Geomorphons - edge effects and difficulties in geographic coordinates

[laurencehawker]

I have been using Whitebox Workflows and it's been fantastic. However I have noticed some odd behaviour with the Geomorphons tool. It could be my misunderstanding or maybe a bug.
My aim has been to calculate geomorphons for a few places using a global 30m DEM in EPSG4326. I found it impossible to get the 'flat' class, despite using a large range of flatness threshold (1-15), search distances (30-200) and the skip parameters. I also tried the residuals parameter. I then tried the r.geomorphon tool in GRASS and I managed to get a classification that looks sensible and includes the 'flat' class.
Next I tried the Grand Junction DEM included in the sample data (image attached). In the projected coordinate system, Whitebox gives similar results to GRASS GIS. However, when converting to EPSG4326, the results are very different. Whitebox also has an edge effect (see white rectangle in attached image) in both types of projections. So in summary, there appears to be 2 potential issues:

• The geomorphons tool does not behave properly with geographic coordinates
• There are edge effects in the Geomorphons tool output

Code version and parameters used in the geomorphons tool indicated in the image.
Could well be some misconception on my part, so apologies if thats the case
Thanks for your great continued great work