Must be wearing the hills down
10-11-13, 01:51 PM
#1
Senior Member
Thread Starter
Join Date: Jul 2013
Location: West Virginia
Posts: 201
Bikes: 2012 Cinelli Xperience
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Must be wearing the hills down
My goal for this month has been to work more on my climbing which brings me to a question. Just how is elevation gain calculated? I have ridden the same route three days this week and the elevation gain on strava has went down every day. Started at 597 ft, 534 ft, and today 498 ft. The only solution I can figure is my fat ass must be flatting the hills out by riding over them. At this rate in a few years West Virginia will be as flat as Kansas and bigger than Texas. I suspect the problem is the precision, or lack there of, that my phone measures everything with, because the grade of the hills is never the same from day to day. Heck, the grade isn't even the same going up as it is going down on the same ride. I guess I'm going to have to get an actual GPS for my bike. Garmin seems to be the only real option right? When I started riding I had no idea that a significant amount of my weight loss would be from my wallet. 
10-11-13, 02:29 PM
#2
Senior Member
Join Date: Jul 2010
Location: Victoria, Canada
Posts: 1,541
Bikes: Cannondale t1, Koga-Miyata World Traveller
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
When we use GPS for timber cruising, we set the machine to require 3 satellites to give us a fix. This increases the accuracy, but we can't always get our coordinates.
The way phone/GPS seems to work all the time, I suspect they provide a fix based on only one satellite
The way phone/GPS seems to work all the time, I suspect they provide a fix based on only one satellite
10-11-13, 02:44 PM
#3
Texas Tornado
Join Date: Aug 2013
Location: Texas
Posts: 278
Bikes: '14 Specialized Roubaix SL 4 Full Ultegra 6800
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I will agree with Strava's elevation numbers being a bit wonky. I previously used Runkeeper and would see 2-3x more elevation on that app than what Strava shows. I also know of people who run Strava and a Garmin unit simultaneously and get very different readings.
10-11-13, 02:52 PM
#4
Senior Member
Join Date: Sep 2012
Location: NW Arkansas, USA
Posts: 1,031
Bikes: 2015 Giant Roam 2 Hybrid
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 91 Post(s)
Likes: 0
Liked 6 Times
in
4 Posts
It's not an uncommon problem. I've got a bare bones GPS computer and export the data to SportTracks software and the same exact ride produces various results like the OP gave. What's funny is that I make a round trip and the rise and fall never match. Wouldn't they have to if I start and stop at the same place? There is sometimes 40-100 foot difference.
10-11-13, 03:50 PM
#5
Senior Member
Join Date: Jun 2011
Location: Buffalo, NY
Posts: 2,401
Bikes: 2012 Surly LHT, 1995 GT Outpost Trail
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
So there are really 2 ways to calculate elevation.
The first is what you'll find on devices like the Garmin Edge computers: Barometric Pressure. These measure the air pressure and calculate the altitude based on that. It's not deadly accurate, but it's usually accurate enough. Changes in the weather will affect it, so typically the longer you ride, the more inaccurate it gets.
Then there's GPS-based elevation. GPS doesn't actually calculate elevation, the signal merely provides a 2D polar coordinate that tells you your latitude and longitude. This value is then interpretted by online mapping software, mixing it in with the National Geographic Survey data to calculate your elevation changes. The problem with this method is that GPS is not as accurate as it looks. Typically it'll be off by up to 20-30 feet, but most mapping software is smart enough to show you on a road when it renders the final graphic. However, most mapping software calculates your elevation before it smooths out the lines. If you ride on a road that follows a ledge, for example, the GPS location can show you as being at the bottom of that ledge in one instant, 30 feet to the right, and then up on the road again a few seconds later. The elevation calculations will say "ok he just descended 20 feet and then climbed 20 feet in the last 20 seconds", thus adding 20 feet to your elevation, and causing your numbers to fluctuate.
The bottom line is; the numbers are "noisy". The equipment required to calculate your precise elevation at all times is far too expensive to put into a cellphone so it merely relies on approximations, and you'll never get a consistent number.
The first is what you'll find on devices like the Garmin Edge computers: Barometric Pressure. These measure the air pressure and calculate the altitude based on that. It's not deadly accurate, but it's usually accurate enough. Changes in the weather will affect it, so typically the longer you ride, the more inaccurate it gets.
Then there's GPS-based elevation. GPS doesn't actually calculate elevation, the signal merely provides a 2D polar coordinate that tells you your latitude and longitude. This value is then interpretted by online mapping software, mixing it in with the National Geographic Survey data to calculate your elevation changes. The problem with this method is that GPS is not as accurate as it looks. Typically it'll be off by up to 20-30 feet, but most mapping software is smart enough to show you on a road when it renders the final graphic. However, most mapping software calculates your elevation before it smooths out the lines. If you ride on a road that follows a ledge, for example, the GPS location can show you as being at the bottom of that ledge in one instant, 30 feet to the right, and then up on the road again a few seconds later. The elevation calculations will say "ok he just descended 20 feet and then climbed 20 feet in the last 20 seconds", thus adding 20 feet to your elevation, and causing your numbers to fluctuate.
