Loading ...
Sorry, an error occurred while loading the content.

[John Muir Trail] Re: Cumulative Elevation gain Hi to Whitney

Expand Messages
  • longritchie
    Best data? I think it is probably an underestimate. Agreed that TOPO! isn t very good at this. Likewise there are issues with using GPS position data which is
    Message 1 of 32 , Jul 10, 2013
    • 0 Attachment
      Best data? I think it is probably an underestimate.

      Agreed that TOPO! isn't very good at this. Likewise there are issues with using GPS position data which is then translated to elevation via mapping data. I think this is likely what postholer.com did to arrive at their numbers. I'm surprised nobody mentioned postholer. If you look at the values they publish for individual sections of the JMT they are very high, likely raw data that has a lot of noise. Here's a closeup of some GPS data that lies between 1000 Island Lake and Garnet, possibly data used by postholer.com:





      Those zig-zags and figure-8 maneuvers translate into noise. For the 1000 Island Lake to Garnet section this dataset produced a value of about 670 feet, the same as what appears on the postholer website. But for a total JMT value for Yosemite to Whitney, postholer.com says 45,029 feet. To arrive at this number they must have used some filtering/smoothing algorithm. The problem is figuring out how to eliminate the noise without losing actual elevation change. Did they succeed in throwing out the bathwater without any of the baby?

      Now I have a few numbers for Yos to Whitney cumulative gain  (in 1000s of feet):

      47.4 TOPO! (my trace)
      46.0 wikipedia (uncredited)
      45.0 postholer.com
      42.6 Wenk

      I also have two cases now where the Wenk data appears to be 5-10% low.

      You decide.


      --- In johnmuirtrail@yahoogroups.com, John Ladd wrote:
      >
      > Longrichie -- How nicely wonky! I was about to comment that the heat
      > probably explained the discrepancy and see you already caught that. I
      > suspect the Wilderness Press data is (while not perfect) probably the best
      > we have once we eliminate the odd error like the one involving Island Pass.
      > Note that when you traverse a steep slideslope, anything derived from a map
      > trace (as in Topo) can be way off. If it is reading just 5 yards to the
      > right of the trail at one point, and 5 yards to the left at the next point
      > and it is a 45 degree slope you would get a spurious 30 foot elevation
      > difference in addition to the true climb of the trail. And we know from
      > comparing GPS traces as walked to the traces of the same trail on a USGS
      > basemap (probably used by Topo), the map trace can be quite far off the
      > real trail (not usually a problem for actual field use, but it makes a
      > difference here). I think the truest readings would come from closely
      > spaced barometric altimeter readings and the WP data does that pretty well.
      > I suspect they may deliver a slight underestimate, but I suspect it is
      > less than other methods.
      >
      > John Curran Ladd
      > 1616 Castro Street
      > San Francisco, CA 94114-3707
      > 415-648-9279
      >
      >
      > On Wed, Jul 10, 2013 at 2:18 PM, longritchie no_reply@yahoogroups.comwrote:
      >
      > > **
      > >
      > >
      > >
      > >
      > >
      > > I borrowed an altimeter watch and walked the first 50 miles of the JMT.
      > > Here is the reported altimeter elevation compared to the Wenk data:
      > >
      > >
      > > (http://imageshack.us/a/img29/5526/0ojq.jpg)
      > >
      > > I set the elevation at Happy Isles and then set it again at Tuolumne. The
      > > altimeter read low between Yosemite Valley and Tuolumne. I suspect that
      > > this was due to the fact that it was a very hot day. High temperature means
      > > lower aid density which results in a smaller change in pressure per change
      > > in elevation. Although the pressure sensor in the altimeter is temperature
      > > compensated I don't think the firmware attempts to adjust its model of the
      > > atmosphere based on the ambient temperature. One of the pitfalls of
      > > barometric altimetry.
      > >
      > > Beyond Tuolumne the altimeter tracked the map elevation reasonably well,
      > > both on the way out and on the return trip. When returning to the
      > > sweltering hell that is Yosemite the altimeter again under-reported the
      > > elevation change.
      > >
      > > Here is the cumulative gain reported by the altimeter, compared to the
      > > Wenk data:
      > >
      > >
      > > (http://imageshack.us/a/img198/967/wz5d.jpg)
      > >
      > > The altimeter matched the Wenk data between Yosemite and Tuolumne despite
      > > reporting about 6% less net elevation change. Beyond Tuolumne the agreement
      > > is again fairly good until around mile 43 when they diverge. This is
      > > because of insufficient Wenk data points between 1000 Island Lake and
      > > Garnet Lake. The Wenk/Ravi calculation is for a total gain of 70 feet here.
      > > The altimeter reported 390 feet (both heading out and on the return trip).
      > > A trace in TOPO! suggests 330 feet.
      > >
      > > By the way, I also came across an error in the original Wenk data. She
      > > somehow inserted a second copy of "Island Pass" at the same mile point as
      > > Donahue Pass. Ravi carried this error over to his spreadsheet resulting in
      > > an additional 165 feet of gain where there is actually little or none.
      > >
      > > The cumulative gain determined by the altimeter was about 5% greater than
      > > that from Wenk. If one were to attempt to compensate for the
      > > under-reporting of elevation change between YV and TM the difference would
      > > be something like 8-10%, but this is assuming that the altimeter was
      > > otherwise accurate in determining cumulative gain. I do not know the
      > > details of its accumulation algorithm.
      > >
      > > For what it's worth.
      > >
      > >
      > >
      >
    • longritchie
      Okay, now I see. I checked the box that hides both my email AND my IP address. I was in a suspicious mood when I checked that. Why do they need my IP address,
      Message 32 of 32 , Jul 10, 2013
      • 0 Attachment
        Okay, now I see. I checked the box that hides both my email AND my IP address. I was in a suspicious mood when I checked that. Why do they need my IP address, I wondered? It doesn't appear that I can chose one or the other, it's got to be neither or both. So I went with neither.

