Identifying downhill cells for dispersing frog?

Question

1. I'm working on a movement/cost analysis for frogs dispersing from 135 wetlands located in Maine's mountains.
2. I have a lot of wetlands, so I'd like to automate this task in ModelBuilder using an iterator (I'm using ArcGIS 10).
3. The cell values (i.e., cost values) of my cost surface are based on the terrain ruggedness index (TRI); cells that correspond with rugged areas (e.g., mountainside) have greater cell values than cells that correspond with flat areas (e.g., plateau).
4. Using the Cost Distance tool, I will delineate an "area of influence" around each wetland; this will be based on the cost surface and species-specific dispersal distances.
5. Finally, I will estimate covariate values (e.g., cumulative stream length [NHD] and wetland area [NWI]) within the "area of influence" around each wetland.

What is the best way to identify those cells that correspond with downhill movements for frogs?

Imagine that a frog leaves a wetland; it will encounter "ups" and "downs" as it disperses. Cells corresponding with "ups" (i.e., elevation gains) should retain their respective TRI value, while cells corresponding with "downs" (i.e., elevation losses) should be reclassified as a 1 (i.e., low cost) because frogs likely have little trouble moving downhill. It seems to me that this analysis should somehow consider the location of each wetland and the direction of the frog's movement because an elevation gain from one direction may be an elevation loss from another direction.

Answer 1

You'll want to use a vertical factor with the Path Distance tool. This incorporates the "cost" of traveling up or down slope, and I've used it to (attempt to) automatically find ideal hiking trail routes.

I think that the required input data is just a DEM (surface) raster, rather than a slope or aspect raster.

Image is from the ArcGIS help page "Understanding path distance analysis"