One of the problems with the maps in my last post is that the vertical scale of the landforms is completely lost, it is impossible to tell whether a slope is shallow or steep.

A possible way to bring this information back is to grey out the shallowest slopes which I have done progressively for slopes less than 5 degrees in the maps below:

Apropos of cycling and Footslope / Slope / Shoulder geomorphons, would it be possible to (a) add a (raster?) category for 'painfully steep slope' (say 1 in 4 or painfuller), and (b) overlay a (vector) highway network and then (c) select out a new (vector) layer formed where the two coincide ? Sort of like (exactly like?) a spatial join. This would give a list of tough (on-road) uphills (or speedy downhills).

ReplyDeleteAs a bonus, and this might be very tricky, would it also be possible to do some kind of field-based join so that the slope value attribute of the slope geomorphon is recorded ? With this, not only would there be a layer of evilly steep slopes, but the lines could be colour-coded (scaled according to range of values) to show exactly how steep and painful ?

To reflect the fact that the painfullest slopes are both steep and carry on thus for ages, perhaps another field could be added, in which is some calculation factoring (segment steepness) x (length of that line).

Not sure what such a map of steepest road slopes might be useful for (apart from people planning a hill climbing race, in which they might need to find lots of steep stretches in fairly close proximity). My ny vynsen vy diwrosa war an fordhow na vyth, ytho dhe les dhymm rag mos yn tylleryow erel, y fia mappa a'n par ma.

It would be fairly easy to select steep slopes by doing a raster to a vector polygons of the slope layer, selecting say 1:5 or steeper, and then intersecting that with the road network.

ReplyDeleteHowever as far as navigating a road network is concerned, it isn't the actual slope value that matters, but the slope along the direction of the road. Its how to work this out and join it to the road network that I am not sure about at the moment.

Good point. I missed that - the road could even be dead flat along the contour of a steep slope.

ReplyDeletehttp://grass.osgeo.org/grass70/manuals/addons/r.geomorphon.html i can't get my head around but would it be the case that if "form" parameter is used to produce the map, at the same time you could use "azimuth" parameter to get a raster of sort-of direction of each geomorphon ? Can't see how to get the gradient value for each geomorphon though. :-z May i suggest :

1) (Somehow) for each road line segment, to make its heading/azimuth known, calculate it in a field using segment's beginning and end-point.

then, ummm.....

2) (Somehow) for the slope's geomorphon, find out heading/azimuth (perhaps using "azimuth" parameter to produce a raster and then add value to a field)

3) Find out geomorphon's prevailing gradient along azimuth (???).

4) For each road segment on a slope, perhaps use trigonometry with (1) and (2) (mmm, tricky - ny wonn vy fatel yllir y wul) to allocate potential gradient from geomorphon to road. Perhaps, using appropriate calculation from that, if the geomorphon's slope azimuth is known, and the geomorphon's gradient is known (?) and the road segment's azimuth is known, where road and geomorphon intersect, assign geomorphon's gradient (via calculation) to road. To allow for overlapping road segments beyond geomorphon boundary (this is a really rough method - and that overlapping might happen often), average out the gradients from the geomorphons with which it overlaps/intersects, if possible proportionately to the percentage extent of its length to which it overlaps with each. That might need some lot of coding.

It might well be that highways data already hold gradient values (d'oh!) :-) On the other hand, if you were doing this along a completely unmapped surface, with only 2D highway data, it might be handy to know how to do this sort of thing from geomorphons . i bet there are easier ways though.

for (4) (don't quote me on this) think the actual road gradient might be just cos (difference of road azimuth from slope azimuth) * slope gradient. So a road 90° to the geomorph's azimuth will have a 0 * slope gradient (i.e. dead level), and a road aligned spot on with the geomorphon will have nearlybout 1 * its slope gradient.

DeleteThat might be obvious, apologies if so. My maths is rubbish.

That would still leave the problem of overlapping, and shortfalling, road segments viz geomorphon edges.