Smoothing Contours for Water Flow

[tpennerNXGJH]

Hi,

   We recently did an incredibly large Lidar for a project my company is doing.  I was able to get it down to a usable size for our project but I'm still having issues with the surface.  These are zoomed in so you can see exactly what I mean;

On the left is a plan view of the contours created by the lidar.  On the right is a snippet of our profile view.  The grey squiggly line is our existing ground from the lidar.   Because it's bumpy and all over the place, it's impossible to do a good water drop analysis of the area.  These faux "low points" are all over the place and it becomes absolutely useless to determine water flow.  How do I smooth out the profile view, and clean up the plan view so it looks more like actual contour lines should? Like this;

[TerryDotson]

---

You could try Dotsoft Lidar Tools. I've had good luck with them.

---

Thanks for the mention. While I'm still absorbing this overall thread, that freeware does contain LIDARMAKE which creates a LAZ file from Civil3D surfaces defined in the drawing. Surface sampler writes out points on a virtual grid (user supplies interval) like the image shown on post #2, draped onto the surface for reducing existing overly dense surfaces. 

That .LAZ file could then be used with the freeware's LIDARTIN to create a reduced Civil3D surface that is (based on my opinion of grid surfaces) going to be smoother than the source.

[tpennerNXGJH]

So much to absorb. Thanks for all the replies. Will be back to tackling this project on Monday. Will let you know how it turns out!!

Travis

[ChrisRS]

Aha! I may not be completely crazy!

[AllenJessup]

@TerryDotson I tried LIDARTHIN on one of my files. This resulted in an overall smoother Surface in Civil 3D. However there were still a few big jogs instead of many small jogs in the contours. I think this is what you're going to get from a Lidar grid with no breaklines. Not necessarily a fauld in the program. Just the way the algorithm processes that kind of data.

I never use the aerial Lidar to create a Surface that I'm going to design against. It's either to fill in outside the Surveyed are or for preliminary planning.

We use Point Clouds from Terrestrial Scanning and create points and breaklines from that. 

[TerryDotson]

---

I tried LIDARTHIN on one of my files. This resulted in an overall smoother Surface in Civil 3D. However there were still a few big jogs ...

Using LIDARTHIN has two methods:

1. The Contour method will remove points that would fall between the designated intervals.  The results actually look like contours if you plot the remaining points.
2. The Distance method (slower) removes points closer than the specified distance and creates and overall looser triangulation.  View the resulting triangles to see the results.

My last post mentioned LIDARMAKE which actually writes a .LAZ file by sampling the in-drawing surface at the user specified interval.  The triangulated results look like a plotted grid file with the squares and one diagonal in each square.

[AllenJessup]

I used the Distance method. Looks like the Contour method would work better. Thanks for the explaination.

[TerryDotson]

---

I used the Distance method. Looks like the Contour method would work better. Thanks for the explaination.

---

Unless the specs are super tight the results of that would still include small triangles that happen to lie close to the contour interval.  The results of the Distance method has loose triangles almost as if a field crew had collected points.  So maybe a looser distance method would be best?

I don't know how the Civil3D water flow tool works but I have to assume it's a follow the triangle approach.  Reaching the triangle edge then pickup and follow the neighboring triangle downhill, etc.   I wouldn't be opposed to trying to implement it, but I expect I would encounter the same problems it does in level areas and ponds, etc.

[AllenJessup]

I don't think an Improved Water Drop tool is what is needed. A totally different Flow Tool would have to take into account the features mentioned above. Most important would be the tools ability to "Fill" a depression and find the outfall. Very different from finding the maximum slope along a triangle.

The Water Drop tool might not even deal with triangles. I believe it existed in Land Desktop. A coworker who started out as an Electrical Engineer said that he thought it worked like calculating something (I don't remember what) across lines of EMF . In areas where there is no defined channel. The Water Drop path runs perpendicular to the Contours.

I used the Water Drop tool to trace the deep channel (Thalweg) of the Hudson River opposite our County. That defines the County Boundary.

[tpennerNXGJH]

I've tried everything everyone has suggested here and learned a lot.  There's a lot of great ideas.  Unfortunately nothing I've done so far has gotten this massive lidar surface to be smooth enough to use the water flow tool effectively.  I'm afraid it's too rudimentary to be useful on a large project like this.

Thanks for all the suggestions.  Keep them coming - perhaps you might know something I haven't thought of yet.

Travis

[AllenJessup]

---

@tpennerNXGJH wrote:

I'm afraid it's too rudimentary to be useful on a large project like this.

---

Yes. This is a very old and simple tool. A new routine would have to be developed to do what you want it to do. In the meantime. I'd  suggest you post this as an ideas on the Civil 3D Ideas forum.

Let us know if you do so we can vote for it.

[Neilw_05]

Perhaps this would be better handled by GIS software? I know there are more robust tools for this. See the comment by ssperales above. You might be able to do the analysis in GIS and export/import the resulting linework you need.

[Joe-Bouza]

You need to weed weed weed the contours

[tpennerNXGJH]

Yeah I basically told my engineer - if he wanted it to work, I'd have to spend a couple days getting very hands on with it.  There's just no way to automate it unfortunately.

[Joe-Bouza]

There is of course reading the contours.

I assume the engineer wants to delineate a water shed. read the contours and draw the water sheds

[tpennerNXGJH]

That's pretty much what he's decided to do after nearly 2 days of me trying to do this the "easy way"