Distance Normalization For TRI

The original definition of TRI made no account for the different distances to the 8 neighbors, in particular longer distances for the diagonals, and the differences in EW and NS spacing for arc second DEMs. As such TRI was very similar to Steepest Adjacent Neighbor slope algorithm.

MICRODEM offers several ways to normalize.

• None--this will match other computations for TRI, with minor differences for how they handle missing and edge data.
• Use the elevation difference as a directional slope derivative, and use a linear estimate of the elevation at the specific distance from the central point
  • NS distance:
  • EW distance
  • 30 m distance
• Interpolate: find the point on the diagonal that is ½√2 pixels out on the diagonal.  For UTM grids, this provides essentially the same results as the directional derivative.

Last revised 3/5/2023