Canadian Forest Service Publications

Novel aerial photography as an aid to sampling secondary structure in pine stands. 2009. Teti, P. Natural Resources Canada, Canadian Forest Service, Pacific Forestry Centre, Victoria, B.C. Mountain Pine Beetle Working Paper 2009-16. 24 p.

Year: 2009

Available from: Pacific Forestry Centre

Catalog ID: 30122

Language: English

Series: Mountain Pine Beetle Working Paper (PFC - Victoria)

CFS Availability: Order paper copy (free), PDF (download)

Abstract

This project describes a new method for sampling secondary structure in beetle-attacked pine stands and reports on a test of its accuracy. The method consists of interpreting aerial photos taken when snow covers the forest floor and the sky is overcast. Winter/overcast aerial photos of beetle-attacked, pine-leading test plots in British Columbia, Canada, were acquired with a digital SLR camera in March 2008, geo-referenced to provincial orthophotos, and interpreted for secondary structure numbers, locations, and sizes. The resulting stem maps were field-checked for accuracy in the summer and fall of 2008. Tree classification tests consisted of determining whether each interpreted tree was in the plot, whether it was a green conifer, and whether any green conifers in the plot were missed. Net errors in counting secondary structure stems ranged from +3% to +21% in four plots; the most common error was the misclassification of dead pine trees or brush as green conifers. The diameters of green conifer crowns were also measured on the aerial photos and field-checked. Photo-interpreted crown diameters averaged 38 cm less than actual crown diameters, with 78% of the variance explained. As these accuracies are based on field-checking only four plots, they are preliminary; however, the results provide the first known demonstration that secondary structure can be interpreted on winter/overcast aerial photos. We conclude that this method can provide useful information on secondary structure at a rate of coverage, cost per unit area, and accuracy suitable for many applications if field-checking is used for training and verification.

Date modified: