1

I am working on creating a cost catchment based on Tobler's equation in ArcGIS10.1.

I feel confident in my cost surface; it produces a raster where each cell equals the time in hours that it would take to cross the cell dependent on slope.

Next I used the Cost Distance tool to create a raster of the accumulated time it would take to reach each cell from a source location in the hope of then creating a contour map of 1hr hiking times from the source location.

However, my cost distance outcome values are significantly greater than they should be. A point I know should only take an hour to get to produces a cost distance value of 30.

Am I using this tool incorrectly or interperating the results incorrectly?

  • 1
    Have you manually calculated the travel time to a nearby point (e.g. add the values to get across three neighboring cells), and the Cost Distance tool is giving a different sum? (This will help determine whether the problem is the cost surface values, or how the Cost Distance tool is interpreting it.) Separate of that, I'm curious how you are dealing with slope directionality (e.g. it's faster to go down a slope than up a slope); is angle of approach accounted for in Tobler's equation? This is simply academic interest, though, and won't help you troubleshoot :-) – Erica May 2 '14 at 14:39
  • Yes I have summed up all the cell values to test and got a different answer from the cost distance tool. With regard to directionality that is really dealt with, I tried using the ToblerAway table in the Path Distance tool as described on mapaspects.org/node/3744 but that also hasn't worked – AHunt May 2 '14 at 18:03
  • I admit I'm stumped. Perhaps a unit/resolution error? (if the Cost Distance tool thinks the cells are a different size, it could "scale" the resulting travel time) – Erica May 3 '14 at 2:50
  • this is not how cost distance works. pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/… – If you do not know- just GIS Jan 17 '16 at 1:12
1

Your Cost Surface could be the problem. The value of the Cost surface cell represents the cost for each map unit, not for the whole cell. The cost-values are multiplied with the resolution when using the cost-distance function.

1

I came up with the table for Tobler values posted on MapAspects in 2009 (in ArcGIS version 9.3) and after a few hours struggling with Path Distance in Arcmap 10.3 my conclusion is it's broken in the current version.

Even their example table on the Esri help file produces the same transformation as a bunch of 0 values. I've tried various values and while the built in COS and LINEAR seem to work I can't get it to accept any format of VF Table.

Consider using the original GRID command line PathDistance or use the new ArcGIS Pro. I just checked and ToblerAway.txt works as expected in ArcGIS Pro 1.0.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.