Strava or Ride With GPS--Accuracy

[Cindy White]

I'm wondering if anyone can explain the difference between elevation reports from Strava and Ride with GPS. I recently got a report from Ride With GPS that says the route I rode was just over 6,000 feet of elevation. Strava sent me a report for the same ride with just under 5,000 feet of elevation. 

I'm not a tech person but this caught my attention. Is this common? What accounts for the discrepancy?

Thanks,

cindy

[eric winkler]

Try correcting the elevation in strava. Three dots on left hand navigation. Weather and your gps  can produce different results.   I cannot confirm this but the two applications may be using different base maps.  

Eric Winkler

Met...@gmail.com

413-530-9713

On Jun 8, 2022, at 9:18 AM, Cindy White <whi...@comcast.net> wrote:

I'm wondering if anyone can explain the difference between elevation reports from Strava and Ride with GPS. I recently got a report from Ride With GPS that says the route I rode was just over 6,000 feet of elevation. Strava sent me a report for the same ride with just under 5,000 feet of elevation. 

I'm not a tech person but this caught my attention. Is this common? What accounts for the discrepancy?

Thanks,

cindy

--
You are receiving this e-mail because you joined the NCC Discussion Group on our Website. By default, every new post is sent to you. This setting can be changed in your group preferences. **REPLIES ARE SENT TO THE ENTIRE GROUP**
---
You received this message because you are subscribed to the Google Groups "NCC Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ncc-discussion-g...@nohobikeclub.org.
To view this discussion on the web visit https://groups.google.com/a/nohobikeclub.org/d/msgid/ncc-discussion-group/377c624d-25e1-4e45-aaad-7fd10342c641n%40nohobikeclub.org.

[Chris Stratton]

There are a couple of differences in the ways ride tracking apps can
handle elevation.

One is to use GPS height data from your device. Another is to use
barometric height data from some cycling computer type GPSes.

And then a third is to recognize your route, and replaced the data
with map elevation data.

None of them are perfect. For example, GPS may see the small system
errors suggest a bunch of little ups and downs that didn't actually
happen. Barometric elevation can only really show change within a
ride, when the baseline air weather-type air pressure hasn't changed.
A map data may be imperfect, coming from various sources, in some
cases including whoever's recording created a segment.

Some of the things specific things you might notice:

- Ride with GPS's dataset doesn't generally know about bridges. So
pretty much every time you cross a river, it credits you with
descending, swimming your bike across, and climbing back up the other
side.

- Some strava segments have absurdly wrong elevation data, though I
think this only effects the segment stats not your ride overall -
often whats in there is a bogus starting elevation several hundred
feet below sea level (so you'd think they could automatically clean
their dataset quite simply...). There was a "cat 4 climb" on a canal
towpath trail in NJ they eventually removed after I reported it.

- If you want to see a local odditiy, open up google maps and have it
give you directions from the main bit of Park Hill Rd to the little
disconnected piece that comes out at Glendale. Switch it to cycling
mode and manually drag the route to use the rocky dirt mtb track
beside the transfer station fence. Suddenly it thinks part of the
route is 700 feet below sea level. In the past I was able to get it
to do this for some journeys on Glendale Road alone, filed a report
and it may or may not be improved, but there's something goofy in the
data they're using, and a lot of these data sets may originate in
things like government radar altimetry (needless to say, the USGS topo
map does not show a huge sinkhole on Glendale, nor does the chimpunk
loop encounter one)

(I have gotten google to fix cycling routes before - for example
filing a report when the connector was built between Westchester's
North County and South County trailways. At first there was silence,
then after a while a notification they were looking into it, and then
one that they'd added the connector)

Anyway, back on the subject of ride recordings, one can download the
GPX/TCX files from one of these sites and try it in another. I
typically record with RWGPS (since Strava started suffering absurd GPS
wander on my phone) but then upload the rides I care about to Strava,
and tend to believe its altitude claims more, especially compared to
what I see while riding.

Also been playing with osmand some lately, it is quite clunky and
non-intuitive but will mostly do live navigation for free with a gpx
file downloaded from the RWGPS website (the rwgps app won't let you
download). Catches are that it announced turns, not intersections -
so it will call out sharp curves even when there's no choice to be
made, and it can miss particularly shallow intersections such as
Russelville to Wyben on the Westfield ride (staying on Russelville
works but is cheating - found myself already in Westfield without ever
doing that last big climb!)

[Jonathan O'Keeffe]

Hi Cindy - 

Yes, this is common.  Elevation measurements can vary considerably across different devices and different services.  It's not uncommon for two people on the same ride to have total elevations that differ by 10% or more, depending on what devices they're using and how they're configured.  It's also not uncommon to plan a route on a service like RideWithGPS with a given elevation total, only to find a different total on your device after you finish the ride, and still a third measurement reported on Strava after you load the ride to Strava.

There are three main ways of getting elevation data:

1) Calculated from maps.  A mapping service like RideWithGPS knows the elevation of each point along a road, so when you plan a route it can calculate the elevation gain you will get for that route.  Note that this is entirely a "calculated" elevation total, based on the mapping data, not any measurements taken during your ride.  Note also that different services can use different sets of mapping data, thus producing different results.

2) Calculated on your device (phone or bike computer) during a ride, using the GPS signal.  The GPS in your bike computer or phone can measure position (latitude and longitude) and elevation, but typically the elevation data is less precise, and can easily be off by 20 feet or more depending on conditions.

3) Calculated on your device (phone or bike computer) during a ride, using a barometric altimeter.  Some computers have a barometric altimeter that measures air pressure and uses this to calculate elevation.  This is usually more accurate than using the GPS signal, although it is susceptible to fluctuations in air pressure over the course of your ride if the weather is changing.

In addition, anything measured on your phone or bike computer, or calculated by Strava after your load your ride data, is subject to variations based on the exact algorithms the device or web service is using.  Some algorithms use very frequent sampling, so if you go up ten feet, down ten feet, and up ten feet again over the course of a few seconds, this would register as twenty feet of climbing (because you went up two small rises, each ten feet high).  Other algorithms might sample less frequently, or smooth out the data to try to reduce random variations, so the same sequence would result in a measurement of only ten feet of elevation gain (because you ended up ten feet higher than where you started).  So there is a lot of potential for variation.

In addition to all of that, the measured elevation data itself can be inaccurate -- GPS data can be bad if there are a lot of trees or buildings around, and barometric elevation is notoriously susceptible to problems with moisture, giving inaccurate readings when rain blocks the tiny hole in the device case that lets air into the pressure sensor.

On Strava there is an option to "correct" your elevation data for your ride if you have bad data, which means throwing out the data measured on your device by either #2 or #3 above, and replacing it with the calculated data from the underlying map from #1.  This is usually a good idea if your data is really badly off.

The net result of all of this is that elevation data is notoriously difficult to measure and calculate, and you shouldn't obsess too much about relatively minor variations.  If you use a consistent source for your numbers, such as always looking at the Strava figure, you will be reasonably close to the true figure, and this will be consistent over time.

One final point is that many of the same points above apply to distance measurement also, but usually to a lesser degree. So when your friend's mileage comes in at 60.5 when you did the same ride at 61, you usually don't need to worry about it, and should feel comfortable saying that you both rode "about 60 miles".

Have fun on the road and enjoy accumulating your elevation gain!

- Jonathan

On Wed, Jun 8, 2022 at 9:18 AM Cindy White <whi...@comcast.net> wrote:

--