I would like to ask if this workflow for obtaining xyz points from raster data is acceptable.
1. Obtain SRTM data from USGS (usually 30m res.)
2. Define or Project to proper coordinate reference system and Clip to reduce the data.
3. Use the tool raster to point to create points on each center of pixels. Point features created will inherit the z value of each pixel.
4. Use the tool Export feature attributes to ASCII to export the points in csv format to have the x, y, and z.
I would then use the ASCII file containing x, y and z columns in other software like Bentley InRoads to create the surface and do the triangulation in Bentley InRoads.
Any comments on the workflow will be much appreciated.
Thanks Dan. It is because that is my main work / software at work for road design. It is just that the surface / profile created by this method is rough terrain. Alot of small undulations compared to lidar data or the one from the EU-DEM. For Hydrology study, this is not good terrain. I think i am missing something.
Not really IMHO, spline perhaps http://pro.arcgis.com/en/pro-app/tool-reference/3d-analyst/comparing-interpolation-methods.htm
but you still have the input data causing the issue. Smoothing can be done after the fact using things like focal mean to take the edge of highs and lows.
Some interpolators are better than others for reducing the impact of real or artificial data 'anomolies'. If you are smoothing for aesthetics then that is a separate issue from smoothing for things like analysis for hydrologically 'correct' data