The bottom line is; the numbers are "noisy". The equipment required to calculate your precise elevation at all times is far too expensive to put into a cellphone so it merely relies on approximations, and you'll never get a consistent number.
10-11-13, 09:23 PM
#6
Ancient Clydesdale
So there are really 2 ways to calculate elevation.
The first is what you'll find on devices like the Garmin Edge computers: Barometric Pressure. These measure the air pressure and calculate the altitude based on that. It's not deadly accurate, but it's usually accurate enough. Changes in the weather will affect it, so typically the longer you ride, the more inaccurate it gets.
Then there's GPS-based elevation. GPS doesn't actually calculate elevation, the signal merely provides a 2D polar coordinate that tells you your latitude and longitude. This value is then interpretted by online mapping software, mixing it in with the National Geological Survey data to calculate your elevation changes. The problem with this method is that GPS is not as accurate as it looks. Typically it'll be off by up to 20-30 feet, but most mapping software is smart enough to show you on a road when it renders the final graphic. However, most mapping software calculates your elevation before it smooths out the lines. If you ride on a road that follows a ledge, for example, the GPS location can show you as being at the bottom of that ledge in one instant, 30 feet to the right, and then up on the road again a few seconds later. The elevation calculations will say "ok he just descended 20 feet and then climbed 20 feet in the last 20 seconds", thus adding 20 feet to your elevation, and causing your numbers to fluctuate.
The bottom line is; the numbers are "noisy". The equipment required to calculate your precise elevation at all times is far too expensive to put into a cellphone so it merely relies on approximations, and you'll never get a consistent number.
The first is what you'll find on devices like the Garmin Edge computers: Barometric Pressure. These measure the air pressure and calculate the altitude based on that. It's not deadly accurate, but it's usually accurate enough. Changes in the weather will affect it, so typically the longer you ride, the more inaccurate it gets.
Then there's GPS-based elevation. GPS doesn't actually calculate elevation, the signal merely provides a 2D polar coordinate that tells you your latitude and longitude. This value is then interpretted by online mapping software, mixing it in with the National Geological Survey data to calculate your elevation changes. The problem with this method is that GPS is not as accurate as it looks. Typically it'll be off by up to 20-30 feet, but most mapping software is smart enough to show you on a road when it renders the final graphic. However, most mapping software calculates your elevation before it smooths out the lines. If you ride on a road that follows a ledge, for example, the GPS location can show you as being at the bottom of that ledge in one instant, 30 feet to the right, and then up on the road again a few seconds later. The elevation calculations will say "ok he just descended 20 feet and then climbed 20 feet in the last 20 seconds", thus adding 20 feet to your elevation, and causing your numbers to fluctuate.
The bottom line is; the numbers are "noisy". The equipment required to calculate your precise elevation at all times is far too expensive to put into a cellphone so it merely relies on approximations, and you'll never get a consistent number.
Fixed one word of it for you.
10-11-13, 09:49 PM
#7
Senior Member
This right here. I'm a reseller for Delorme and the more sats that can be locked onto the more accurate it is. Additionally there is a "corrective" data source that is used to correct various anomalies but one would only care about that if precision was a concern.
10-12-13, 04:35 AM
#8
Full Member
Join Date: Aug 2007
Location: Florida Gulf Coast
Posts: 396
Bikes: 1967 Carlton team/ 2013 BMC GF02 Disc/ 2013 Bianchi San Jose SS commuter
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 7 Post(s)
Likes: 0
Liked 3 Times
in
2 Posts
I have used a few different apps the latest being map my ride...I dont have elevation gain im in Fl lol however the average speed and time was wrong.Data told me i averaged 23.5 mph and did it in 1hr 40 min. 30 miles.However i actually did it in 2 hours and that puts me at 15 mph and that would be where i normally ride.The calories were correct i use a calculator after each ride and it was close so just the speed and time is way off.
10-12-13, 05:48 AM
#9
Senior Member
Thread Starter
Join Date: Jul 2013
Location: West Virginia
Posts: 201
Bikes: 2012 Cinelli Xperience
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Thank you. I never even thought about the "wobble" in the GPS signal. If it were to put me 30 feet either way in some places I could be 50-60 ft lower than the road, down in the river, or 100+ft higher than the road, up on the cliffs. That makes a lot of sense now. I guess I will just try not to worry so much about numbers and just push myself to climb more and faster.
Thread
Thread Starter
Forum
Replies
Last Post
AllWeatherJeff
Long Distance Competition/Ultracycling, Randonneuring and Endurance Cycling
26
07-02-19 01:36 PM
