PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
February 2014
179
Combining RapidEye Satellite Imagery and Lidar
for Mapping of Mining and Mine Reclamation
Aaron E. Maxwell, Timothy A. Warner, Michael P. Strager, and Mahesh Pal
Abstract
The combination of RapidEye satellite imagery and light
detection and ranging (lidar) derivatives was assessed for
mapping land-cover within a mountaintop coal surface mine
complex in the southern coalfields of West Virginia, USA.
Support vector machines (
SVM
), random forests (
RF
), and
boosted classification and regression trees (
CART
) algorithms
were used. Incorporation of the lidar-derived data increased
map accuracy in comparison to using only the five imagery
bands, and
SVM
generally produced a more accurate classi-
fication than the ensemble tree algorithms based on overall
map accuracy, Kappa statistics, allocation disagreement,
quantity disagreement, and McNemar’s test of statistical
significance. Based on measures of predictor variable impor-
tance within the ensemble tree classifiers, the normalized
digital surface model (
nDSM
) was found to be more useful
than first return intensity data for differentiating the classes.
Introduction
Commercial satellite imagery such as Ikonos, GeoEye, and
RapidEye provide high spatial resolution but low spectral
resolution compared to sensors such as Landsat Thematic
Mapper (
TM
), Enhanced Thematic Mapper Plus (
ETM+
), or
Moderate Resolution Imaging Spectrometer (
MODIS
) (Warner
et al
., 2009). Although high spatial resolution can yield fine
detail for land-cover and vegetative mapping, classification is
complicated by the increased spatial resolution and decreased
spectral resolution. Fine spatial resolution tends to generate
high internal variability within land-cover classes, which
can lead to decreases in classification accuracy (Townshend,
1981; Cushnie, 1987; Townshend, 1992; Baker
et al
., 2013).
This research investigated a potential means to enhance clas-
sification accuracy by combining high-resolution commercial
satellite imagery with light detection and ranging (lidar) data.
The analysis focused on mapping land-cover classes in
a mountaintop coal surface mine complex in the southern
coalfields of the eastern United States. Because surface mine
complexes experience rapid change due to human distur-
bance and reclamation, they are particularly good examples
of disturbed landscapes. Although this research focuses on
mapping land-cover within a mountaintop coal mine, the
challenges in mapping mining landscapes are typical of other
disturbances, such as timber harvesting, urban sprawl, etc.
This work adds to prior remote sensing of surface mines
research by investigating information gained by combining
lidar and commercial satellite data for mapping land-cover
(Cowen
et al
., 2000). This research had two components.
First, we assessed lidar-derived inputs as predictor variables
when combined with commercial satellite imagery to enhance
land-cover mapping. Second, we compared three machine
learning algorithms for the classification: support vector ma-
chines (
SVM
), random forests (
RF
), and boosted classification
and regression trees (
CART
). The image data consisted of com-
mercial RapidEye imagery. Lidar-derived predictor variables
included the normalized digital surface model (
nDSM
) gener-
ated by subtracting ground return data from the first return
data, first return intensity data, and the first return intensity
range within raster grid cells.
Background
Machine Learning Classification
Research has highlighted the improvement in classification
accuracy when lidar is combined with optical data, suggest-
ing that lidar can provide important predictor variables for
mapping land-cover (Cowen
et al
., 2000; Brennan and Web-
ster, 2006; Bork and Su, 2007; Chust
et al
., 2008; Chen
et al
.,
2009; Guo
et al
. 2011). The combination of imagery and lidar
data has been investigated in heterogeneous rangeland envi-
ronments (Chen
et al
., 2009), urban landscapes (Brennan and
Webster, 2006; Guo
et al
., 2011), and coastal estuary environ-
ments (Brennan and Webster, 2006; Chust
et al
., 2008). Guo
et
al
. (2011) specifically noted the usefulness of nDSM data for
mapping urban landscapes.
Combining disparate data such as imagery and lidar poses
distinct challenges because the combined data set may not
meet distribution assumptions required for traditional para-
metric classifiers. Machine learning algorithms have emerged
as an alternative to parametric classifiers and have been
shown to be more accurate and efficient when faced with high
dimensional and complex data (Hansen et al., 1996; Huang
et al., 2002; Rogan
et al
., 2003; Pal, 2005; Na
et al
., 2010;
Ghimire
et al
., 2012). Machine learning algorithms, such as
artificial neural networks (Del Frate
et al
., 2003),
SVM
(Pal
and Mather, 2005; Pal, 2005), and decision trees (Waske and
Braun, 2009), do not make assumptions regarding the data
distribution (Loosvelt
et al
., 2012). In summary, in remote
Aaron E. Maxwell is with Alderson Broaddus University, 101
College Hill Drive, Philippi, WV 26416, (
.
Timothy A. Warner is with West Virginia University, Depart-
ment of Geology and Geography, West Virginia University,
Morgantown, WV 26506-6300.
Michael P. Strager is with West Virginia University, Division
of Resource Management, 2004 Agricultural Science Building,
West Virginia University, Morgantown, WV 26506-6108.
Mahesh Pal is with the Department of Civil Engineering, Na-
tional Institute of Technology, Kurukshetra, 136119, Haryana,
India.
Photogrammetric Engineering & Remote Sensing
Vol. 80, No. 2, February 2014, pp. 179–189.
0099-1112/14/8002–179
© 2013 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.80.2.179
