0.2 



0.0 



-1 1 r~ 



5 10 15 



Years Since Last Disturbance 



20 



Figure 18 — Use of the calibration feature in the Regeneration Es- 

 tablishment fvlodel. Line A is the predicted probability of stocking 

 curve. Point b represents an inventory from the same stand but 

 stocking is better than predicted. This information is used to ad- 

 just the probability stocking curve upward to line B. Point c repre- 

 sents an inventory from the same stand as line A but stocking is 

 poorer than predicted. The probability of stocking curve is ad- 

 justed downward (line C). 



from year to year. Variation can also be expected among inventories of the 

 same stand. If different starting points are chosen for a transect or samphng 

 grid, one would expect different results from the inventory. 



The use of pseudo-random numbers in the regeneration model also resulted 

 in inconsistencies among projections of the same stand where projections 

 differed only in cycle lengths. For example, when two 5-year cycles are used 

 instead of one 10-year cycle, the pseudo-random numbers are used in a differ- 

 ent sequence between projections. Consistency was obtained by using pseudo- 

 random numbers in the same order for all tallies within a tally sequence. 



To further improve consistency, species composition is held constant at 

 10 years since last disturbance. This is necessary because of the relative 

 increases and decreases over time in the probability of species presented in 

 appendix B. Species composition is relatively stable at 10 years, and none 

 of the species in the model has long delays to establishment. 



An example of consistency among runs is given in table 8. Three projec- 

 tions are compared — four 5-year projections, two 10-year projections, and one 

 20-year projection. The probability of stocking, density, and species composi- 

 tion are nearly identical, regardless of cycle length. The small differences 

 that do occur are attributable to very slight differences in probabilities when 

 regeneration from one tally adds basal area to the stand in the next tally — 

 the probability of stocking and species probabilities are shifted slightly. 



35 



