# How should I correctly exaggerate a hillshade raster (ArcGIS10, Spatial Analyst)?

I once wanted to generalize a high resolution raster and instead of MEAN i accidentally selected the default SUM technique, using a cell factor 5 (Lidar DEM is 1m resolution). When I used this result as input to the hillshade funciton what I got was what I can only describe a highly exaggerated hillshade. I deleted the results and recreated the generalized DEM but now I have the need to show more exaggerated terrain characteristics in a topographic map and it seems to produce the results I want albeit generalized. Is this a common way to do this or should I be using a different technique? How would I exaggerate a DEM without having to aggregate the raster?

Hillshade created from generalized (Mean, Cell factor 5) LiDAR DEM

Hillshade created from generalized (SUMm Cell factor 5) LiDAR DEM

-

## 1 Answer

Since hillshade is just an overlay, you can do it anyway that gives you the desired result. However, the usual way to emphasise the terrain is to apply a vertical exaggeration. In ArcGIS you use the 'Z-factor' in the hillshade tool to do this as per the documentation here. You multiply the z conversion factor by the exageration factor. If your z-units are the same as your distance units then it will default to 1 and you can then treat it just as an exageration factor.

-
Thanks. That works pretty well. I have to use a high exaggeration factor (10 or more) to get similar results. Thanks for pointing it out, I overlooked that feature. I have to say though the the SUM technique yields a more dramatic result but perhaps too dramatic... – Jakub May 1 '12 at 15:40
Jakub, Summing with a 5 by 5 square neighborhood is algebraically the same as multiplying the mean by 25, that's all. The Z-factor does exactly the same thing: it multiplies all values in the DEM before computing the hillshade. One implication is that the effect of using a sum instead of a mean depends on the neighborhood size. – whuber May 1 '12 at 16:15