Digital Mapping
I like using Memory Map - a great product - but since completing our route for the TGO Challenge, I have been reflecting on the ascent figures generated by Memory Map, and the huge variation between our figures (obtained by contour counting) and those generated by the software. The digital mapping invariably overestimates by 20% to 25% or more on average terrain, but where we are doing a lot of steep ascent in open country it is almost exactly right. Why?
Clearly a track on the ground, even if it does vary a bit between two contours, will be much more level in reality than the software model suggests. For example, a route along the canal path from the top of Neptune’s Staircase in Fort William to Gairlochy shows 70m of ascent. I’ve walked along there – and, like the water in the canal, the path is flat. In fact a route along the canal itself still gives 50m of ascent – and no locks!
Where bridges occur we will in walk over them, and not down into the ravine and up the other side as the mapping software assumes (a route across the Forth rail bridge gives 43m of ascent).
Another suggested cause of error is inaccurate alignment of the map and altitude data, giving ascent and descent where there is none. This misalignment can sometimes be seen in 3D mode, where lochs extend up hillsides and streams run above the bottom of their valleys.
Given this, it is clear that any route contouring along a hillside can produce ascent figures so way out as to be pretty useless. A steep up and down route over a two or three Munros might well return a more accurate figure as the terrain mapping is much more likely to reflect reality, but in many other cases, perhaps the majority, it will be wrong.
So just how useful is the software’s version of Naismith’s Rule? If you were to factor in an inaccuracy of up to 25%, then a mountain day where the ascent is correctly rendered could come as a nasty surprise!
Alan & I decided to apply a bit of common sense, and worked out the ascents for our TGO route manually. It looks a lot less tiring now!
Clearly a track on the ground, even if it does vary a bit between two contours, will be much more level in reality than the software model suggests. For example, a route along the canal path from the top of Neptune’s Staircase in Fort William to Gairlochy shows 70m of ascent. I’ve walked along there – and, like the water in the canal, the path is flat. In fact a route along the canal itself still gives 50m of ascent – and no locks!
Where bridges occur we will in walk over them, and not down into the ravine and up the other side as the mapping software assumes (a route across the Forth rail bridge gives 43m of ascent).
Another suggested cause of error is inaccurate alignment of the map and altitude data, giving ascent and descent where there is none. This misalignment can sometimes be seen in 3D mode, where lochs extend up hillsides and streams run above the bottom of their valleys.
Given this, it is clear that any route contouring along a hillside can produce ascent figures so way out as to be pretty useless. A steep up and down route over a two or three Munros might well return a more accurate figure as the terrain mapping is much more likely to reflect reality, but in many other cases, perhaps the majority, it will be wrong.
So just how useful is the software’s version of Naismith’s Rule? If you were to factor in an inaccuracy of up to 25%, then a mountain day where the ascent is correctly rendered could come as a nasty surprise!
Alan & I decided to apply a bit of common sense, and worked out the ascents for our TGO route manually. It looks a lot less tiring now!
posted by Phil at
Saturday, January 19, 2008
2 Comments:
So much for a "Digital Challenge" then!
Post a Comment
<< Home