The design of the receiver part of the laser ranging unit is particularly important as it may determine the type of lidar data the receiver records—discrete return measurements or the full waveform� In the first case, a laser pulse may provide multiple returns depend- ing on the type of surface it intercepts� When the laser beam hits porous objects, such as the forest canopy, it may intercept foliage or tree branches over part of the laser footprint, which may backscatter enough energy to trigger the recording of the travel time by the laser receiver (Figure 3�8)� After hitting the top of the canopy, part of the laser beam may continue its travel through openings in the canopy until it again hits another layer of foli- age or branches, or possibly the ground, which may generate secondary returns of the same pulse� Depending on the complexity of the forest canopy and the settings of the laser receiver, a laser pulse may generate up to four or five discrete returns, sometimes with less dependence on the limitations imposed by the receiver�
Ideally, a laser pulse hitting the forest canopy would provide a return from the top of the canopy—the first return in Figure 3�8—and it would still be able to penetrate to the ground and record a last return from the forest floor—the third return in Figure 3�8� Such measurements allow us to accurately characterize vegetation height and the terrain eleva- tion under the canopy� Some of the laser pulses intercepting the canopy may provide only one return when foliage, branches, or tree trunks block the entire footprint or when these pulses hit the bare ground without intercepting tall layers of vegetation� Similarly, when the laser footprint covers completely nonporous objects, such as roofs, sides of buildings, or other human-made structures, the laser pulse will provide only one return� Therefore, discrete-returns lidar data include first returns, intermediate returns, and last returns�
Most discrete-returns lidar sensors use constant fraction discriminators (CFDs) to mini- mize the “range walk” or systematic variation in range with signal level� Backscattered laser signals have varying amplitudes depending on the initial pulse energy, size of the intercepted object, and target reflectance characteristics� In order to handle such varia- tions, most laser receivers use a constant amplitude ratio to identify a laser return and record its travel time and power, denoted as amplitude or intensity� The CFD is used to define the leading edge of the pulse (Figure 3�9), which, as explained in Section 3�3, is not well defined but generally considered to be a fraction of the signal peak to avoid issues caused by various pulse amplitudes (Baltsavias 1999)�
The CFD-based receivers need to reset their detectors to prepare for the next pulse or the next echo returned from the same pulse; therefore, there is a time separation between returns recorded for the same echo� Although the reset time, sometimes referred to as “nominal dead time,” varies with sensors and manufacturers, it is most commonly around 8–10 nano- seconds� This reset time translates to a range separation of 1�2–1�5 m between the recorded returns of the same pulse, when considering to- and from-target travel times�
The reset time and the minimum range separation between multiple returns have implications for detecting ground covered by vegetation� When the ground is covered by tall grasses or shrubs with heights less than 1�2–1�5 m, the laser beam may provide a return from the top of the vegetation cover and may penetrate to the ground and gener- ate a secondary ground return� This ground return may not be detected due to the fact
Discrete returns Full waveform Path of laser beam
First return
Second/intermediate return
Third/last return Return not recorded
0 5 m 6.3 m 12 m
Time delay Height (ns)
(m)
% Intensity
0 100
0 ns
40 ns 19 ns
23.3 ns
FIgure 3.8
Laser beam interaction with vegetation and variation of the backscattered laser signal�
that no returns are recorded by the receiver within the reset time� This situation does not mean that characterizing ground elevation is always biased toward higher elevation val- ues when there is a low layer of vegetation� A significant number of laser pulses will pen- etrate to the ground, and laser point classification algorithms will identify lower pulses that most likely hit the ground and use them to generate digital elevation models�
Some airborne lidar sensors manufactured during the mid-1990s (e�g�, Optech ALTM 1020, Optech, Inc�, Vaughan, Ontario, Canada) could be toggled to record either the first or the last return, and two flights over the same area were necessary to get the bare ground terrain model and the top of the canopy surface, when flown over forest vegetation� Surveys in the U�S� Pacific Northwest carried out using the Optech ALTM 1020 scanning system indicated a minimum 2030% penetration of coniferous canopies (Flood and Gutelius 1997)� In the same region, with conifer-dominated stands and dense overstory, Means (2000) observed a very low penetration to the ground of only 1–5%, for a small-footprint lidar� Kraus and Pfeifer (1998) estimated a penetration rate
Full waveform
First return
Second/intermediate return
Third/last return Return not recorded
0 5 m 6.3 m 12 m
Time delay Height (ns)
(m)
0 100 100
% Intensity % Intensity
Discrete returns
0 ns
40 ns 19 ns
23.3 ns
0 Reset/dead time
range equivalent
Returned intensity is sampled at 1–ns time interval
No reset/dead time Entire waveform is recorded
FIgure 3.9
Conceptual differences between discrete-return and waveform lidar systems�
of less than 25% for their lidar study in the Vienna Woods (Wienerwald), in Austria, using an Optech ALTM 1020 lidar system�
A study by Popescu, Wynne, and Nelson (2002) conducted in Virginia over forests of varying age classes including deciduous, coniferous, and mixed stands estimated the pen- etration rate for the last return laser hits, or first return when there was only one return, to be approximately 4%� The laser point density on the ground, for one flight line, was 0�47 points per square meter for the first return, and the last return when there was only one return; 0�20 points per square meter for the second return, less than half compared to the first return point density; 0�02 points per square meter for the third return; and 0�0001 points per square meter for the fourth return� None of the pulses were able to produce a fifth return for the given vegetation conditions, although the sensor, an AeroScan system that later became Leica’s ALS40 sensor (Leica Geosystems, Inc�, Heerbrugg, Switzerland), was configured to receive up to five returns�
Since the early 1990s, discrete-returns lidar sensors have experienced major technologi- cal advances, reflected mainly in an increased pulse frequency, the recording of multiple returns for each pulse, the recording of intensity information, and the positional accuracy�
The latest generation of airborne laser scanners has added waveform-recording capability and the ability to handle multiple pulses in the air� Systems that are able to track echoes from multiple pulses in the air have the potential to significantly increase the productivity of airborne lidar data acquisition systems as these systems, do not depend on receiving the target reflection before starting the next range measurement cycle� More pulses providing range measurements will enable lidar data users to fly a notably wider swath while main- taining the same point densities as conventional systems, or acquire significantly increased point densities for the same swath widths, leading to appreciably reduced flight costs in the end� Due to such innovative technological achievements developed by commercial laser systems manufacturers and the increased number of service providers, airborne lidar is used routinely for topographic mapping, vegetation assessment and forest inventory, 3D urban modeling, wireless communications planning, corridor mapping of power lines and oil pipes, and transportation planning, to mention just a few of the applications� Ground- based laser scanners have been used mainly for surveying and industrial 3D mapping�
Despite the advances in scanning lidar technology, a number of research groups are using airborne lidar profiling systems to extract elevation profiles along flight lines, mainly due to the lower cost of the sensor and the reduced data volume acquired during flight time� Such a system has been developed at NASA for forest research, called a “portable airborne laser system” (PALS), by Nelson, Short, and Valenti (2003)� This system is in fact based on off-the-shelf components, including a Riegl laser range finder, a Garmin GPS receiver, and a video camera