r.los

[Jianping Xu]

> >> Does anybody have ideas about the following : we all know
> >> that r.los can show you the area that is visible from a certain
> >> observation point, but can it also show you from what area you can observe
> >> something (I am told that this is not necessarily the same)?

> Jianping,
> 
> I think he's talking about the "one-way visibility" phenomenon that occurs
> due to geometric interpretation of cells and the algorithms used to determine
> los, as you demonstrated in your paper this year at the GRASS conference.

Thank you Bill. So my comment on Philip's message is if you run r.los
using something as observation point, then there are some area 
will not be able to see that something even they are "seen" from it.
The percentage of the "bad" area may vary depending on surface condition,
among other things.
 
> Have you contributed your los program using different alogrithms yet?  Also,
> have you been able to define under what conditions (resolution vs. height)
> the one-way visibility is more pronounced?
> 
> Bill
> 

Not yet. I have a quite tight schedule, and probably do it late this summer.

Jianping Xu