Something I've been playing with for a few months - using channel and valley relationships in different terrain units to infer relative elevation.
Low order --> high order = upstream --> downstream = higher elevation --> lower elevation.
While many of the terrain units could be placed on a relative scale, others are speculative. So this is partly artistic speculation rather than hard science. Still, most of the observed relationships seem to fit the model. I also creatively blurred things to soften the transitions. Pixel values are relative, not absolute.
Relative Elevation Model of Sikun Labyrinthus, Titan:
Click to view attachment
Overlay of PIA10219 (crop of T39 SAR RADAR Swath) and the derived Relative Elevation Model of Sikun Labyrinthus, Titan:
Click to view attachment
-Mike