm

Postholer Google maps now has mileages for the PCT, AT, GET and JMT!!! Move your mouse over the blue trail trace and mileages will appear. Try it with the PCT:

http://postholer.com/gmap/gmap.php?trail_id=1

About the new mileage markers
At higher zoom levels with the map scale greater than 2 miles the markers are spaced a couple miles apart. As you zoom in the markers are spaced about 1/4 mile apart. This provides a great level of mileage detail at any zoom level.

This feature can be demanding on older, slower computers thus the 'Hide/Show Distances' link for the ability to turn it off or on.

Accuracy
The overall length of the calculated trail is 2,644 miles. The data used for the trace and obtaining the mileage is the USFS pct.kml. As a quality control check I've compared the calculated mileage with the PCT Data Book to see just how different they are. Overall, it's pretty impressive. The data is very close with only one big, glaringly obvious descrepancy.

Between Big Bear and Agua Dulce, the calculated mileage loses about 15 miles compared to the data book. That's alot! I haven't compared the guidebook and kml trace intensively for the error, yet. Maybe it's a frog detour, flooding or fire during mapping that is the culprit.

Here are a few comparisons of the mileage so you can get a feel of how close it really is.

Code: Select all
Segment                                  Data Book / Postholer Map
Cascade Locks(I-84) to Snoqualmie (I-90)     246.7 / 247.3
Castle Crags(I-5) to Crater Lake (Hwy62)     323.9 / 326.4
Kennedy Mdw to Toulomne Mdw                  239.9 / 247.1
Mexico to Pine to Palms (Hwy 74)             151.3 / 151.3


More than you wanted to know section
Early last year I was going to do something with elevations. When working with elevations or mileages each lat/lon point needs to be in consecutive order, that is, they can't be in some mish-mashed order. Each coordinate has to follow one after the other from the beginning of the trail to the end.

Consider the clean, well organized PCT kml file with 1501 segments (clumps of consecutive coordinates) totaling 257,276 coordinates.

Or worse, the very messy, unorganized AT (Appalachian Trail) data with 4,934 segments and 297,340 coordinates. This data set was night-marish and has become my standard by which all badly organized data is compared.

Your goal is to turn all these coordinates into one big, clean, consecutive, multi-thousand mile segment. They have to fit together correctly without gaps or overlaps to get the correct mileage, otherwise what's the point?

2 options, do it by hand or create a script to do this heavy lifting. Well the former wasn't an option, period. So it sat on the shelf until I could wrap my head around it. 3 weeks ago after a couple intense weeks of trial and error, I created the script to do the task. The best part is, it works on any kml file of any length. If the trail can be rearranged into one long segment, the script can do it. The result are a set of encoded polylines and distance encoding for google maps consumption. (Pssst, the script doesn't live on the web server.)

That's pretty much it.

Have fun!
-postholer 8)
Last edited by postholer on Mon Sep 28, 2009 5:47 pm, edited 2 times in total.
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

Postby p-fitz » Thu Sep 24, 2009 5:01 am

Wow, you da man Postholer! The site just keeps getting better and better!
p-fitz
 
Posts: 44
Joined: Tue Dec 18, 2007 5:27 pm

Re: Google trail maps WITH mileages!

Postby postholer » Thu Sep 24, 2009 9:14 am

Get the distance between any 2 points. Click once on the blue line, click a second time and the Start, End and Distance between the 2 points will appear.

This is real handy for planning a hike of a given length or getting distances between resupply points.

-postholer 8)
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

Most impressive!

Postby brianle » Thu Sep 24, 2009 10:08 am

Agreed, this is very cool beans --- thanks as always for your work.

I wonder if there's a way for your script to take in some additional reference points to periodically recalibrate to make the mileage data more accurate?

Looking at the AT plot, I randomly zoomed to the town of Pearisburg (southern part of Virginia). The 2009 AT Companion lists this at 625.9 miles from the (NOBO) trail start (access this for free in pdf form at http://www.aldha.org/comp_pdf.htm). Your map shows Pearisburg about 5 miles farther at about 630.8.

It seems to me that if you could load up the databook mileages for road crossings and shelter locations, those particular points could be synced to known NOBO distances and all other points in between recalibrated accordingly.

Yes, easy for me to suggest this when I have no personal sweat equity involved, but ... a suggestion, anyway! 8)

The particular reason I do think this could be helpful is with journal writers tending to list their current locations in terms of official trail mileages, if their readers look on the map they'll certainly get in the right ballpark, but sometimes perhaps be a little confused (FWIW).


Hmm, and while I'm on a roll, I know that the AT data that you wrestled so mightily with has nearly 300 shelters locations, and AT journalers tend to list their current location relative to specific shelters. Given that the waypoints for these are named in the data stream, how hard would it be to get those to show up on the AT map? In your copious spare time, of course ... :-)

One way to look at this is that if you were to focus on making Postholer clearly superior for AT hikers in particular, you hit a bigger audience, to include a lot of people who go on to do other triple crown trails. By not getting that audience from the start, inertia likely keeps a lot of folks on trailjournals even if they later hear about this site.

