Total Climb or % - how do you calculate it?

[summerdays]

At the moment I'm having a hilly commute, and I would like to know how do I calculate either/both the total climb and the % incline of a particular hill.

I have a memory map track log of the route - can I do it within there? or is there a web-site on line that I can put in the route to find out?

For example the hill I would like to know the % of is:
200 feet in .5 miles - to me its really really steep, but I don't know if that's a mere bump to many of you here.

 

[garrilla]

some of the sites, such as mapmyride.com, give you elevation data for free. Does memory map?

we can work out the gradient using elevation / horizontal distance - but we don't have horizontal distance (yet) only the distance travelled up the incline


we can get the horizontal distance with pythagoras' theorem

a[2] + b[2] = c[2] [square brackets idicates squared]

where a = vertical, b = horizontal, c=slope

we know

a[2] = elevation = 200[2] = 40000

and c[2] = distance travelled up the incline (or the hypotenuse) = 2640[2] = 6969600

so

b[2] = 6969600 - 40000 = 6929600

and b = square root of 6929600 = 2,632

thus

the gradient is 200 / 2632 = 0.76 = 7.6%

 

[Moose]

Approx 7.58%

 

[RedBike]

Most online mapping programs under calculate the gradient / ascent. This is particularly evident on longer routes that involve short sharp climbs.

Have a try at Bike Hike.
http://www.bikehike.co.uk/mapview.php

Plot the route then somewhere on the right hand side is a little check box labelled elevation. Put a tick in this and it will list all the figures you need.

I normally go on ft climbed when comparing one hill against another.

 

[fossyant]

I think some of the mapping programmes under estimate a bit, especially when you have a lumpy ride - something to do with plot points - OK on long climbs, but then you are going up and down in short succession then I believe it can underestimate.

Can someone confirm ? I use mapmyride.

 

[Greenbank]

Can someone confirm ?

It depends on the terrain (and the program/data).

From experience Bikely underestimates by 25% in rolling terrain. In London it can overestimate wildly.

In very hilly terrain, especially if the road passes along the side of a valley or along a gorge with steep sides, it can overestimate wildly.

Various programs use different DEM data and models, none of which will be accurate as, by definition, they're a model of the terrain, not a perfect representation of it.

GPS plots are quite accurate, but you need to be careful to remove the obvious calibration problems, flutter due to poor signals, etc.

Finally, there's the definition of "climb". My commute follows the Thames and, bar a couple of bridges over it, rarely goes 5m above sea level and with no hills at all. Yet bikely, mapmyride, and my GPS all claim over 50m of climbing.

There's a 10m "climb" in Richmond Park (just past Ham Gate going anti-clockwise) that, if you get an uninterrupted run at, hardly takes any effort to get up as your momentum does most of the job. Further round there's a whopping 4m "climb" between Kingston and Roehampton Gates that is much more work since you can't carry enough momentum into it to clear it. The latter feels like more climbing, but is much less height wise.

Don't get hung up on climbing, unless it's something big in which case you can just look at the height at the bottom, and at the top on an OS map.

Rides are one of: pan-flat, flat, easy, undulating, rolling, hilly, challenging, epic or brutal.

These often change over time, despite the topography not changing one bit. Stuff that I found challenging two years ago now barely registers as undulating.

 

[2Loose]

That's pretty steep (to me anyways) Summerdays.
I have one bit that is just under 11% (thanks for the maths and explanations guys), fortunately under 300m but it burns me!

 

[potsy]

I think some of the mapping programmes under estimate a bit, especially when you have a lumpy ride - something to do with plot points - OK on long climbs, but then you are going up and down in short succession then I believe it can underestimate.

Can someone confirm ? I use mapmyride.

Fossy what % would you say the hill is through Brinnington that's the 1 I'm struggling with at the mo?

 

[wesa]

I use MemoryMap and Bikely, I agree with Greenabank on the 25% under estimate from bikely - this is compared to the figure from a Garmin 305.

You can import your route from MemoryMap to bikely (via .gpx) but last time I tried this Bikely refused to give me the elevation profile, I don't know if there are any options to play with, I just re-drew the route.

 

[jimboalee]

some of the sites, such as mapmyride.com, give you elevation data for free. Does memory map?

we can work out the gradient using elevation / horizontal distance - but we don't have horizontal distance (yet) only the distance travelled up the incline


we can get the horizontal distance with pythagoras' theorem

a[2] + b[2] = c[2] [square brackets idicates squared]

where a = vertical, b = horizontal, c=slope

we know

a[2] = elevation = 200[2] = 40000

and c[2] = distance travelled up the incline (or the hypotenuse) = 2640[2] = 6969600

so

b[2] = 6969600 - 40000 = 6929600

and b = square root of 6929600 = 2,632

thus

the gradient is 200 / 2632 = 0.76 = 7.6%

It's 200/( 0.5 x 5280 ) x 100 = 7.57.

That is a 'gentle incline'.

A nice 64 rpm on a 42 x 21 ( 53" ).

simples....

 

[fossyant]

Fossy what % would you say the hill is through Brinnington that's the 1 I'm struggling with at the mo?

Ah, the one from Portwood up to Brinnington...... map my ride says the steep bits are from 9% to 13% - although the average from the roundabout at portwood to the 'top' of the hill is only about 4% over the 0.5 mile, as most of it's flat - it's the steep bit into the village that's hard.

Stockport Road, up towards Bredbury is a good 7-8% drag - that's what I'm going up in a 74" fixed (44 x 16)

 

[RedBike]

Stockport Road, up towards Bredbury is a good 7-8% drag - that's what I'm going up in a 74" fixed (44 x 16)

I'm currently running 48x18, so slightly lower. I have two 12% hills to climb on my way to work. Although it might actually be quicker to walk the steep bits I do just about get up them. It's amazing just what hills you can get up.

 

[garrilla]

It's 200/( 0.5 x 5280 ) x 100 = 7.57.

that is exactly why the mapping programmes underestimate incline

you have done elevation / distance travelled, but distance travelled is not the same as the true horizontal distance, and to calculate that you must use the theorem

The gradient is a/b. The OP only has a and c

 

[Greenbank]

that is exactly why the mapping programmes underestimate incline

Mapping programs underestimate incline because their DEM data is only a model, it isn't an exact model of the heights along the route.

The effect of using the hypotenuse distance only changes the percentage by a tiny amount.

Consider a climb that is 20km along the road (i.e. distance travelled) and climbs 2000m.

So, in your diagram, a=2, c=20

b=sqrt((2*2)+(20*20))=20.099751

100*a/b = 9.95037%
100*a/b = 10%

So the difference between the two measurements, even on a 10% hill, is only 0.05%.

So, this could account for an underestimate by a factor of 1/200th of the real climbing figure.

 

[garrilla]

Yes. You're correct. This method is weak. The gradient would have to be 60% before it got to 25% error.

I think before I write next time.


I tricked myself into believe Pythagoras would find the horizontal distance, which is need to create a correct gradient. But that's best taken from a topological map.