摘要

A simple study was designed to assess the magnitude of model error in estimates of total green weight per acre in a sample loblolly pine plantation. A model was fitted to a large felled-tree data set and used to predict weight per ha in the sample stand. This is a simple case and, even here, ignoring model error would result in confidence intervals on the order of 10-40% too narrow.

  • 出版日期2014-8