Infer the topobathymetric surface of a reservoir

Question

I have a DEM that just represents the water surface of a reservoir, and I would like to infer its depth geometrically, as shown below:

ideally i would like to get a 3d line of this reservoir bottom, just like that:

Does anyone know of a tool capable of doing this?

I can use QGIS, ArcGIS, Civil3D... and others

What software are you using? You can do this in AutoCAD Civil 3D. — Pointdump, May 09 '22 at 15:53
@Pointdump i'm using GIS softwares, but i can use CIVIL3D as well. What is the function? — Vinicius B. de S. Moreira, May 09 '22 at 15:55
Create a Surface from your DEM and WSE, then add the red line as a Breakline. — Pointdump, May 09 '22 at 17:39
Are you interested in numerical depth estimates or just the line geometry? Looking at your sketch, it seems that you are interested in just the line geometry. Perhaps, then, you could use the Polygon to Centerline feature in ArcGIS or similar tool. — Gary Lucas, May 11 '22 at 13:22
@GaryLucas hey! i want just some tool that give me the reservouir bottom surface, so, i want a numerical depth estimate — Vinicius B. de S. Moreira, May 12 '22 at 00:01
How do you know the reservoir continues underwater at the same slope as the sides of the hill? Do you have an old topographic map suggesting it does or are you making up data? — GBG, May 17 '22 at 18:01

score 1 · Answer 1 · answered May 13 '22 at 12:31

I have one approach to this problem, though I am not sure it would suit your needs.

Basically, you are dealing with an area of interest where the terrestrial data is known, but the bathymetric regions are unknown. To fill in the missing raster cells, I would start with a polygon for the body of water (shapefile, etc.) and then run a series of scan lines across your raster. The polygon would allow you to determine where the lines transition from known data to unknown data. I would collect sample points along each line where data was known and define them as "control points". Then I would use a curve-fitting interpolation to fill in points where the the data was unknown. I think I would use a B-Spline interpolation rather than a cubic spline or other technique since you want to avoid "overshoots" for your curve fit. Finally, you could build contours from the fitted points using conventional GIS tools.

You could probably do this using tools like the ArcGIS Python package or an equivalent. You might have to supplement them with NumPy or other numerical package to perform the curve fit. I don't have enough experience with the standard ArcGIS analysis tools to know if you could do this using the application interactively, but the idea I describe might be supported through the right combination of tools.

This approach would give you a rough estimate of the shape of the water. Just visualizing the problem it's easy to see that it would not be particularly accurate. But, considering that you have a very limited description of the lake bottom, it's probably the best you can hope for.

Infer the topobathymetric surface of a reservoir

1 Answers1