Extract elevation of ridge crest of a mountain range

[DavidWampler]

I'm trying to find the best way to do this.

I started by just placing points 1km apart on the highest point of the DEM for the mountain range, using an aspect map to easily see where the slope shifts. It worked pretty well for much of the range since the highest point is easy to determine. My plans were to then link up these points into a series of 1km polylines to analyze.

It wasn't a perfect solution, but because much of the mountain range is very simple, it was workable. However, part of the mountain range has no clear single crest, and instead has a series of peaks. How to "draw" the crest between peaks isn't obvious.

What I would like to do is draw a rough line parallel to the general trend of the mountain range, and then grab the highest DEM value orthogonal to each point of the line. This would easily give me something I would can feel confident calling the "ridge crest" of the mountain range, with less opportunity for human error than my current plan.

Is there a way to do this? A tool to look orthogonally from a line I draw, grab the highest value from a DEM in that direction, and attribute it to the line (or a new layer)? I haven't found any direct or indirect ways to do this from my searches.

Thanks a ton!

[RobertBorchert]

You could convert the dem to points.  it would give the points the value of each raster cell.  the end result would be the highest number indicates the highest point and so on. 


You could also play with the colors for the DEM.

If you open the layer properties and go to the symbology tab.

For example I have a DEM of the Black Hills near Custer.  The low end is 4095 and the high is 6031

select Classified and give it say 20 classifications.   It will break it so that each group contains roughly 5% of the total number of values.  In the Black Hills DEM the top group has a range of 293 feet.  The bottom one has 163  another in the middle has only 75 feet of elevation change.

If you hit the classify button you can then lengthen or narrow the range that would represent the ridge.  You can give it  distinct red color to really make it pop out. 

You then draw your line along the ridge.

I'm trying to find the best way to do this.

I started by just placing points 1km apart on the highest point of the DEM for the mountain range, using an aspect map to easily see where the slope shifts. It worked pretty well for much of the range since the highest point is easy to determine. My plans were to then link up these points into a series of 1km polylines to analyze.

It wasn't a perfect solution, but because much of the mountain range is very simple, it was workable. However, part of the mountain range has no clear single crest, and instead has a series of peaks. How to "draw" the crest between peaks isn't obvious.

What I would like to do is draw a rough line parallel to the general trend of the mountain range, and then grab the highest DEM value orthogonal to each point of the line. This would easily give me something I would can feel confident calling the "ridge crest" of the mountain range, with less opportunity for human error than my current plan.

Is there a way to do this? A tool to look orthogonally from a line I draw, grab the highest value from a DEM in that direction, and attribute it to the line (or a new layer)? I haven't found any direct or indirect ways to do this from my searches.

Thanks a ton!

[DavidWampler]

Thanks for the response, but this does not particularly help me - the range crest varies by enough that in order to use a color ramp to isolate the ridge crest I would have to be constantly adjusting it every few dozen meters in some cases. Which is exactly the sort of time-consuming, prone-to-error approach I'm trying to avoid.

Converting to points also doesn't help - although it may open up some analysis options. I'm looking for a way to automatically extract the highest points in elevation of a DEM orthogonal to an arbitrary line I draw. If I create a massive array of points, the only way I could do this (that I can think of) would be to manually delete points until I am left with a set of points that represent the highest orthogonal value. Which seems extremely time-consuming too.

In plain English, the tool I am looking for would do something like this:

1) Take input of a polyline/line
2) Start at one end of the line
3) Look orthogonally to it in both directions - find the highest value from the DEM layer
4) Assign this value to a point/cell on this line
5) Move one cell along the line and repeat

This would easily find the elevation of the ridge crest, which is all I really need. I'm not sure if the capacity to do this exists in ArcMAP though.

[AliciaMein]

Would they hydrology tool set help?  If you find the divided (watershed boundaries) in the dataset wouldn't that be the ridge you are looking for?
Otherwise, I have to think longer. 
Please also look at my question for today. Please!
Thanks

[DavidWampler]

Would they hydrology tool set help?  If you find the divided (watershed boundaries) in the dataset wouldn't that be the ridge you are looking for?
Otherwise, I have to think longer. 
Please also look at my question for today. Please!
Thanks

Unfortunately, due to the nature of the DEM I have, establishing pour points would be an extensive process. This was actually the first thing I tried to do : ).

Sorry, I can't help you with your question. I'll think about it some, but it's not something I have done before.

[CarrieDavis]

Hi,

You may be interested in checking out the following Knowledge Based Article.

http://support.esri.com/en/knowledgebase/techarticles/detail/39093