COMPONENT MODELS 



Full implementation of this prognosis algorithm requires three kinds of models for: 

 Development of individual trees 



Development of regeneration stands including ingrowth into existing stands 



Transition from regeneration phase to individual tree phase 



Only the model for the first of these three processes is described in this paper. The 

 latter two processes will require some additional silvicultural research before general 

 models can be derived. Growth functions should be based on data derived from the area 

 to which the model is to be applied. The self-calibration feature of this model only 

 partially mitigates this admonition. 



Models for Development of Individual Trees'* 



The growth model for saplings and larger trees is a set of functions that predict 

 the rate of increase of tree d.b.h., the rate of increase in tree height, the change 

 in crown dimensions, and the change in bark thickness. The change in number of trees 

 per acre represented by each sample tree is based on a function estimating the 

 mortality probability. 



The nature of the variables that should be included in the growth functions is 

 controlled by the purposes for which the prognoses are to be used. For exan^ile, if 

 the only course of development to be modeled is the unmanaged trends of natural stands, 

 without catastrophic disturbance, then a variable representing past growth rate would 

 be a very effective predictor of succeeding growth rates. It would include most of 



'*The logic of the program provides for distinct functions for 11 different 

 species or species groups . To increase the number of species would not be particu- 

 larly difficult. 



8 



