sakau2007

I have a couple different apps and mapping utilities and I am curious how much climbing I am actually doing, however the elevation changes can be wildly different from one app or service to the next. I don't have a dedicated cycling computer... just a phone I carry with me on my rides.

ShortLegCyclist

The higher end dedicated GPS cycling computers have barometric altitude measurement which will probably be the most accurate way to determine that.

For me, close enough is close enough. I have recorded rides with phone app, base level Garmin and Cateye computers all at once and noted significantly different distances and elevation changes recorded by all three, but none of them have barometric altimeters.

sakau2007

I use the iPhone 6+ with mapmyride app. When I get home from my ride, I can then view the total climb by looking at the route I took. I have no idea how accurate this is, however, whenever I ride the app does a really good job of keeping me on my route. It never has me jumping from side to side. In fact, it can even detect pretty well when I say, drift from the shoulder towards the middle of the road. So I feel like the route is very accurate, but I just have no idea if the altitude built into the maps that it uses is accurate.

Elvo

It isn't. I've found that on the same ride, GPS elevation gain can vary as much as 100% compared to the gain calculated by devices with bariometric altimeters.

A third low cost version and most accurate way would be to obtain as-built drawings of the roadways/highways you ride from your local/county/state transportation department.

sakau2007

I think I'm doing a poor job explaining how I'm getting my theoretical elevation change.

I don't mean that I am using GPS and using data from a specific ride.

I am simply mapping the route on my PC using mapmyride and given the streets I take, it calculates the elevation; It is the same every time I do it for a given route... but I just don't know how accurate mapmyride's database is.

ShortLegCyclist

He answered your question, though, not very.

MapMyRide's elevation numbers will not be as accurate as a barometric altimeter and neither will any other program's.

WNCGoater

Just curious, does not a barometric altimeter require regular calibration with known elevations and being affected by barometric variations from weather? I would think considering the accuracy of GPS that a GPS reading of elevation would be more reliable day in and day out. Not sure, just wondering.

MapMyRide, Strava and similar apps use GPS along with cellular data to track stats, I would assume that to also include elevation. I, like the OP, have been impressed with the accuracy of the route tracking on these apps. However, my experience has been also like the OP, in that the elevations recorded have varied significantly between these two apps even while running them on the same ride.

gregf83 

Yes, a barometric altimeter needs to be periodically 'synced' to actual elevation. Garmin allows you to program in a location with a known elevation that will be used to calibrate the barometric altimeter. If you go on a long ride where there is enough time for the barometric pressure to change you will get an error in the absolute value of elevation but changes in elevation will continue to be decent. That means if you climb a 300m hill it will still register an elevation gain of 300m even if the value at the top or bottom of the climb is off.

GPS can be more accurate for elevation but generally contains more 'noise' which, if not filtered out, results in abnormally high elevation gain measurements. The elevation reported by GPS bounces around a lot above and below the actual elevation. If you sit it one place and average the values over a reasonable period of time it will be very accurate.

There will always be a variation in the elevation gain reported by apps or even barometric altimeters as there is no standard way of filtering and calculating elevation gain. It's somewhat like measuring the length of coastline. The closer you look the longer the coastline.

WNCGoater

Thanks for that explanation.

RFEngineer

I have to disagree about the barometric pressure altitude sensor. I had one on my previous sigma rox computer and the altitude reading was garbage. Yes, you basically "zero" the reading when you start your ride, but I got wildly varying elevation readings throughout an hour long ride. The number was basically useless to me.

When you track your ride with mapmyride or another GPS app, it tries to get elevation through GPS, but that is also a very error prone measurement.

I believe, IF you go out on a ride, then come back and map it manually on mapmyride (which I'm pretty sure uses google data), THAT will be the most accurate elevation data. It comes straight from topographical data that doesn't change every time you ride. That's where my money is.

Alan

scott967

In the US you can get free DEM data from the USGS for use in GIS software. Most of the US has 1/3 arc-sec horizontal resolution (10m) with some areas having 1/9 arc-sec (3m). The vertical accuracy is hard to compute. It was originally digitized from contour lines on topo maps using interpolation. So you could overlay 2d line (track) data on a DEM and extract a 3d profile and then compute elevation change from that.

Getting good DEM data outside the US isn't so easy. There is SRTMv2 which has 3 arc-sec (90m) horizontal resolution, the quality of the vertical resolution varies. More recently there is an additional dataset ASTER2 which might improve accuracy. SRTM is only available for areas under the Space Shuttle orbit (+60 deg lat to -60 deg lat) so isn't available for high lat areas. I'm not sure of the spatial coverage of the ASTER data.

scott s.
.