Am I shamelessly saying this because I personally am trying the AT next year and would love to see AT-specific improvements? You bet I am :wink: But it's true, nevertheless.
brianle
 
Posts: 26
Joined: Wed Apr 09, 2008 5:13 pm

Re: Most impressive!

Postby postholer » Thu Sep 24, 2009 11:41 am

Hey Brian, thanks for the QA!

brianle wrote:take in some additional reference points to periodically recalibrate to make the mileage data more accurate?


That falls into the manipulate by hand category and I'm way too lazy for that. Ultimately, it's a band-aid for correcting a difference in 2 different data sets.

...town of Pearisburg .... at 625.9 miles from the (NOBO) trail start ... Your map shows Pearisburg about 5 miles farther at about 630.8.


I re-ran the AT data with a new skew value. Obivously, the results will never be exact, but it is closer. Pearisburg is now at 625.9/624.7 and Delaware Water Gap is 1283.6/1289.9

In both cases this is less than 1% error. In context, hiking a 20 mile day means you will walk at most +/- a quarter mile any given day.


... has nearly 300 shelters locations, how hard would it be to get those to show up on the AT map?


Dude, they're already there! :) From the 'Location Type' list box select 'Shelter/Campsite' and you'll get icons and a list of names in the 'Location' list box.

One way to look at this is that if you were to focus on making Postholer clearly superior for AT hikers in particular, you hit a bigger audience, to include a lot of people who go on to do other triple crown trails. By not getting that audience from the start, inertia likely keeps a lot of folks on trailjournals even if they later hear about this site.


Well, I add features on a whim. Those features are incorporated across all trails. Yes, I understand that catering to one group or the other will increase 'attendance', I don't want that to be an objective per se (although I do have a soft spot for the PCT). TJ's has more AT journalists, that's it.

Hey, thanks again Brian and have fun planning that AT hike buddy!

-postholer
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

... hmm, but never too-o-o far off ...

Postby brianle » Thu Sep 24, 2009 11:48 am

Looking at a few more sample points, I've found other points much closer to the published mileage. But at least some points farther north on the trail seem to diverge a bit more.

About mile 1381 on your map, 1374.1 in the data book (where the AT crosses I-87) is about 7 miles off. Then about mile 1900 on your map (1911.4 in the data book) is 11 miles off (highway 26 crossing).

Yet somewhat in between those two points the U.S. 5 crossing at 1739.63 (yours, 1735.0 in data book) is just under 5 miles off, and close to there the Vt. 12 crossing at 1716.6 (yours, 1714.1 databook) is off just 2.5 miles. So it's not like there's some monotonically increasing error going either north of south.

I don't mean to make overmuch of this, but if a closer calibration were not too difficult, it would be nice to see. The feature is incredibly nice anyway, putting things in great perspective as I move the map around at different levels of detail and swapping between road, terrain, and satellite views. Thanks.
brianle
 
Posts: 26
Joined: Wed Apr 09, 2008 5:13 pm

Re: ... hmm, but never too-o-o far off ...

Postby postholer » Thu Sep 24, 2009 1:29 pm

For a different view, get a distance between 2 points that are 200 miles apart, or whatever, and see how close the mileages are to the published mileages.

So far we've comparing distances relative to the beginning or ending of the trail. That's how you'll find the largest errors.

-postholer
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

Mileages relative to start

Postby brianle » Thu Sep 24, 2009 2:34 pm

For most long distance hikers, "current location mileage" is relative to the start --- that's how they're typically listed in data tables. NOBO (or SOBO) mileages relative to the start are an unambiguous number to use for a "current location" in a trail journal, in email, etc.
brianle
 
Posts: 26
Joined: Wed Apr 09, 2008 5:13 pm

Re: Mileages relative to start

Postby postholer » Thu Sep 24, 2009 3:55 pm

brianle wrote:For most long distance hikers, "current location mileage" is relative to the start --- that's how they're typically listed in data tables.


You bet! Mileages relative from the start for keeping track of your trip makes the most sense.

Planning your trip and determining mileage/days between resupplies is pretty important and it's super easy between any 2 points. It beats subtracting mileages from the data book.

On that note, when you click on point 'A', then click on point 'B' not only does it tell the start mile/end mile/distance, it will tell you the number of days at 20 miles a day (just added that).

-postholer
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

days at X mpd

Postby brianle » Thu Sep 24, 2009 4:22 pm

This is also very cool. In particular, my wife will be able to click on this map and get a decent estimate (barring of course Nero/Zero time) for when I'll be at point Y --- nice!
brianle
 
Posts: 26
Joined: Wed Apr 09, 2008 5:13 pm

Re: Google trail maps WITH mileages!

Postby kdtsuei » Fri Oct 23, 2015 4:02 am

I plan to hike the JMT SOBO in the summer of 2016. Found your very nice software working with Google Maps. I wish to know the mileage at any point on the trail from the north terminus Happy Isles while the apparent mileage on the map shows from the south terminus Mt. Whitney instead. Of course I can do a little math to subtract off the known total mileage. Or I can measure the distance by two clicks on the map but making the first click on a part of the zoomed-in map then making the second click by scrolling or panning far away seems inconvenient too. Is there a more handy way to reverse the default starting point in calculating mileage? Thanks a lot.