        But feel free to email me: longritchie AT yahoo DOT com


        --- In johnmuirtrail@yahoogroups.com, Roleigh Martin <roleigh@...> wrote:
        >
        > Your postings, come across like this:
        > longritchie <no_reply@yahoogroups.com>
        >
        > It means you asked Yahoo not to display your email address. It's fine, it
        > just means people can't contact you individually, even moderators.
        >
        > I'm replying via the group even though you spelled out your email address
        > so that others understand the same implication of making that
        > choice.
        >
        > Elizabeth Wenks created the XLS except for the advertisement page which she
        > asked me to do to please the publisher and get their permission.
        >
        > She created the other two spreadsheet tabs in that file.
        > -------------------------------------------------
        > Visit my Google Profile (lots of very interesting research
        > links)<https://plus.google.com/104440166440169700478/about>
        > _
        >
        >
        >
        > On Wed, Jul 10, 2013 at 8:14 PM, longritchie <no_reply@yahoogroups.com>wrote:
        >
        > > **
        > >
        > >
        > > "Longritchie, you make it impossible for people to reach you offline
        > > since your handle is not associated with an email address."
        > >
        > > How did I make this impossible?
        > >
        > > I'm new to this group. I signed up with a yahoo email: longritchie AT
        > > yahoo DOT com.
        > > What else was I supposed to do exactly?
        > >
        > > I checked the "_Elizabeth Wenk Master Spreadsheet for JMT.xls" file
        > > which I obtained from the JMT Yahoo files section and found that she did
        > > have an errata regarding the location of Island Pass. But on the "Data"
        > > page in the spreadsheet there are in fact two seperate entries for
        > > Island Pass. And that is where Ravi got his data. His value of 42,773
        > > feet total gain from Yosemite to Whitney, based on Wenk's data, should
        > > be reduced by 165 feet to 42,608 feet. Yeah, I know, big deal.
        > >
        > > I can send an email to Ms. Wenk if it was in fact her hand that created
        > > the XLS file. Was it?
        > >
        > >
        > > --- In johnmuirtrail@yahoogroups.com, Roleigh Martin wrote:
        > > >
        > > > Longritchie, you make it impossible for people to reach you offline
        > > since
        > > > your handle is not associated with an email address. I wanted to
        > > reply
        > > > offline that Elizabeth provides contact information for error
        > > corrections
        > > > in that spreadsheet listed below:
        > > >
        > > > When you post the spreadsheet and errata to the websites, you can go
        > > ahead
        > > >
        > > > and give the following e-mail address for me, asking people to
        > > contact me
        > > > only with errors
        > > >
        > > > they have found or information they'd like included in future
        > > editions:
        > > >
        > > > wenk [dot] jmtbook [at] gmail [dot] com.
        > > >
        > > >
        > > > -------------------------------------------------
        > > > Visit my Google Profile (lots of very interesting research
        > > > links)
        > > > _
        > > >
        > > >
        > > >
        > > > On Wed, Jul 10, 2013 at 7:04 PM, Roleigh Martin roleigh@ wrote:
        > > >
        > > > > Did you look at the errata page Elizabeth Wenk provided us in the
        > > master
        > > > > spreadsheet in our file library?
        > > > >
        > > > >
        > > > >
        > > http://groups.yahoo.com/group/johnmuirtrail/files/Planning%20and%20Trans\
        > > portation/_Wilderness%20Press%20and%20Other%20Data%20Points/<http://groups.yahoo.com/group/johnmuirtrail/files/Planning%20and%20Transportation/_Wilderness%20Press%20and%20Other%20Data%20Points/>
        > >
        > > > >
        > > > > _Elizabeth Wenk Master Spreadsheet for JMT.xls
        > > > >
        > > > > If she did not correct the error, I can send you her email address,
