163 
RADIOMETRIC CALIBRATION OF FULL-WAVEFORM 
SMALL-FOOTPRINT AIRBORNE LASER SCANNERS 
W. Wagner a ’ *, J. Hyyppä b , A. Ullrich c , H. Lehner 3 , C. Briese 3 , S. Kaasalainen b 
Christian Doppler Laboratory, Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, 
Gusshausstrasse 27-19, 1040 Wien, Austria - ww@ipf.tuwien.ac.at 
b Finnish Geodetic Institute, Geodeetinrinne 2, 02431 Masala, Finland-juha.hyyppa@fgi.fi 
c RIEGL Laser Measurement Systems GmbH, Riedenburgstraße 48, 3580 Horn, Austria - aullrich@riegl.co.at 
KEY WORDS: Laser scanning, Lidar, Calibration, Radiometry, Point Cloud, Classification 
ABSTRACT: 
Small-footprint airborne laser scanners (ALS) are lidar instruments originally developed for topographic mapping. In recent years 
ALS sensors are increasingly used also in other applications (forest mapping, building extraction, power line modelling, etc.) and 
their technical capabilities are steadily improving. While the first ALS systems only allowed determining the range from the sensor 
to the target, current ALS sensors also record the amplitude of the backscattered echoes (peak power of the received echo), or even 
the complete echo waveform. To fully utilise the potential of the echo amplitude and waveform measurements in applications, it is 
necessary to perform a radiometric calibration. The calibration process involves the definition of the physical quantities describing 
the backscattering properties of objects and the development of practical calibration techniques. These issues are currently addressed 
by an EuroSDR (http://www.eurosdr.net/) project which aims at developing ALS calibration standards. This paper reviews the 
definition of common scattering (reflectance) parameters and concludes that in the case of small-footprint airborne laser scanning, 
the cross section <j[m 2 ] and the backscattering coefficient /[m 2 m 2 ], which is defined as the cross section normalised with the cross- 
section of the beam hitting the larget, are the preferred quantities for describing the scattering properties. Hence, either a or /should 
be used in the calibration. Also, some results of converting full-waveform data acquired with the R1EGL LMS-Q560 to cross section 
data over urban and rural test sites in Austria are shown. 
1. INTRODUCTION 
Airborne laser scanners (ALS) designed for topographic 
mapping - also referred to as topographic lidar - transmit 
narrow-beam laser pulses with a high pulse repetition frequency 
and measure the round-trip time of the pulses travelling from 
the sensor to the ground and back (Wehr and Lohr, 1999). After 
converting the time measurements to range and precise 
geolocation, an irregular but dense 3D point cloud representing 
the scatterers is obtained. While first ALS systems only 
measured the round-trip time, more advanced systems also 
record the echo amplitude (peak power of the received echo, 
most commonly referred to as “intensity”) or the complete echo 
waveform (Wagner et al., 2004). In this. way, not only 
information about the 3D location of the scatterers is collected, 
but also information about the physical backscattering 
characteristics. This opens the possibility for identifying target 
classes (e.g. vegetation, asphalt, or gravel) and target properties 
(size, reflectivity and orientation of scatterers). However, the 
echo amplitude and waveform measurements depend not only 
on the backscattering properties of the targets but also on sensor 
and flight parameters such as the flying altitude, beam 
divergence, laser pulse energy, atmospheric conditions, etc. 
(Hopkinson, 2007). Therefore, amplitude and waveform 
measurements from different sensors, acquisition campaigns 
and flight strips are not directly comparable. It may even not be 
possible to compare the measurements taken within one 
individual flight strip because of topographic height variations 
and variable atmospheric conditions along the flight path. 
For segmentation and classification purposes it would in 
general be sufficient to perform a relative correction of the ALS 
amplitude and waveform measurements. Relative correction 
methods such as proposed by (Coren and Sterzai, 2006), 
(Ahokas et al., 2006) and (Hofle and Pfeifer, 2007) aim at 
reducing echo amplitude variations by correcting the 
measurements relative to some reference, e.g. relative to a 
reference range R ref . However, it is much more desirable to 
convert the echo amplitude and waveform measurements into 
physical parameters describing the backscatter properties of the 
scatterers in a quantitative way. As pointed out by (Freeman, 
1992) for the case of Synthetic Aperture Radar (SAR) imaging, 
this is because one would like to compare measurements from 
different sensors and/or flight strips, extract geophysical 
parameters from the backscatter measurements using models, 
carry out multi-temporal studies over large areas, and build up a 
database of backscatter measurements for different types of 
land cover and incidence angles. 
To convert the sensor raw data into physical parameters it is 
necessary to apply calibration procedures. In remote sensing, 
calibration normally involves monitoring of sensor functions 
(internal calibration) and correction of the measurements with 
the help of known external reference targets (external 
calibration). Because current ALS instruments do not monitor 
sensors functions crucial for the radiometric calibration of the 
measurements (e.g. laser pulse energy), their calibration has yet 
to rely solely on external reference targets. First results using 
external reference targets have been presented by (Ahokas et al., 
2006; Kaasalainen et al., 2005; Kaasalainen et al., 2008; 
* Corresponding author