I have been trying to greatly reduce or eliminate the contour-like texture that appears in my output from gdaldem slope
. The image, below, demonstrates this textural effect.
The texture of the slope values appears to follow the contours, making a terraced effect. I am curious, does this output reflect a reasonable calculation of slope (e.g. reasonable isoclines)? Or are these patterns an artifact?
Is it possible that I will not be able to eliminate this effect because this texture is present in the original DEM (image, below)?
I have tried a number of suggestions for reducing patterns in the slope values, for instance:
- warping (reprojecting) the DEM so that the vertical units match the horizontal units (as suggested in Slope analysis from NED data gives crazy histogram?)
- warping (reprojecting) the DEM, as above, but using different resampling methods (as suggested in What is the source of horizontal and vertical striping in USGS DEMs?)
- using various "ratio of vertical units to horizontal" values in
gdaldem slope
(e.g. 111120 or 370400, as suggested in Getting incorrect slope values from an ASTER DEM in QGIS?)
Also, I have seen this texture in many other slope rasters, see for example:
- aldo_tapia's Slope results in Hillshade shows grid texture artifacts)
- This link to a blog about Curvature
Any ideas on what causes this?