I'm at a bit of a loss here. I am attempting to generate a multi-directional hillshade, and the output raster cells seem to have a 'border' around them. I cannot for the life of me work out why and what to do about it.
What am I talking about? This:
The processing history behind this dataset:
1 - downloaded 1 second STRM elevation grid from USGS
2 - as SRTM is in an angular coordinate system, I resampled to a 20 x 20m grid (bilinear) - size not a neat match, but I thought it would be ok as the end use was hillshade rather than analysis
3 - as the STRM data is integer rather than floating, I ran Focal Statistics (average over 3x3 neighbourhood) to smooth
4 - down the multidirectional hillshade function, add to Image Analysis window, run the function on my smoothed data
The grid / borders only seem to appear on cells that have differing values in the cells adjacent to them.
Any ideas on what is causing this and how I can rectify or avoid it?