For the reasons Leslie noted, the Garmin units like the 205 or the older etrex Legend that just use GPS to measure altitude changes work quite terribly; I have seen errors by as much as a factor of 2. On the other hand, the units (like the 705) that have the barometric altimeter seem to be roughly about as accurate as other bike computers that just have a barometric altimeter but no GPS capability.
In principle, Garmin could be using a proprietary algorithm that does data fusion on GPS measurements, topo databases, and the barometric altimeter. What they actually do, AFAIK, is not disclosed. However, my guess based on observation is the models with barometric altimeter are mostly using that. ...Roy On Tue, Feb 16, 2010 at 10:20 AM, Leslie <leslie.bri...@gmail.com> wrote: > On Feb 15, 4:32 pm, Jim Thill - Hiawatha Cyclery <thill....@gmail.com> > wrote: > > GPS should record elevation changes more accurately, and both GPS and > > cycle computers should record distance traveled more accurately than > > mapping software does. In the old days (10-15 years ago), GPS > > elevation measurements were suspect because there weren't always > > enough satellites to triangulate accurately in the 3rd dimension. I > > think this has improved lately. > > I've never used a bicycle computer, so I can't speak to comparing its > data to web-based map service, but I can touch upon the GPS elevation > issue. > > While it's true that the satellite constellation has filled out (and > is currently being rearranged from a 21+3 to a 24+3 configuration over > the next year or so), and GPS receivers are now running at least 12 > channels or more, that accuracy has improved. Unfortunately, the > vertical accuracy will always lag the horizontal, because the earth > isn't transparent. > > If you were to stand at a window with a GPS unit and look at the > 'skyplot' view, you would notice that the receiver is picking up > signals from satellites out in front of you that you can see from the > window (if you could actually 'see' them), but that the ones behind > you are blocked by the building, which thus skews your calculated > location. If you then went outside and were standing in the middle of > a field, you could see satellites all around, then the calculated > location can be more precise. > > The same thing is going on with elevations, but in that case, it's the > earth that's obscuring signals from satellites on the other side of > the earth. Because all of the satellites within view are being used > are all overhead, and not 'behind' you, there's a bit of a skew; the > smallest bit of error makes for a much larger change in elevation, > than you would get for horizontal calculations. If the satellites > that you're viewing/using are spread out around the horizon, then > you'll be better off than if they were all clustered directly overhead > - that would help minimize that error - but, that error will still be > greater than your horizontal. > > FWIW, the maps that I usually make, I will pull my horizontal location > off the GPS, plot it on a topo, then use the elevations off the map, > in order to compare to other locations (that aren't GPS'd). [Unless, I > can actually GPS the other locations myself, too, in which case I'll > then emphasize their relative elevations over absolute elevations.] > > With that said, yes, it's improving. I've got a few Trimble units > here in my office that are sub-meter, and I can pull off sub-foot with > them, given enough time (and even cm-level accuracy w/ a LOT of time > and post-processing); but, vertically, I wouldn't guarantee anything > more than sub-foot, and that's w/ a long residence time, not an > instantaneous value, as you would have if you're biking along. > > (Hope that's not totally confusing....) > > -L > > -- > You received this message because you are subscribed to the Google Groups > "RBW Owners Bunch" group. > To post to this group, send email to rbw-owners-bu...@googlegroups.com. > To unsubscribe from this group, send email to > rbw-owners-bunch+unsubscr...@googlegroups.com<rbw-owners-bunch%2bunsubscr...@googlegroups.com> > . > For more options, visit this group at > http://groups.google.com/group/rbw-owners-bunch?hl=en. > > -- You received this message because you are subscribed to the Google Groups "RBW Owners Bunch" group. To post to this group, send email to rbw-owners-bu...@googlegroups.com. To unsubscribe from this group, send email to rbw-owners-bunch+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/rbw-owners-bunch?hl=en.
