Finding Sinks with SA Hydrology Tools

[jeffnolan]

I'm looking for sinkholes in a karst area.  I thought I would take the elevation raster created from a lidar file and run the Sink tool to find all the closed depressions.  All I get with Sink are tiny sinks a few pixels in size, even though I can see large closed depressions on the contour map made from the same elevation file.   When I run Fill, it fills all the closed depressions I see in the contours, so it seems like Fill is working fine.  But Sink isn't identifying the sinks that Fill is filling.  I can get at what I need if I create a second contour map from the Fill raster, and overlay the contour maps from the filled and unfilled rasters.  With the filled contours on top, I can identify the sinkholes where the unfilled raster contours are visible.  However, I would like to understand why the Sink routine isn't identifying these sinks.

[BenjaminMittler]

Fill is an iterative process meaning once the program has filled all of the sinks it will then check for new sinks that have been created by the process of filling the old sinks and continue to fill in this fashion until all sinks are filled.

When you identify sinks it only identifies the first round of sinks which are often smaller sinks subset within larger sinks. It is for this reason that when you run identify sinks it only shows sinks that are a few pixels in size.

If you really want to see the larger sinks try running fill with a small z value to rid yourself of the smaller sinks and then run identify sinks on the new grid.

[jeffnolan]

Thanks.  That would explain it.  I ended up subtracting the original raster from the raster with the filled sinks, which yields a layer with just the sinks.  Works fine.