The digital timber volume estimation method was developed primarily for use with Landsat MSS scanner data. The technique makes use of a vector field clustering algorithm, nearest neighbor classification, and regression analysis. When such a technique is to be applied to reduce the cost for a given level of precision in a forest inventory, the interface between sampling methods and the digital estimation model must be considered. The candidate models and sampling methods must be evaluated using test data as closely related to actual circumstances as possible.
Date of this Version