Testing a mechanistic model for predicting stand and tree growth

R. L. Korol, K. S. Milner, S. W. Running

Research output: Contribution to journalArticlepeer-review

36 Scopus citations


Given the uncertainty of future climate regimes, it has become necessary to develop growth and yield models that can respond to potential changes in climate. TREE-BGC, a derivative of the physiological process model FOREST- BGC, was used to simulate the growth of 998 trees in uneven-aged stands near Kamloops, B.C. Stand variables were derived from a tree list. The model used a disaggregation logic to allocate stand level estimates of carbon gain and respiration costs to individual trees. Increments in height and diameter were estimated so as to maintain their allometric relation, but scaled to produce an estimate of stem volume equivalent to the carbon allocated to the trees. Mortality was simulated when the maintenance respiration of the tree exceeded the carbon allocated to the tree. Plot level estimates of 20 yr basal area increment and volume increment were highly correlated with actual measurements (r2 = 0.94 and 0.96, respectively; n = 24). The cumulative modeled diameter and height distributions were compared to measured distributions. The simulated cumulative diameter and height distributions were not significantly different from the actual cumulative diameter and height distributions for 23 of the 24 plots (α = 0.05). The model was found to have an accuracy of about 0.16 m2 ha-1 yr-2 for basal area and 1.51 m3 ha-1 yr-1 for volume (α' = α' = 0.05), averaged over the 20 yr period. Long-term model behavior was influenced by stand density, basal area, and volumes. Stand mortality appeared to emulate the so-called self-thinning rule, and a maximum size-density relationship was found.

Original languageEnglish
Pages (from-to)139-153
Number of pages15
JournalForest Science
Issue number2
StatePublished - May 1996


  • Net primary production
  • carbon balance
  • maintenance respiration


Dive into the research topics of 'Testing a mechanistic model for predicting stand and tree growth'. Together they form a unique fingerprint.

Cite this