and shading trees is recalculated for each distance, provid-
ing a reduced bias in the estimation of the number of trees of
smaller
DBH
classes.
The Relaskop-based method for correction of instrument
bias combined with the Poisson attenuation model for correc-
tion of occlusions has proved to substantially reduce bias in
N
(Strahler
et al.
2008) and
BA
(Strahler
et al.
2008; Lovell
et al.
2011) estimated with single scan
TLS
data. Likewise, Yao
et al.
(2011) reported correlations of 0.902 and 0.656 with field data
when estimating
N
and
BA
with the Poisson attenuation model
for correction of occlusions without the Relaskop sampling.
However, in our analysis using ForeStereo data this method
yields biased results when compared with
HPC
estimates. In
contrast to
TLS
, where tree detection is better at the lower part
of the stem, stem sections viewed under an inclination angle
over the crown of more distant trees—seen against the sky—
are most accurately detected in the hemispherical images.
The Relaskop-based method selects trees depending on
DBH
,
whereas the method based on image segmentation considers
the diameter at the height corresponding to the inclination
angle of view, making the latter preferable for the RGB photos
acquired using ForeStereo. However, the Relaskop-based and
Poisson attenuation model methods performance may be
improved through the adjustment of
BAF
and
D
E
.
The results described in this work demonstrate the ef-
fectiveness of ForeStereo as a
PS
technology for field sampling
in operational forest inventories. Cost analyses are beyond
the scope of this study, but the performance capacity shown
by ForeStereo in field surveys for this study—a team of two
people surveyed 16 sampling points located on a systematic
grid of 400
×
400 m, covering 256 ha per day on average—en-
ables increased sampling and cost optimization of the forest
inventory field stage. Two stage sampling design, combining
PS
technologies and spatially explicit remotely sensed data,
may improve the accuracy of inference estimations in forest
inventories (Brosofske
et al.
2014). Precise and unbiased
estimators for plot level variables are still to be developed.
The approaches tested in this work can be used to deal with
bias, our results pointing to the distance-samp
as covariate and the
HPC
method proposed her
approaches with ForeStereo data. Other source
fecting ForeStereo estimates—light conditions,
or species on image segmentation, or the limited accuracy of
distance estimations in the proximity of the base line—are
beyond the scope of this study and should be the subject
of future research. As an alternative, remote sensing and
PS
data may be integrated during data processing: Korpela
et
al.
(2007) combined photogrammetry and field triangula-
tion for complete mapping of the stand, whereas Liang
et al.
(2016) suggested the use of
TLS
for retrieval of individual tree
attributes and airborne remote sensing data to extend the es-
timation of tree attributes to all trees within the stand, which
renders the estimation of stand-level summary statistics
from
TLS
data unnecessary. Whichever approach is chosen,
the next step is to implement these methods in user-friendly
software packages—such as the ForeStereo software presented
in Sánchez-González
et al.
(2016), which has been used for
this study—so that these methods can be made available for
foresters and stakeholders.
Acknowledgments
The authors wish to thank Roberto Vallejo and the National
Forest Inventory team of the Spanish Ministry of Agriculture,
Fisheries, Food and Environment and Mancomunidad Forest-
al Ansó-Fago for their support with field data. This research
has been funded by the National Parks Autonomous Agency
of the Spanish Ministry of Agriculture, Fisheries, Food and
Environment through the project 979S/2010 of National Parks
and the Spanish Ministry of Economy and Competitiveness
through the project AGL2016-76769-C2-1-R.
References
Astrup, R., M. J. Ducey, A. Granhus, T. Ritter and Lüpke. 2014.
Approaches for estimating stand-level volume using terrestrial
laser scanning in a single-scan mode.
Canadian Journal of Forest
Research
44:666–676. doi:
0535.
Bitterlich, W. 1947. Die Winkelzählmessung.
Allgemeine Forst- und
Holzwirtschaftliche Zeitung
58:94
Bottalico, F., G. Chirici, R. Giannini, S. Mele, M. Mura, M. Puxeddu,
R. E. McRoberts, R. Valbuena and D. Travaglini. 2017. Modeling
Mediterranean forest structure using airborne laser scanning
data.
International Journal of Applied Earth Observation
57:145–153.
Bouvier, M., S. Durrieu, R. A. Fournier and J.-P. Renaud. 2015.
Generalizing predictive models of forest inventory attributes
using an area-based approach with airborne
LiDAR
data.
Remote
Sensing of Environment
156:322–334.
Brosofske, K. D. R. E. Froese, M. J. Falkowski and A. Banskota. 2014.
A review of methods for mapping and prediction of inventory
attributes for operational forest management.
Forest Science
60:733–756. doi:
Brunner, A. and B. Gizachew. 2014. Rapid detection of stand density
tree positions and tree diameter with 2D terrestrial laser scanner.
European Journal of Forest Research
133:819–831. doi: 10.1007/
s10342-014-0799-1.
Buckland, S. T., D. R. Anderson, K.,P. Burnham, J.,L. Laake, D.,L.
Borchers and L. Thomas. 2001. Introduction to distance-
sampling: Estimating abundance of biological populations.
Oxford, U.K.: Oxford University Press.
Clark, N. A., R. H. Wynne, D. L. Schmoldt and M. F. Winn. 2000.
An assessment of the utility of a non-metric digital camera
for measuring standing trees.
Computers and Electronics in
Agriculture
28:151–169. doi: 10.1016/S0168-1699(00)00125-3.
Clark, R. G. 2016. Statistical efficiency in distance-sampling.
PLoS
ONE
11:e0149298. doi: 10.1371/journal.pone.0149298.
ier, G. W. Frazer and K. O. Niemann. 2012. A
ctural model of trees to enhance
LiDAR
-derived
f forest canopy structure.
Agricultural and Forest
Meteorology
166/167:72–85.
Dassot, M., T. Constant and M. Fournier 2011. The use of terrestrial
LIDAR technology in forest science: Application fields benefits
and challenges.
Annals of Forest Science
68:959–974.
Ducey, M. J. and R. Astrup. 2014. Adjusting for nondetection in forest
inventories derived from terrestrial laser scanning.
Canadian
Journal of Remote Sensing
39:410–425. doi:
doi/abs/10.5589/m13-048.
Forsman, M., N. Börlin and J. Holmgren. 2012. Estimation of tree stem
attributes using terrestrial photogrammetry. Pages 261–265 of the
International Archives of the Photogrammetry Remote Sensing
and Spatial Information Sciences, XXII ISPRS Congress
, held in
Melbourne, Australia. .
Garestier, F., P. Dubois-Fernández and I. Champion. 2008. Forest
height inversion using high resolution P-band Pol-InSAR data.
IEEE Transactions on Geoscience and Remote Sensing
46:3544–
3559.
Gea-Izquierdo, G., F. Montes, R. Gavilán, I. Cañellas and A. Rubio.
2015 Is this the end? Dynamics of a relict stand from pervasively
deforested ancient Iberian pine forests.
European Journal of
Forest Research
134:525–-536. doi: 10.1007/s10342-015-0869-z.
Gómez, C., M. A. Wulder, F. Montes and J. A. Delgado. 2011. Forest
structural diversity characterization in Mediterranean pines
of central Spain with QuickBird-2 imagery and canonical
correlation analysis.
Canadian Journal of Remote Sensing
37:628–642. doi:10.5589/m12-005.
506
July 2019
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
