OP
OP
GuyBoden
Guru
- Location
- Warrington
No, I didn't take my GPS as I've known the route since a teenager, but the cobbles make it more interesting. It's not steep compared to the climbs with @ColinJ in Calderdale on Monday.
Accuracy of max hill gradients using route mapping software?
- Thread starter GuyBoden
- Start date
Page may contain affiliate links. Please see terms for details.
No, I didn't take my GPS as I've known the route since a teenager, but the cobbles make it more interesting. It's not steep compared to the climbs with @ColinJ in Calderdale on Monday.
Dogtrousers
Kilometre nibbler
To make matters even more complicated, RWGPS offers you a choice of base maps. And depending on which map you use, it will calculate different routes and will get different elevations.
For example, if re-create @Alex321 's favourite, Witch's Hill above then if I use Google maps I get an average gradient of 10.4% with a max of 18.3%. But if I use OSM cycle I get a gentle average of 7.6% with a max of 10.4%. Bear in mind that according to Veloviewer it has an average of 12.0% and a max of 27.9% Your OSM Cycle user is in for a bit of a shock!
So the tool is only as good as its input. Clearly google maps has better base elevation data than OSM Cycle.
(btw OSM Cycle is just OSM with cycleways and stuff shown. I get the same elevation results if I use plain OSM.)
Google Maps
OSM Cycle
For example, if re-create @Alex321 's favourite, Witch's Hill above then if I use Google maps I get an average gradient of 10.4% with a max of 18.3%. But if I use OSM cycle I get a gentle average of 7.6% with a max of 10.4%. Bear in mind that according to Veloviewer it has an average of 12.0% and a max of 27.9% Your OSM Cycle user is in for a bit of a shock!
So the tool is only as good as its input. Clearly google maps has better base elevation data than OSM Cycle.
(btw OSM Cycle is just OSM with cycleways and stuff shown. I get the same elevation results if I use plain OSM.)
Google Maps
OSM Cycle
Last edited:
Alex321
Veteran
- Location
- South Wales
I think the only way to be sure is to go out there with a protractor, spirit level, and a weighted piece of string.
The sign at the top, if you head down it, says 20%
The sign at the top, if you head down it, says 20%
Last edited:
Dogtrousers
Kilometre nibbler
Ah yes, you can't always trust the signs. This sign is just before the (uphill) steep bit of Mote Rd near Sevenoaks. I don't quibble with the 10% bit, that's probably pretty close. But the direction of the hill? Nope.
https://www.google.co.uk/maps/@51.2...G7Er1rIDn-hr5mCQ!2e0!7i16384!8i8192?entry=ttu
https://www.google.co.uk/maps/@51.2...G7Er1rIDn-hr5mCQ!2e0!7i16384!8i8192?entry=ttu
Last edited:
presta
Guru
Plotaroute puts Hard Knott west side (signposted 30%) at 26% and Rowen YHA 29%, but I'd say they're the other way around, as I've done Rowen, but had to push on the steepest parts of Hard Knott. I was pushing on the steepest bit of Rosedale Chimney too, which is signposted as 1 in 3, but Plotaroute has it at 23%. I think gradient measurements for road signs are going to be measured over the wheelbase of a vehicle, whereas Plotaroute measures over 90m.
This looked like it's going down, but it's actually 3% - 9% up.
Really? All you need is a spirit level, and a piece of packing if the floor's not level.
- Location
- Inside my skull
Garmin's dont seem to process well very short descents, Ive seen on some of those stupid segments (usually called 'speed trap') values of the likes of 433mph
I’d say the 433mph is almost certainly an artefact of someone digitally enhancing (doping) their ride, to make themselves look faster than they are.
HLaB
Marie Attoinette Fan
Or a GPS error with a stupidly short down hill segment. Looks like Strava have tidied up the one I'm thinking of, instead of 433mph I now show an average of 39.6mph with a max of 33.5mph. How does that work an average far greater than the maximum
https://www.strava.com/segments/1645861
- Location
- Inside my skull
Or a GPS error with a stupidly short down hill segment. Looks like Strava have tidied up the one I'm thinking of, instead of 433mph I now show an average of 39.6mph with a max of 33.5mph. How does that work an average far greater than the maximum
https://www.strava.com/segments/1645861
You’ll have to ask Strava !
Richard Fairhurst
Guru
- Location
- Charlbury, the Cotswolds
It’s not so much the data - pretty much everyone uses the same SRTM source data - as the processing algorithm for that data, and that varies from site to site. I use SRTM data for cycle.travel but the processing I do is quite different to (say) RWGPS.
Dogtrousers
Kilometre nibbler
Well, you know more about this stuff that I do, but in this case we have one site (RWGPS) , with a choice of two separate base data sets (Google and OSM). So we remove the "site to site" variation. The variation is between the two data sets, Google and OSM within the same site.
In an attempt to make Witch's hill the most profiled hill in Britain. In addition to Strava, Veloviewer, RWGPS (Google) and RWGPS (OSM) here's what cycle.travel makes of the hill.
Dogtrousers
Kilometre nibbler
I'd also note that even if there was a magic route planner that was always right about max gradients that would agree with a plumb line and protractor, I reckon that the road surface is worth a good few percentage points either way. And some hills just have a nastier personality than others, regardless of their max gradient. They do sneaky things like holding back a steep bit then surprising you when you think it's all over.
Richard Fairhurst
Guru
- Location
- Charlbury, the Cotswolds
Kind of.
Google don’t give you their dataset, they just expose an elevation API. You ask it “what’s the elevation on the route I’ve just planned” and it says “600m climb and max gradient of 67%”.
OSM (as in actual OpenStreetMap) doesn’t have any elevation data bar a few spot heights. It’s not in scope for the OSM project. So if you’re building a bike routeplanner with OSM data you need to bring your own elevation data and write (or borrow) some software to do the calculations. In RWGPS’s case I’m pretty sure that’s SRTM data with the open source Graphhopper software for processing, but they might have some of their own code on top.
(“OSM Cycle” is just what RWGPS calls the OpenCycleMap display map, which is Andy Allan’s independent product from his company thunderforest.com. It doesn’t do routing or elevation calculations - those are things RWGPS do themselves.)
HLaB
Marie Attoinette Fan
I wasn't expecting an answer, it was more of rhetorical question with the answer being its extremely stupid
neil_merseyside
Guru
- Location
- Wirral
Here's one that I did for the Humber Bridge ride a few years ago...
View attachment 700418
Er, no... the bridge is NOT floating on top of the river!
Are you repeating that ride this year?
ColinJ
Puzzle game procrastinator!
- Location
- Todmorden - Yorks/Lancs border
That's a coincidence... I was just thinking about it when I fired up my tablet to check back in!
It is a bit of a trek both ways by train from Todmorden but I reckon that I might be able to persuade my pal Bill to drive us out there. He enjoyed doing the Cheshire ride on his singlespeed bike and Humber Bridge is another ideal route for that. Watch the rides forum for updates!
