Open Collections

Abstract

Colour-infrared positive transparencies, at scales of 1:1000, 1:2000, 1:3000 and 1:4000, were obtained of a lodgepole pine (Pinus contorta Dougl.) stand which was undergoing an active mountain pine beetle (Dendroctonus ponderosae Hopkins) infestation. Visual and densitometric photo-interpretation techniques were applied to these transparencies to determine if symptoms of incipient mortality could be detected from green-foliaged trees following attack by mountain pine beetle. The purpose of this study was to determine the efficiency of photo-interpretation at scales 1:1000, 1:2000, 1:3000, and 1:4000 for early detection of mountain pine beetle-attacked trees. Visual interpretation involved classification of trees into damage types. Subsequent ground-checking allowed a determination of accuracy of damage type classification at each scale. Results indicated that 15% fewer trees were correctly classified as healthy at 1:4000 than at 1:1000. Separation of healthy from attacked trees was achieved with an overall accuracy of 89% at 1:1000, 83% at 1:2000, 91% at 1:3000, and 83% at 1:4000. The interpretation process at 1:4000, however, was considerably slower than at the larger scales. Total green (TG), total red (TR) and total infrared (TIR) film response values of the images of newly-infested trees differed from measurements of the healthy trees. At 1:1000, images of newly-stressed trees indicated significantly higher (p