Sampling Aerial Photographs for Estimation of Pest-Caused Tree Mortality
Six methods of estimating pest-caused tree mortality were evaluated by computer-based sampling of the aerial photo and numerical database of a 9,320 ha area of ponderosa pine (Pinus ponderosa) in the central Sierra Nevada, California. Two sampling universes were defined, one comprising 9 X 9 in photo units, the other 70 mm photo units. Of the sampling designs simulated, 2-stage designs involving subsampling of photo strips with either equal or variable probability provided estimates with the lowest standard deviations, and hence would require smaller sample sizes (and less photo-interpretation time) for a specified precision. The findings of this study have application to operational surveys of pest damage for which complete aerial photo coverage is available, or to the design of surveys for which only sample aerial photo coverage is feasible. West J. Appl. For. 1:84-89, July 1986.
No Supplementary Data
No Article Media
Document Type: Journal Article
Affiliations: Biostatistician-programmer, Environmental Health Associates, Oakland, CA
Publication date: 01 July 1986