Comparisons of five indices for estimating local terrain surface roughness using LiDAR point clouds

TL;DR

This study compares five common terrain surface roughness indices derived from LiDAR data across different terrains, highlighting their similarities and differences to improve roughness assessment accuracy.

Contribution

It provides a comparative analysis of five roughness indices using LiDAR-derived DEMs across various terrain complexities, emphasizing the importance of multiple indices.

Findings

Global patterns of roughness maps are similar across indices.

Local pattern distinctions vary significantly among indices.

Considering multiple indices enhances roughness analysis accuracy.

Abstract

Terrain surface roughness is an abstract concept, and its quantitative description is often vague. As such, there are various roughness indices used in the literature, the selection of which is often challenging in applications. This study compared the terrain surface roughness maps quantified by five commonly used roughness indices, and explored their correlations for four terrain surfaces of distinct surface complexities. These surfaces were represented by digital elevation models (DEMs) constructed using airborne LiDAR (Light Detection and Ranging) data. The results of this study reveal the similarity in the global patterns of the local surface roughness maps derived, and the distinctions in their local patterns. The latter suggests the importance of considering multiple indices in the studies where local roughness values are the critical inputs to subsequent analyses.