Point data generated from helicopter surveys are used to determine the location and magnitude of mountain pine beetle infestations. Although collected for tactical planning, these data also provide a rich source of information for scientific investigations. To facilitate spatial research, it is important to consider how to best represent spatially explicit mountain pine beetle infestation data. This paper focuses on the spatial representation of point-based aerial helicopter surveys, which can be difficult to represent due to issues associated with large data quantities and data uncertainty. In this paper, the benefit of using a kernel density estimator to convert point data to a continuous raster surface is demonstrated. Field data are used to assess the accuracy of the point-based aerial helicopter survey data and the kernel density estimator is extended to incorporate data uncertainty. While the accuracy of point-based aerial surveys is high, with 92.6% of points differing by no more than ± 10 trees, there is a general tendency to overestimate infestation magnitude. The method developed for incorporating uncertainty into the kernel density estimator reduces overestimation and improves the correspondence between estimated infestation intensities and field data values. RESUME Des points de donnees tirees de sondage par helicoptere sont utilises pour determiner la localisation et l’importance des infestations de dendroctone du pin. Meme si elles ont ete recueillies pour des raisons de planification tactiques, ces donnees constituent egalement une importante source d’information pour les etudes scientifiques. Afin de faciliter la localisation spatiale, il est important d’etudier comment on peut representer le mieux les donnees explicites en terme de localisation des infestations de dendroctone du pin. Cet article porte sur la representation spatiale des points tires des sondages par helicoptere, ce qui peut etre difficile a realiser compte tenu des questions entourant les grandes quantites de donnees et l’incertitude qu’elles comportent. Dans le cas present, l’avantage de l’utilisation d’un estimateur de noyau de densite pour la conversion des points de donnees en surface continue est demontre. Des donnees de terrain sont utilisees pour evaluer la precision des points de donnees du sondage aerien par helicoptere et l’estimateur de noyau de densite est concu pour comprendre l’incertitude entourant les donnees. Meme si la precision des points de donnees par sondage aerien est elevee, 92 % des points ne differant par pas plus de ± 10 arbres, on retrouve une tendance generale de surestimer l’etendue de l’Infestation. La methode elaboree pour incorporer l’incertitude au sein de l’estimateur de noyau de densite reduit la surestimation et ameliore la correspondance entre les intensites d’infestation estimees et les valeurs en provenance des donnees du terrain.
[1]
L. Safranyik.
Management of Lodgepole Pine to Reduce Losses From The Mountain Pine Beetle
,
2002
.
[2]
P. Diggle,et al.
Non-parametric estimation of spatial variation in relative risk.
,
1995,
Statistics in medicine.
[3]
John Krygier,et al.
Visualization in Geographic Information Systems
,
1995
.
[4]
P. Diggle.
A Kernel Method for Smoothing Point Process Data
,
1985
.
[5]
C. D. Kemp,et al.
Density Estimation for Statistics and Data Analysis
,
1987
.
[6]
A. Fall,et al.
Provincial Level Projection of the Current Mountain Pine Beetle Outbreak: An Overview of the Model
,
2004
.
[7]
Elliott D. Kaplan.
Understanding GPS : principles and applications
,
1996
.
[8]
Ziheng Yang.
Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: Approximate methods
,
1994,
Journal of Molecular Evolution.
[9]
Trevor C. Bailey,et al.
Interactive Spatial Data Analysis
,
1995
.
[10]
Michael A. Wulder,et al.
Detection and monitoring of the mountain pine beetle
,
2004
.
[11]
R. H. Silversides,et al.
Dispersal of released mountain pine beetles under the canopy of a mature lodgepole pine stand
,
1992
.
[12]
Spatial-Temporal Analysis of Mountain Pine Beetle Infestations to Characterize Pattern , Risk , and Spread at the Landscape Level
,
2004
.