Assessing uncertainty in mechanistic models

Abstract

Concern over potential global change has led to increased interest in the use of mechanistic models for predicting forest growth. The rationale for this interest is that empirical models may be of limited usefulness if environmental conditions change. Intuitively, we expect that mechanistic models, grounded as far as possible in an understanding of the biology of tree growth, may be more useful in an altered environment. Unfortunately, such models often produce only point estimates, with no associated credible or confidence intervals. The Bayesian synthesis (BSyn) method provides a solution to this dilemma. We present a summary of the BSyn, and the results of an application of BSyn to PIPESTEM, mechanistic model of forest growth calibrated for loblolly pine.

Parent Publication

• Integrated tools for natural resources inventories in the 21st century

Citation

Green, Edwin J.; MacFarlane, David W.; Valentine, Harry T. 2000. Assessing uncertainty in mechanistic models. In: Hansen, Mark; Burk, Tom, eds. Integrated tools for natural resources inventories in the 21st century. Gen. Tech. Rep. NC-212. St. Paul, MN: U.S. Dept. of Agriculture, Forest Service, North Central Forest Experiment Station. 501-506.