distance, gradient and a little trigonometry

5 watchers

Icelandic Trigirl Apr 2016 1:05pm, 29 Apr 2016 2,452 posts	0 Icelandic Trigirl Bear with me. I've been doing some long runs of late so my mind has been far too active. My question is: when Garmin measures distance (and when a race official measures distance), do they do it on a flat map basis or more like a tape-measure basis? What I am thinking about are the hills and whether the distance is measured along the flat or along the hypotenuse. This does not matter much most of the time, but since I have been running where there are ONLY hills, I think it matters to me. And I ran for long enough to start wondering. This is the trigonometry part. It really isn't as hard as it sounds: Point A is the starting point; Point C is the top of the hill; Point B is therefore the point of the right-angled triangle that joins A and C. So, if you go in a dead-straight line horizontally from point A; and a dead straight vertical line down from point C, they will meet at Point B. Hence the question: is the distance measured as A to B (straight line); or as A to C (the hypotenuse; slope up and actual distance covered)? It is not going to make a massive difference but it makes some difference. My long run this week was 35k; but I reckon there was about 3.5k of climb/drop. So the hypotenuse distance (if we just take an average) would be 35.175m: square route of ((35x35)+(3.5*3.5)) . I have a feeling Fetch might have said something about this in a newsletter of late but I might have dreamt that. (Did I say I've been running a lot recently? It has taken over my life. Nothing wrong with that!) Fortunately for me, next long run is Edinburgh marathon which is nice and flat
Wednesday Mouse Apr 2016 1:08pm, 29 Apr 2016 36 posts	0 Wednesday Mouse I have no answer for you, but I've often thought about the same thing.. usually when running up a hill and wondering/hoping I can cut a run a bit short because running up the hypotenuse is further than the flat distance..
Uyuni Apr 2016 1:17pm, 29 Apr 2016 446 posts	0 Uyuni For races on the road at least the course measures use calibrated bikes ridden along the road itself - see Manchester marathon thread for full details So in your case that would be C. For XC and fell races I'm not sure how it is done but those distances tend to more approximate anyway For GPS watches and the like I've no idea. Many do show ascent/descent although that isn't reckoned to be that accurate. You would need to find a really steep preferably straight hill to test it out
thing Apr 2016 1:20pm, 29 Apr 2016 95 posts	0 thing I don't think a race measurer uses GPS, but rather something derived from counting the number of times a wheel turns, so that gives ground distance, ie does account for hills. GPS gives you a bunch of points of latitude/longitude and assumes you travel on a flat surface between them, so does not account for hills. Even on that hilly course though (35k, 1750m up + 1750m down?), the "error" is around 0.5%, which is probably within the standard operating error of GPS anyway.
rf_fozzy Apr 2016 1:26pm, 29 Apr 2016 7,691 posts	0 rf_fozzy I think the GPS measures A to C - i.e. actual distance travelled. However when imported to an online thing, such as fetch or ~~strava~~ I think they plot it onto a map and convert (although I'm not 100% certain on this), but the error is usually so small, unless you're running up Mont Ventoux or something. A second however though - GPS measurements of altitude changes are awful, so I do believe that Fetch plots that onto some form of Mapping database and extrapolates the ascent/descent numbers, and I'm fairly certain that doesn't work properly - I've seen some odd numbers both in mine and other people's data. I have a feeling there's a bug to do with the starting/finishing altitude.
Dvorak Apr 2016 1:27pm, 29 Apr 2016 11,694 posts	0 Dvorak As said, races are literally measured on the ground, so this is taken into account. I've had a think about that as well (in fact immediately before this I was using the route measurement tool to work out the slope of sections I ran a couple of days ago) and decided that the amount of added distance is in most cases negligible. If it isn't negligible, a few extra metres will be the least of my worries ;-).
Icelandic Trigirl Apr 2016 1:27pm, 29 Apr 2016 2,454 posts	0 Icelandic Trigirl Thanks for your replies! Good to know on race day there won't be any extra-extra bits to run (though I did read the newsletter about how they add 1m for every km "just in case" (no fairsies) and of course you never get to run along the shortest race line as there are crowds. And thing, it is VERY important! If you'd seen the hills around here... (and it is all hills; the only flat section I have found is about 400m although I suppose I could jump the 50cm high airport security fence and run along the runway) Back home: plenty hills to choose from but also lots of nice flat runs as well.
thing Apr 2016 1:28pm, 29 Apr 2016 96 posts	0 thing Here's Fetch waffling on about GPS accuracy, btw: http://www.fetcheveryone.com/cms-30
Icelandic Trigirl Apr 2016 1:33pm, 29 Apr 2016 2,455 posts	0 Icelandic Trigirl shhhh.... thing. I also don't want reminded that GPS gets more unreliable the further north you go because there is poorer satellite coverage (I imagine it is the same at the South Pole). I ran my 35k PLUS lots of hilly bits so for sure that was at least, well, at least 35.1k
Old Croc Apr 2016 1:33pm, 29 Apr 2016 27,289 posts	0 Old Croc http://www.fetcheveryone.com/forum__58948__2__gps_garmin_accuracy has a summary of how courses are measured by a certain local measurer...............