Cull Estimation 



The logistic function was used twice to estimate cull volume on the Kaniksu 

 National Forest. First it was used to predict the probability that a tree of given 

 characteristics was a cull tree. The dependent variable assumed the value 1 when the 

 tree was a cull tree and otherwise assumed the value 0. The independent variables 



were DBH and a measure of stand density, jy^jj • (V/a is gross cubic foot volume per 

 acre.) For hemlock the resulting model is 



P = {1 + exp[9.1010 - 0.3696(Z?Bfl) - . 00348 (^) ]} ~ 1 



The second use of the function was to estimate the probability that a noncull tree 

 had some cull volume. The dependent variable assumes the value 1 if a tree classified 

 as noncull has some cull volume and otherwise assumes the value 0. The independent 



variables were DBH and (^577) • For the cedar-hemlock species mix, the model for pre- 

 dicting the probability of "cubic-foot defect in noncull trees is 



P = {1 + exp[4.494 - . 3776 (DBH) ] }~ * 



Cubic-foot defect in noncull trees was estimated by the product of P and a linear func- 

 tion that estimates cubic- foot defect in trees having cull as a function of DBH and 



V/a 

 DBH * 



Canker Inactivation 



The algorithm was also used to investigate levels and trends of natural inactiva- 

 tion of western white pine blister rust cankers and thereby to determine causal rela- 

 tionships for natural inactivation and to explain the variation that occurs in natural 

 inactivation . 



The following model demonstrates the relationship between the probability of a 

 canker becoming inactive in the coming year (P) , stand elevation (EL), canker age (CA) , 

 and branch diameter (BD) : 



P = {1 + exp[0.883 + 0.0323(a) - 0.304(C4) + 1 . 61 7 (BD) ] } " 1 



Readers unconcerned with the theoretical development of the algorithm and its modi- 

 fications may omit reading the next four sections. Enough detail has been provided in 

 the section "Guides for the Use of Risk", page 12, to run the program. 



4 