Doug :)


postholer wrote:Postholer Google maps now has mileages for the PCT, AT, GET and JMT!!! Move your mouse over the blue trail trace and mileages will appear. Try it with the PCT:

http://postholer.com/gmap/gmap.php?trail_id=1

About the new mileage markers
At higher zoom levels with the map scale greater than 2 miles the markers are spaced a couple miles apart. As you zoom in the markers are spaced about 1/4 mile apart. This provides a great level of mileage detail at any zoom level.

This feature can be demanding on older, slower computers thus the 'Hide/Show Distances' link for the ability to turn it off or on.

Accuracy
The overall length of the calculated trail is 2,644 miles. The data used for the trace and obtaining the mileage is the USFS pct.kml. As a quality control check I've compared the calculated mileage with the PCT Data Book to see just how different they are. Overall, it's pretty impressive. The data is very close with only one big, glaringly obvious descrepancy.

Between Big Bear and Agua Dulce, the calculated mileage loses about 15 miles compared to the data book. That's alot! I haven't compared the guidebook and kml trace intensively for the error, yet. Maybe it's a frog detour, flooding or fire during mapping that is the culprit.

Here are a few comparisons of the mileage so you can get a feel of how close it really is.

Code: Select all
Segment                                  Data Book / Postholer Map
Cascade Locks(I-84) to Snoqualmie (I-90)     246.7 / 247.3
Castle Crags(I-5) to Crater Lake (Hwy62)     323.9 / 326.4
Kennedy Mdw to Toulomne Mdw                  239.9 / 247.1
Mexico to Pine to Palms (Hwy 74)             151.3 / 151.3


More than you wanted to know section
Early last year I was going to do something with elevations. When working with elevations or mileages each lat/lon point needs to be in consecutive order, that is, they can't be in some mish-mashed order. Each coordinate has to follow one after the other from the beginning of the trail to the end.

Consider the clean, well organized PCT kml file with 1501 segments (clumps of consecutive coordinates) totaling 257,276 coordinates.

Or worse, the very messy, unorganized AT (Appalachian Trail) data with 4,934 segments and 297,340 coordinates. This data set was night-marish and has become my standard by which all badly organized data is compared.

Your goal is to turn all these coordinates into one big, clean, consecutive, multi-thousand mile segment. They have to fit together correctly without gaps or overlaps to get the correct mileage, otherwise what's the point?

2 options, do it by hand or create a script to do this heavy lifting. Well the former wasn't an option, period. So it sat on the shelf until I could wrap my head around it. 3 weeks ago after a couple intense weeks of trial and error, I created the script to do the task. The best part is, it works on any kml file of any length. If the trail can be rearranged into one long segment, the script can do it. The result are a set of encoded polylines and distance encoding for google maps consumption. (Pssst, the script doesn't live on the web server.)

That's pretty much it.

Have fun!
-postholer 8)
Last edited by kdtsuei on Fri Oct 23, 2015 8:48 am, edited 1 time in total.
kdtsuei
 
Posts: 2
Joined: Wed Oct 07, 2015 7:23 am

Re: Google trail maps WITH mileages!

Postby postholer » Fri Oct 23, 2015 6:14 am

kdtsuei wrote:I plan to hike the JMT SOBO in the summer of 2016. .... Is there a more handy way the reverse the default starting point in calculating mileage? Thanks a lot.

Doug :)


I realize the JMT is often hiked southbound, even more so than northbound. What I should have done and will probably do, is change this to southbound. Having both sets of mileages would be difficult, I think.

Let me put this on my todo list.

Thanks for the feedback!
-postholer
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA

Re: Google trail maps WITH mileages!

Postby kdtsuei » Fri Oct 23, 2015 8:44 am

Thanks for your positive reply. I believe more JMT hikers would benefit from your changing mileage southbound. Look forward to seeing it even though it might not benefit my planned 2016 JMT hike.

Doug


postholer wrote:
kdtsuei wrote:I plan to hike the JMT SOBO in the summer of 2016. .... Is there a more handy way the reverse the default starting point in calculating mileage? Thanks a lot.

Doug :)


I realize the JMT is often hiked southbound, even more so than northbound. What I should have done and will probably do, is change this to southbound. Having both sets of mileages would be difficult, I think.

Let me put this on my todo list.

Thanks for the feedback!
-postholer
kdtsuei
 
Posts: 2
Joined: Wed Oct 07, 2015 7:23 am

Re: Google trail maps WITH mileages!

Postby postholer » Sat Oct 31, 2015 8:45 am

The mileages on the JMT Google Map have been changed to southbound:

JMT: http://postholer.com/gmap/gmap.php?trail_id=4&dist=1

Have fun!
-postholer
postholer
Site Admin
 
Posts: 732
Joined: Mon Apr 17, 2006 11:30 am
Location: Crescent City, CA


Return to Postholer Google Maps

Who is online

Users browsing this forum: No registered users and 1 guest