Validating the southern variant forest vegetation simulator height predictions on southeastern hardwoods in Kentucky and Tennessee

Abstract

The accuracy of forest growth and yield forecasts affects the quality of forest management decisions (Rauscher et al. 2000). Users of growth and yield models want assurance that model outputs are reasonable and mimic local/regional forest structure and composition and accurately reflect the influences of stand dynamics such as competition and disturbance. As such, simulation models should be subjected to a process of evaluation in order to build confidence in their validity (Reynolds et al. 1981, Vanclay et al. 1996). Operational validation, as defined by Rykiel (1996), tests whether the model output conforms with its stated purpose. Operational validation may be viewed from two perspectives: (1) hypothesis testing and (2) confidence interval estimation (Rykiel 1996, Rauscher 2000). Confidence interval estimation expresses the degree of reliability that can be placed in model predictions through estimates of the direction, magnitude, and variability of the prediction error and is the approach used in this study.

Citation

Parresol, Bernard R.; Stedman, Steven C. 2004. Validating the southern variant forest vegetation simulator height predictions on southeastern hardwoods in Kentucky and Tennessee. In: Joint 2004 annual general meeting and convention: 1-11