Original User: r.sullivan
I've encountered this issue in the past but I've been able to work around it, however for this data set I don't have that luxury.
If I interpolate a line (either by drawing a line using the 3D Analyst Tool or by Adjusting the 3D Z value of polyline) that runs over a gap in my raster, the resulting Profile Graph is severely skewed.
What I've found is that if I click on the Interpolate Line Tool on the 3D Analyst toolbar and draw a line that crosses a gap in my elevation model, I get a graph with distances and elevations all out of sync. (See attached Image). If I, however, go to the properties of that graph and set the X Field to Ascending (under the Series Tab), the graph straightens itself out. However again, the distance over the gap becomes equal to the distance prior to the gap. In other words, if my line runs 1000m before it hits the dropped pixel 3m wide, when the graph picks up again it interpolates the distance over that dropped pixel as being 1000m. In the second attached image note the straight slope running between 1200m and 2400m. That missing segment should only be 50m long not 1200. The overall length of that profile is elongated by 1,150m.
I've also found that when interpolating a line using the Line Tool, if I go from North to South rather than from South to North, the jumbling problem goes away (changing Ascending to Descending in the Properties has no affect). The distance skewing problem also seems very inconsistent. Sometimes closing and reloading ArcMap makes it go away. The same profile line that gave wacky distances before closing ArcMap will give accurate results after reopening. But again, this is not a consistent solution. The moral is that I can sometimes get very usable data out of this tool even with the gaps... sometimes.
I'm working with a series of LiDAR coverages that are not complete. There are missing sections, pixel dropouts, whole regions of bad data. But since I'm interested in the overall trend of the elevation along very specific lines, interpolating across the gaps should be fine. It's just that ArcMap doesn't seem to always want to do it.
I'm curious if anyone else has encountered this problem and if anyone knows why it is a problem and what can be done about it. As I said, I need to generate profiles along very specific lines to show an overall trend in topography, so interpolating across a gap 3m or 50m wide in the middle of my line is not a concern.
-Richard
