How accurate is a GPS with factoring hill distance elevation?

(1) Of course GPS does NOT take elevation change into total distance

(2) It has minuscule effect on total distance - in your case less than half a mile

First, if your pace was going down on the downhills, taht kills your theory. The downhills are equally "long" in comparison to the uphills, not short. I assume it was dropping since you mention not "striding out."

Regardless, simple Pythagorean Theorem says that your 14 miles "assuming measured "flat" with your GPS) and your 2500 ft of elevation change gives you an actual distance of 14.008 miles. In other words, you ran 42 feet farther than your watch said. Considering the inaccuracy of GPS watches to begin with, especially if you're looking instant pace, I'd say the extra 42 feet is negligible.

You probably really are slowing down and speeding up considerably on hills.

I believe you are right. GPS registers a shorter distance on hills so your pace is computed to be slower. Picture yourself running up and down a 45 degree angled very short hill. The GPS measures distance on two dimensions. So you get a straight line distance from the start to finish but you don't get credit for the hypotenuse. It registers like there was no hill but just a straight line from before the hill to after the hill.

a gps measures altitude...

It is harder to run up a hill, so you will slow down if you maintain the same effort. I'm not saying the GPS is absolutely precise, especially over short distances, but slowing down by 15-20 secs per mile is not surprising.

GPS sensor inherently measure position and velocity in three dimensions. Whether your particular watch uses three dimensions in computing the displayed distance and velocity depends on the particular device. There are good reasons that they wouldn't. First that the difference is generally negligible for any typical running route. Perhaps more importantly the vertical GPS measurements are worse than horizontal measurements so including them would degrade typical performance of the device.

That said, looking to run constant pace on a hilly route seems silly, particularly for "an easy effort run". You're going to slow down going up hill at any constant effort.

I take back my earlier mathematics. My low estimate of 14.008 assumes that your hill was a gradual 14 mile long hill, or a continuous series of equally sloped up and down hills.

Since we don't know the steepness/slope of your hill, all I can say is that your course was somewhere between 42' long, and 2500' long....if your run had a literally vertical 2500 wall in it. I think we can assume your course was much closer to a continuous series of equally sized hills than it was to having a 2500' wall in it, so I would still lean toward your course being around 50-100' long.

We could figure it out more precisely if you tell us how much of the course was perfectly flat.

GSP almost always measures long compared to the standard way to measure distance courses with a properly calibrated Jones counter on a bike.

It is true that the GPS does measure elevation (I believe this started w/ the Forerunner 300 series, but not 100% sure). However, this is limited by two very important factors. GPS is not a very good way to measure altitude very accurately. This is true regardless of what the device is. For some applications (think airplanes) they put more robust sensors and more robust software in to give a more accurate indication of altitude. Making the altitude reading more robust increases cost and size and makes very little sense in running, where the absolute chances in length are going to be very small.

I think people have a very poor judement of the steepness of hills, as I somewhat frequently see people refer to running up hills of 20-45%. You are pretty much never going to see those sorts of slopes on streets. The only application where the altitude is going to make much difference at all is going to be in mountain running, where issues with reception would be at least as big a factor as the altitude. Also pace is less important as it is so variable.

Coopington wrote:

Regardless, simple Pythagorean Theorem says that your 14 miles "assuming measured "flat" with your GPS) and your 2500 ft of elevation change gives you an actual distance of 14.008 miles. In other words, you ran 42 feet farther than your watch said. Considering the inaccuracy of GPS watches to begin with, especially if you're looking instant pace, I'd say the extra 42 feet is negligible.

That math is incorrect. You are assuming a steady change of 2500 ft over the 14 miles. It is more likely there some shorter steeper changes in elevation combined with some flat running and not a constant elevation change, then the extra distance can be substantially more than 42 feet. For instance, consider a 14 mile run where there was no elevation change for the first 13.5 miles, then the last half mile of horizontal distance had the 2500 ft elevation change. Then the Pythagorean Theorem tells us that his 14 mile run would have been about 14.2. Now two tenths of a mile is not a lot but it certainly changes the pace per mile substantially. So while he may be slowing down and speeding up, if the GPS does not account for elevation change it can make its measurement off by a noticeable amount.

2500 feet elevation change over 1/2 mile isn't running, it is mountaineering. Even with your extreme example the error in horizontal distance is about 1.5%. If anything your example illustrates how little relevance elevation change has to a GPS running watch.

Run with a GPS that uses a footpod or get an Ambit for accurate hill measurement.

Citizen Runner wrote:

2500 feet elevation change over 1/2 mile isn't running, it is mountaineering. Even with your extreme example the error in horizontal distance is about 1.5%. If anything your example illustrates how little relevance elevation change has to a GPS running watch.

I think gps on my garmin is stupid. Ill do an out-and-back and it'll show me that i gained or lost 200ft on the run when I started and ended in the exact spot.

I'm in agreement that if the gps does not take slope into account it does not affect the distance much unless you are talking about some pretty serious hills.

Also, those that said that altitude measurements by a gps unit are much less accurate than the lat/long estimates are correct due to the way a GPS triangulates from overhead satellites.

I believe airplanes and altimeters use barometric sensors that are coupled with ground level barometric pressure mapping to accurately gauge altitude. There is an equation that relates altitude to pressure but you need to know what the barometric pressure is at ground level to make practical use of it since the ground pressure varies with weather patterns. Some of the newer cell phones have barometric pressure sensors so it will be interesting to see if some of the GPS and run mapping apps start to make use of these to get a better altitude estimate.

I did see an app that just uses this sensor to get altitude and it claimed an accuracy of +/- 6 inches which is impressive.

I have a Garmin eTrex (?) handheld GPS that's targeted for hiking which has an integrated barometric sensor. It's several years old, so the technology is there and it definitely does better on altitude than unassisted GPS. Not sure why they don't use it in watches, but I could see counter arguments based on size, cost, water resistance, as well as the previous argument that it doesn't meaningfully improve the horizontal performance.

You aren't too bright, are you...

Government Cheese wrote:

(1) Of course GPS does NOT take elevation change into total distance

(2) It has minuscule effect on total distance - in your case less than half a mile

Citizen Runner wrote:

I have a Garmin eTrex (?) handheld GPS that's targeted for hiking which has an integrated barometric sensor. It's several years old, so the technology is there and it definitely does better on altitude than unassisted GPS. Not sure why they don't use it in watches, but I could see counter arguments based on size, cost, water resistance, as well as the previous argument that it doesn't meaningfully improve the horizontal performance.

Yeah, decently accurate until big front comes through. I have a watch on barometric pressure reader and elevation is kinda screwed if there are quickly changing conditions.

Assuming all hills even out is ridiculous, if you can assume that from "striding out" you should look into physics, maybe you can solve other complex problems.

Bro... the post is three years old. Give it up.

Clearly, only the displacement in the X and Y dimensions are considered in the calculation of total distance. Picture, if you will, and elevator. Does taking an elevator register a change in "total distance"??? I doubt it.