I have partial bathymetric / topographic coverage at 1m resolution of a lake bed and shore and am trying to find an appropriate method for 'filling in the gaps'. I have tried using the following in ArcGIS:
- IDW: weight = 1, number of points = 12, search radius = variable
I chose IDW weight of 1 because I understood that this would do a straight line interpolation between known points. However it seems that there is some form of distance limit (even though I chose variable radius, which means the end result is more like a Thiessen polygon (i.e. a step between horizontal extrapolations from known points)
I then tried a spline with following inputs:
- Spline: type = regularized, weight = 0.1, number of points = 12.
This produced a similar break around the halfway point between the known data, suggesting some form of distance limit, but also introduced 'corrugations' on the shore side, which I assume are due to undulations at the lake shore being extrapolated and magnified.
Views are shown in the images below overlain with the original point data from combined topo/bathymetry (done in QGIS so I could use the Profile tool).
Options I have considered are:
- fiddling around with spline parameters to see if changing these would produce a more realistic slope
- Set a barrier just behind the lake shore and use the 'Spline with barriers' tool (perhaps with more points) to force the tool to interpolate across the full extent
- Try and use the interpolation tools in QGIS (I previously got error message about k values when using IDW so used ArcGIS)
- Use some publicly available data (e.g. ASTER / SRTM) to try and get some intermediate points and redo using IDW or spline
I don't know what the shape of the profile should be but would assume than a straight line interpolation or maybe an S-curve would be more realistic than a steep cliff which both methods are producing at the moment.
Just wondering if anyone out there has faced a similar problem and what worked for them?