In: Wagner W„ Sz6kely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
TERRAIN ROUGHNESS PARAMETERS FROM 
FULL-WAVEFORM AIRBORNE LIDAR DATA 
M. Hollaus 3 ’*, B. Hofle b 
a Institute of Photogrammetry & Remote Sensing, Vienna University of Technology, 
GuBhausstr. 27-29, A-1040 Vienna, Austria, mh@ipf.tuwien.ac.at 
b Department of Geography, University of Heidelberg, Berliner Str. 48, D-69120 Heidelberg, Germany, 
hoefle@uni-heidelberg.de 
KEY WORDS: Three-dimensional, LIDAR, Surface, DEM/DTM, Retrieval, Geomorphology, Roughness 
ABSTRACT: 
As an active remote sensing technique airborne laser scanning (ALS) is able to capture the topography with high precision even for 
densely forested areas. Due to the high pulse repetition frequency of up to 400 kHz a high sampling rate on the ground can be 
achieved, which allows the description of the terrain surface in decimeter scale. In this contribution two approaches to characterize 
terrain roughness are described. In the first approach the standard deviation of detrended terrain points is calculated. To achieve a 
high spatial resolution of the derived roughness layer a high terrain point density is essential, which requires especially in dense 
forested areas a very high sampling rate. In addition to the 3D position of backscattering objects, full-waveform ALS systems 
provide the width of each detectable echo, which provides information on the range distribution of scatterers within the laser 
footprint that contribute to one echo. It is therefore, an indicator for surface roughness and the slope of the target. In comparison to 
the roughness layer derived from the first approach using high point densities, the derived echo width image shows similar spatial 
patterns of terrain roughness even for moderate point densities. The results show that both the echo widths and the vertical 
distribution of terrain echoes are useful to derive reliable geometric terrain roughness layers of large areas. 
1. INTRODUCTION 
For the modeling of natural hazards e.g. avalanches (e.g. 
Margreth and Funk, 1999), rock falls (e.g. Dorren and 
Heuvelink, 2004) and floods (e.g. Govers et al., 2000), 
information about the terrain roughness is required. For these 
different natural processes different levels of detail of the 
terrain roughness from micro-level (e.g. millimeters to 
centimeters), to meso-level (e.g. decimeter to meters) to macro 
level (e.g. meter to kilometers) are required. For practical 
applications the terrain roughness is commonly estimated by 
field investigations, which are typically based on thematic 
roughness classes as for example described by Markart et al. 
(2004). Furthermore, the terrain or landscape roughness in the 
macro-level can be determined from digital terrain model 
(DTM) analyses as for example shown in Smith et al. (2004). 
Also for the meso-level investigations exist that use airborne 
laser scanning (ALS) and spectral remote sensing data for 
floodplain roughness parameterization (e.g. Straatsma and 
Baptist, 2008). 
Especially for forested areas ALS, also referred to as LiDAR, 
has been proven as the state of the art technology for the 
acquisition of topographic information. As an active remote 
sensing technique ALS is able to capture the topography with 
high precision even for densely forested areas. The transmitted 
nanosecond-long (e.g. 4 ns) laser pulses in the near-infrared 
range of wavelengths (e.g. 1.0 or 1.5 |im) have a typical beam 
divergence of 0.5 mrad, resulting in footprint diameters of 0.2 
to 0.5 m for typical flying heights above ground of 400 to 
1000 m. Due to the high pulse repetition frequency of up to 
400 kHz a high sampling rate on the ground can be achieved, 
which allows the description of the terrain surface in decimeter 
scale. 
In this contribution two approaches to characterize terrain 
roughness with geometric quantities are described. In the first 
approach the standard deviation of detrended terrain points is 
calculated. To achieve a high spatial resolution (e.g. 
1.0 x 1.0 m) of the derived roughness layer a high terrain point 
density is required. Especially in dense forested areas where the 
laser beam penetration rates can decrease to 10% to 20% (e.g. 
Hollaus et al., 2006) a very high sampling rate is, therefore, 
mandatory. In addition to the 3D position of backscattering 
objects, full-waveform (FWF) ALS systems provide (i) the 
signal amplitude characterizing the reflectance of the scanned 
surface and (ii) the width of each detectable echo, which 
provides information on the range distribution of scatterers 
within the laser footprint that contribute to one echo (Wagner et 
al., 2004). Consequently, it is assumed that the echo width is an 
indicator for roughness and the slope of the target. Thus, in the 
second approach the potential of the derived echo widths are 
analysed for terrain roughness characterization. Finally, both 
derived terrain roughness layers are compared and discussed. 
* Corresponding author. 
287