Creating least cost path based on slope and turning radius

[geo_ellaine]

Hi! I am trying to generate least cost paths between multiple points using slope as the primary input, and where the expected output paths should meet a minimum turning radius of 100m (refer to photo below). 

 

I am currently using the Optimal Regions Connection tool to generate the paths between the points, and uses a cost raster from weighted slope and planform curvature (to achieve this turn requirement). However, I might be using the plan curvature incorrectly for this purpose as the output paths are still traversing those below 100m which we do not want as these turns are too tight and difficult to traverse.

Does anyone have an idea how to best achieve this in ArcGIS Pro?

[JamesTenbrink]

Hello Ellaine:

This is an interesting question. If possible, could you elaborate on your use case, if it can be shared?

At this point I can make these comments:

1. The weighted distance analysis tools, out-of-the-box, will not directly solve this problem. Marking input friction cells with a higher weight if they have higher planform curvature, as determined by the landscape, may in some cases help to avoid the problem of generating tight curved paths, but it won't be a general solution. Incidentally, here is some information on the new Surface Parameters tool, which includes a capability to compute round-earth-aware tangential curvature, using standard vector calculus definitions of curvature.

2. As a workaround, I recommend an approach based on direct application of the Distance Accumulation Tool, Optimal Path As Line tool, and the new Least Cost Corridor tool. In a nutshell, you should calculate paths and corridors directly between pairs of points, smooth the paths using whatever smoothing method you want, then check to see if the smoothed path stays within the corridor. If it does, you can add that path to your solution.  If it doesn't you can either widen the corridor or manually edit the path. I'll be happy to elaborate on this approach if you send me a private message, and you can post a summary to the community if you find it acceptable.

3. i think, "theoretically", the distance analysis tools could be enhanced to add a path curvature consideration. You should post the request to the spatial analyst ideas community! The spatial analyst team has done a lot of great work to add directional awareness to the modern distance analysis operations. Curvature awareness might be a natural extension of that work. 

Regards,

-Jim TenBrink

PS: I retired from esri so i'm no longer in a position to talk about spatial analyst product planning. 😁

[EBhn]

Hello Jim,

Thanks for taking time to look into this case. 

We are doing least cost path analysis for our road design and construction, where we have multiple POIs that we want to connect (thus, we are using Optimal Regions Connection tool instead of Optimal Path as Line where there is a path source and destination). 

For our weighting, our main parameters or factors for road construction are:

• Slope: our roads/paths should be constructed on areas where slope is < 40% (percent rise)
• Horizontal curve: we want our paths to make turns where the turning radius is ≥ 100m; turns less than that would be too narrow
• Vertical curve: we want out paths to vertically traverse where the radius is ≥ 500m; turns less than that would be difficult and ‘costly’ to construct

I have read about the Distance Accumulation tool, but not sure how to implement it. I may need to dig deeper onto this and see if it suits our requirement.

This will definitely be a good idea to share to the Spatial Analyst team. 🙂

[JamesTenbrink]

Hi:

You could use Optimal Regions Connection tool to establish connectivity between points, then the corridor based approach outlined above for each pair of connected points. For example, if ORC identifies a path of cost $100 between points A and B, based on slope, you could then create a corridor of, say, cost $105 between those points . The geographic area identified by the corridor would let you visualize where your options for alternative routes between points A and B are, while incurring up to a known amount of additional cost. This approach assumes that the more costly paths that you end up using wouldn't have changed the connectivity choices made by ORC. Here is some more information on creating and interpreting the corridors created by arcgis software.

regards,

-jt