6 replies to this topic

[chinhngt]

Hi All,

I'm a newcomer here and eventually in GIS work also. Hope I can seek for your advice for the following problem.
We have a shapefile with point features (converted from XYZ, Z is the depth of sea bed) and another shapefile with polylines (coastlines). We would like to do the interpolation with boundaries (polylines) to get bathymetry data in regular grid format.
The problem is, the boundary may be a beach (height=0) or a wall (height=10 m for example). So the interpolation with "beach" boundary should be different from interpolation with "wall" boundary. The boundary may contains part of "beach" and part of "wall".
The Spatial Analyst Tools has Interpolation/IDW tool which can do interpolation with boundary, but it's not clear how it treats the boundary.
Do you know if there's any plugin to ArcGIS can solve this tricky interpolation problem? Or any software?

Thank you very much for your attention.

Chinh

[Hans van der Maarel]

I'm a newcomer here and eventually in GIS work also. Hope I can seek for your advice for the following problem.
We have a shapefile with point features (converted from XYZ, Z is the depth of sea bed) and another shapefile with polylines (coastlines). We would like to do the interpolation with boundaries (polylines) to get bathymetry data in regular grid format.
The problem is, the boundary may be a beach (height=0) or a wall (height=10 m for example). So the interpolation with "beach" boundary should be different from interpolation with "wall" boundary. The boundary may contains part of "beach" and part of "wall".
The Spatial Analyst Tools has Interpolation/IDW tool which can do interpolation with boundary, but it's not clear how it treats the boundary.
Do you know if there's any plugin to ArcGIS can solve this tricky interpolation problem? Or any software?

How is your data built up? Is it possible for a "beach" line to join up with a "wall" line? (essentially creating 1 point with 2 z values)

[chinhngt]

Thanks for your question. We currently have some polylines which describes the coastline of islands (no Z value). Now we have another set of data in XYZ (z=height of the wall), the walls will overlap the coastline in some parts only. I'm quite confusing now how to combine both data to use. Will spend tomorrow researching this
Thanks for any idea if you have.

[Hans van der Maarel]

Thanks for your question. We currently have some polylines which describes the coastline of islands (no Z value). Now we have another set of data in XYZ (z=height of the wall), the walls will overlap the coastline in some parts only. I'm quite confusing now how to combine both data to use. Will spend tomorrow researching this
Thanks for any idea if you have.

Well...

One thing you could try is assign a z of 0 to the coastline, and the actual z value to the walls, then have the software generate a terrain model out of that plus the bathymetry points and see what happens.

You'll probabely have some vertical bits (coastline = wall) that may cause trouble, or parts where a coastline suddenly becomes a wall. It's hard to predict what's going to happen without seeing the data, but as I'm fond of saying: "Just give it a try, the worst thing that can happen is that it works right away!"

[François Goulet]

I'd do the interpolation with the height (or elevation) of the base of the wall. If it's vertical, that's where your sea bed ends... should be (or could) near 0.

[chinhngt]

Hi,

I tried to create a 3D polyline in ArcMap, but IDW tool does not count the 3D part. The interpolation just stops at the barrier (as a "wall") ....
The problem when a coastline suddenly becomes a wall from a beach is still not counted for the time being, as I still want to see how it works first

[chinhngt]

Hi All,

I just wanna share with you the method that I used. So we have polyline as the barrier ("wall", where interpolation will stop immediately). So if we want to make a beach, we just put a set of points with Z=0 along the polyline, then IDW or Kriging (ArcGIS) will generate a gentle slope towards the barrier.
This seems not a nice solution, but at least it works for us ...