Distances based on flow accumation path will probably not be useful

[dschwilk]

There may be no errors in the calculations, but it appears the flow accumulation based distances will not be very useful. For example, ldist_valley 2 shows sharp value transitions at ridges because of different distances to valley bottoms I assume. ldist_valley does not show this issue.

I opened this issue to propose eliminating these variables from the analyses. But please disagree if I am missing something:

ldist_valley2 ldist_ridge2

The ridge definition has other problems, so maybe we need to solve those first before abandoning this variable.

[hpoulos]

That is fine. I did some more literature searching and the methods outlined in the workflow are the generally-used methods for defining valley and ridge top pixels. I.E. using flow accumulation and flow direction grids, etc. Yes. Remove them.

[dschwilk]

Is the ridge top issue solved? Let's not worry about this issue until we solve #29.

[dschwilk]

I'm hoping these are useful once the ridge definition is fixed. Go ahead and post what these look like with the new ridge definition.

[hpoulos]

will do

[dschwilk]

These still exist in the topo_grids data but ldist_valley2 still has sharp transitions in values. I don't really understand why but I suppose we can just delete ldist_valley2 and ldist_ridge2?

[hpoulos]

I noticed this from the first generation of these grids, which is why I calculated them both ways. They are redundant, so yes, we can delete them.

[dschwilk]

Ok, We will delete them. Thanks