Measurement error effects on estimates from linear and nonlinear regression whole-stand yield models

Research output: Contribution to journalArticlepeer-review


Systems of whole-stand yield models facilitate projections of forest attributes, but their inputs may be difficult to measure accurately. This study conducted sensitivity analyses to examine the effect of systematic and stochastic measurement errors on outputs from a representative system of equations. Simulated error was added to explanatory variables stand age, site index, or both. Results showed that large systematic error in one variable tended to produce moderate to large percent changes in all models, particularly the height and volume equations (often >50% change). Systematic error in both variables amplified this effect, especially for young, less productive stands. Stochastic error dramatically increased estimate variability (some relative standard errors >50%), particularly in the height and volume models at young ages and low site indices. These results suggest that measurement error may considerably alter projections and increase uncertainty when using whole-stand yield models, highlighting the need for careful crew training.

Original languageEnglish (US)
Article numbere12384
JournalNatural Resource Modeling
Issue number1
StatePublished - Feb 2024

Bibliographical note

Publisher Copyright:
© 2023 The Authors. Natural Resource Modeling published by Wiley Periodicals LLC.


  • errors-in-variables
  • sensitivity analysis
  • stochastic measurement error
  • system of equations
  • systematic measurement error
  • whole-stand yield models


Dive into the research topics of 'Measurement error effects on estimates from linear and nonlinear regression whole-stand yield models'. Together they form a unique fingerprint.

Cite this