PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
August 2015
625
Understanding the Effects of ALS Pulse Density
for Metric Retrieval across Diverse Forest Types
Phil Wilkes, Simon D. Jones, Lola Suarez, Andrew Haywood,
William Woodgate, Mariela Soto-Berelov, Andrew Mellor, and Andrew K. Skidmore
Abstract
Pulse density, the number of laser pulses that intercept a sur-
face per unit area, is a key consideration when acquiring an
Airborne Laser Scanning (
ALS
) dataset. This study compares
area-based vegetation structure metrics derived from multi-re-
turn
ALS
simulated at six pulse densities (0.05 to 4 pl m
-2
)
across a range of forest types: from savannah woodlands to
dense rainforests. Results suggest that accurate measurement
of structure metrics (canopy height, canopy cover, and verti-
cal canopy structure) can be achieved with a pulse density of
0.5 pl m
-2
across all forest types when compared to a dataset
of 10 pl m
-2
. For pulse densities <0.5 pl m
-2
, two main sources
of error lead to inaccuracies in estimation: the poor identi-
fication of the ground surface and sparse vegetation cover
leading to under sampling of the canopy profile. This analysis
provides useful information for land managers determining
capture specifications for large-area
ALS
acquisitions.
Introduction
The accurate and timely retrieval of vegetation structure met-
rics is a key component of vegetation management, monitor-
ing and reporting activities, and ecosystem modeling by land
management agencies and forest scientists. For example, cano-
py height is routinely gathered by land management agencies
around the world, e.g.,
Forest Inventory and Analysis
in the
US, the Canadian
National Forest Inventory
, and the
Victorian
Forest Monitoring Programme
in Australia. This information
is needed for the assessment of vegetation condition to fulfill
statutory and non-legislative reporting obligations such as
that agreed by the Santiago Declaration (Miles, 2002). Canopy
height is also widely used by forest scientists as a proxy to
estimate forest biomass (Asner
et al
., 2010; Drake
et al
., 2002),
and it is an Essential Climate Variable (Sessa, 2009).
Over the past decade, Light Detecting and Ranging (lidar)
and in particular, small-footprint Airborne Laser Scanner (
ALS
)
systems have progressed from an experimental technique to an
operational tool for area-based (e.g., plot or stand scale mapping
unit) attribution of vegetation structure (Wulder
et al
., 2012a
and 2012b).
ALS
allows for synoptic capture of large areas of the
landscape where remoteness or terrain complexity complicates
and increases the cost of establishing inventory plots or where
variance in structure is not captured with traditional sampling
(McRoberts and Tomppo, 2007; Mora
et al
., 2013). Furthermore,
ALS
is now recognized by some authors as a more accurate
method of directly measuring vegetation attributes as compared
to traditional forest inventory methods (Holmgren and Jonsson,
2004; Magnusson
et al
., 2007; Maltamo
et al
., 2006).
Acquisition over very large areas that transect regional to
continental extents are becoming viable as the ability to cap-
ture and process large volumes of data improves. For example,
the Canadian Forest Service captured data along a ~25,000 km
transect (Wulder
et al
., 2012a) and the Department of Environ-
ment, Land, Water, and Planning (
DEWLP
) captured lidar along
~27,000 km of riparian corridors in Victoria, Australia (Quad-
ros
et al
., 2011). However, the cost of large-area
ALS
acquisi-
tions remains significant. There are a number of possibilities
for reducing acquisition expenditure, such as using a sample
based approach where lidar “plots” are targeted to capture the
variance in forest structure (Wulder
et al
., 2012a) or optimiz-
ing acquisition parameters, such as sampling frequency, to
capture vegetation type and terrain conditions.
ALS
sampling
frequency is referred to as “pulse density” and is defined as
the number of emitted laser pulses that intercept a given area
of open ground (Evans
et al
., 2009); pulse density is expressed
here as pulses per m
2
(pl m
-2
). Pulse density is determined
by the pulse repetition frequency (
PRF
) of the instrument and
acquisition parameters such as aircraft flying height and air-
craft speed. Although a number of factors mean
PRF
and flying
height or speed are not independent; at a constant
PRF
, de-
creasing pulse density (on the ground) is achieved by increas-
ing flying height and/or aircraft speed (Baltsavias, 1999).
Phil Wilkes is with the School of Mathematical and Geospa-
tial Sciences, RMIT University, GPO Box 2476, Melbourne,
VIC 3001, Australia; the Cooperative Research Centre for
Spatial Information, Level 5, 204 Lygon Street, Carlton, VIC
3053, Australia; and ITC, University of Twente, P.O. Box 217,
NL-7000 AE Enschede, The Netherlands (phil.wilkes@rmit.
edu.au).
Simon D. Jones, Lola Suarez, and William Woodgate are with
the School of Mathematical and Geospatial Sciences, RMIT
University, GPO Box 2476, Melbourne, VIC 3001, Australia;
and the Cooperative Research Centre for Spatial Information,
Level 5, 204 Lygon Street, Carlton, VIC 3053, Australia.
Andrew Haywood is with the Cooperative Research Centre for
Spatial Information, Level 5, 204 Lygon Street, Carlton, VIC 3053,
Australia; and the Food and Agricultural Organisation of the
United Nations, Viale delle Terme di Caracalla, 00153 Rome, Italy.
Mariela Soto-Berelov is with the School of Mathematical and
Geospatial Sciences, RMIT University, GPO Box 2476, Mel-
bourne, VIC 3001, Australia.
Andrew Mellor is with the School of Mathematical and Geo-
spatial Sciences, RMIT University, GPO Box 2476, Melbourne,
VIC 3001, Australia; the Department of Environment, Water,
Land and Planning, P.O. Box 500, East Melbourne, VIC 3002,
Australia; and the Joint Remote Sensing Research Program,
School of Geography, Planning and Environmental Manage-
ment, University of Queensland, St Lucia, QLD 4072, Austra-
lia.
Andrew K. Skidmore is with ITC, University of Twente, P.O.
Box 217, NL-7000 AE Enschede, The Netherlands.
Photogrammetric Engineering & Remote Sensing
Vol. 81, No. 8, August 2015, pp. 625–635.
0099-1112/15/625–635
© 2015 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.81.8.625
