PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
January 2014
59
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
January 2014
59
Abstract
With the acquisition of lidar data for over 30 percent of the
US, it is now possible to assess the three-dimensional distri-
bution of features at the national scale. This paper integrates
over 350 billion lidar points from 28 disparate datasets into a
national-scale database and evaluates if height above ground
is an important variable in the context of other national-
scale layers, such as the US Geological Survey National Land
Cover Database and the US Environmental Protection Agency
ecoregions maps. While the results were not homoscedastic
and the available data did not allow for a complete height
census in any of the classes, it does appear that where lidar
data were used, there were detectable differences in heights
among many of these national classification schemes. This
study supports the hypothesis that there were real, detectable
differences in heights in certain national-scale classification
schemes, despite height not being a variable used in any of
the classification routines.
Introduction
The three-dimensional arrangement of vegetative and non-
vegetative structure has a profound effect on how ecosys-
tems function and cycle carbon, water, and nutrients. The
increased need to understand the range of local to global
dynamics of ecosystems in three dimensions has created a
demand for extensive ecosystem structure data (Shugart
et al
.,
2010). Accurate mapping of vegetative structure, specifically,
has important implications for natural resources management
and forest harvesting activities (Dubayah and Drake, 2000;
Lim
et al
., 2003), assessing the impacts of natural and anthro-
pogenic change on ecosystems (e.g., Weishampel
et al
., 2007),
and carbon, water, and energy cycling (Lefsky
et al
., 2005;
Chasmer
et al
., 2011).
Recently, trends in ecological research have been driven
by the need to develop predictive capabilities to understand
how ecosystems will respond to increasing anthropogenic
impacts. Predicting changes in community composition and
Jason M. Stoker is with the US Geological Survey (USGS)
Earth Resources Observation and Science (EROS) Center,
Sioux Falls, SD, and the Geographic Information Science
Center of Excellence, South Dakota State University,
Brookings, SD 57007
).
Mark A. Cochrane and David P. Roy are with the Geographic
Information Science Center of Excellence, South Dakota State
University, Brookings, SD 57007.
ecosystem dynamics requires a better understanding of the
ecological importance of structural characteristics of today’s
vegetation (Chapin
et al
., 1997). Because the Earth processes
function in more than two dimensions (or two dimensions
over time), the need to incorporate the third dimension into
today’s remotely sensed modeling and mapping efforts have
become very important (Levick, 2009).
Two frequently used vegetative properties are vegetation
height and canopy cover (Stojanova
et al
., 2010). Vegetation
height can be described as the average (mean or median)
height of the vegetation in a stand, relative to the ground.
Vegetation height is primarily a function of the species com-
position, climate and site quality, and time since disturbance
and can be used for land-cover classification or in conjunc-
tion with vegetation indices (Green
et al
., 2013). If combined
with species composition and site quality information, in
absence of field data, vegetation height serves as an approxi-
mate estimate of stand age or successional stage (Falkowski
et al
., 2009). Vegetation height is also a useful indicator of
forest age and habitat quality. It is an important input variable
for ecosystem and forest fire models, and is highly correlated
with vegetation biomass and productivity. Biomass is the key
component of the carbon circle (Zhao
et al
., 2009), especially
in forests, which are a primary reservoir of terrestrial carbon
(Skole and Tucker, 1993) and a surrogate for fuel loading esti-
mation (Finney, 2005).
While in the past most remote sensing applications have
involved two-dimensional raster-based analyses, light detec-
tion and ranging, or lidar, has become a widely used technol-
ogy that provides the capability to record the third dimension
of features such as bare earth, vegetation, and manmade struc-
tures. In the past decade, lidar has emerged as a powerful tool
for remotely sensed forest canopy and stand structure, includ-
ing the estimation of aboveground biomass and carbon storage
(Lefsky
et al
., 2005). Lidar instruments can produce outputs
of discrete returns of three-dimensional structure, or can even
digitize the entire vertical signal in a continuous waveform.
Two distinct lidar platforms have been pursued to achieve
large-area characterization of the Earth in three dimensions:
disparate, project-based airborne collections and continuous
space-borne lidar instruments, such as
ICESAT
.
Integrating Disparate Lidar Data at the
National Scale to Assess the Relationships
between Height Above Ground, Land Cover
and Ecoregions
Jason M. Stoker, Mark A. Cochrane, and David P. Roy
Photogrammetric Engineering & Remote Sensing
Vol. 80, No. 1, January 2014, pp. 59–70.
0099-1112/14/8001–59/$3.00/0
© 2014 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.80.1.59