        > > I'm
        > > > > sure she'd like to update the errata page. I'll do this offline but
        > > I
        > > > > might not be able to do it immediately as I'm leaving for my JMT
        > > hike
        > > > > tomorrow (at least the fly to CA part).
        > > > > -----------------------------------------------
        > > > > Visit my Google Profile (lots of very interesting research links)
        > > > > _
        > > > >
        > > > >
        > > > >
        > > > > On Wed, Jul 10, 2013 at 5:18 PM, longritchie
        > > no_reply@...:
        > > > >
        > > > >> **
        > >
        > > > >>
        > > > >>
        > > > >>
        > > > >>
        > > > >>
        > > > >> I borrowed an altimeter watch and walked the first 50 miles of the
        > > JMT.
        > > > >> Here is the reported altimeter elevation compared to the Wenk data:
        > > > >>
        > > > >>
        > > > >> (http://imageshack.us/a/img29/5526/0ojq.jpg)
        > > > >>
        > > > >> I set the elevation at Happy Isles and then set it again at
        > > Tuolumne. The
        > > > >> altimeter read low between Yosemite Valley and Tuolumne. I suspect
        > > that
        > > > >> this was due to the fact that it was a very hot day. High
        > > temperature means
        > > > >> lower aid density which results in a smaller change in pressure per
        > > change
        > > > >> in elevation. Although the pressure sensor in the altimeter is
        > > temperature
        > > > >> compensated I don't think the firmware attempts to adjust its model
        > > of the
        > > > >> atmosphere based on the ambient temperature. One of the pitfalls of
        > > > >> barometric altimetry.
        > > > >>
        > > > >> Beyond Tuolumne the altimeter tracked the map elevation reasonably
        > > well,
        > > > >> both on the way out and on the return trip. When returning to the
        > > > >> sweltering hell that is Yosemite the altimeter again under-reported
        > > the
        > > > >> elevation change.
        > > > >>
        > > > >> Here is the cumulative gain reported by the altimeter, compared to
        > > the
        > > > >> Wenk data:
        > > > >>
        > > > >>
        > > > >> (http://imageshack.us/a/img198/967/wz5d.jpg)
        > > > >>
        > > > >> The altimeter matched the Wenk data between Yosemite and Tuolumne
        > > despite
        > > > >> reporting about 6% less net elevation change. Beyond Tuolumne the
        > > agreement
        > > > >> is again fairly good until around mile 43 when they diverge. This
        > > is
        > > > >> because of insufficient Wenk data points between 1000 Island Lake
        > > and
        > > > >> Garnet Lake. The Wenk/Ravi calculation is for a total gain of 70
        > > feet here.
        > > > >> The altimeter reported 390 feet (both heading out and on the return
        > > trip).
        > > > >> A trace in TOPO! suggests 330 feet.
        > > > >>
        > > > >> By the way, I also came across an error in the original Wenk data.
        > > She
        > > > >> somehow inserted a second copy of "Island Pass" at the same mile
        > > point as
        > > > >> Donahue Pass. Ravi carried this error over to his spreadsheet
        > > resulting in
        > > > >> an additional 165 feet of gain where there is actually little or
        > > none.
        > > > >>
        > > > >> The cumulative gain determined by the altimeter was about 5%
        > > greater than
        > > > >> that from Wenk. If one were to attempt to compensate for the
        > > > >> under-reporting of elevation change between YV and TM the
        > > difference would
        > > > >> be something like 8-10%, but this is assuming that the altimeter
        > > was
        > > > >> otherwise accurate in determining cumulative gain. I do not know
        > > the
        > > > >> details of its accumulation algorithm.
        > > > >>
        > > > >> For what it's worth.
        > > > >>
        > > > >>
        > > > >>
        > > > >
        > > > >
        > > >
        > >
        > >
        > >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.