Ensuring an Adequate Sample at Each Location in Point Sampling
Abstract:A current application of point sampling (in stand type inventories for example) controls the number of trees selected at a point by changing to a different basal area factor (BAF) if the initial BAF selects either too few or too many trees. The commonly used estimate of basal area per hectare for this sampling rule is biased. We derive the bias under an assumption which is reasonable for 20 sampling points established in one compartment of the Stanislaus National Forest. For this sample the bias was substantial (positive) even though the procedure investigated is less seriously biased than the one commonly used in actual practice. To ensure an adequate sample size at each sample location, we recommend using large fixed-area primary sampling units (PSU's) which are subsampled by a cluster of point samples in which the BAF is specified prior to sampling. National surveys conducted by the U.S. Forest Service currently use this procedure with 0.39 ha (one acre) PSU's subsampled with 10 points. Although little is known about either optimum PSU size or subsampling rate, it seems likely that for estimating timber variables efficiently, either PSU's should be increased in size or subsampling rate should be reduced, or both. Forest Sci. 27:567-573.
Document Type: Journal Article
Affiliations: Mathematical Statistician, Pacific Northwest Forest and Range Experiment Station, Portland, OR 97232
Publication date: September 1, 1981
More about this publication?
- Membership Information
- ingentaconnect is not responsible for the content or availability of external websites