APPENDIX

TABLE A1.  

Number of observations, mean squared error, standard error, 
mean increment, and mean residual values for  height growth potential and 
competition model verification and model validation by species.
						Species
                		  PP	SP	IC	DF	WF	RF
Verification & Fitting  (PERM1a) 
    Potential
	Number of observations	 174	347+    *	320	141	**
	MSE			 9.62	15.50	*	19.86	15.60	**
	Standard error		 3.10	3.94	*	4.46	3.95	**
    Competition
	Number of observations	 1261	347	820	585	1464	271
	MSE			 21.37	15.63	17.97	23.89	16.54	15.00
	Standard error		 4.62	3.95	4.24	4.89	4.07	3.87
	Mean HT increment	 5.49	5.32	3.66	5.64	5.70	4.83
    Mean Residual		-0.006	.046	0.130	0.182	-0.001	1.353
Validation  (PERM1b)
	Number of observations	 1178	373	841	791	1748	***
	Mean HT increment	 4.50	5.56	3.56	4.95	5.22	***
	Mean Residual		-1.02	0.17	-0.02	-0.48	-0.30	***

*	Old model coefficients used
**	White fir model coefficients used
***	All RF trees are in PERM1a with no trees left for the validation data 
        set.
+	Number of observations used in SP Potential  Adjustment, PP Potential 
        model used.

APPENDIX TABLE A2.  Number of observations, mean squared error, standard error, 
mean increment, and mean residual values for diameter growth potential and 
competition model verification and model validation by species.
				Species
			        PP	SP	IC	DF	WF	RF
Verification & Fitting  (PERM1a) 
    Potential
	Number of observations 	549	140+	820	215	582	271
	MSE			424.2	1074.9	315.5	588.1	551.7	569.6
	Standard error		20.6	32.8	17.8	24.3	23.5	23.9
    Competition
	Number of observations	1261	743	*	*	1464	271
	MSE			313.4	534.1	*	*	373.6	553.2
	Standard error		17.7	23.1	*	*	19.3	23.5
	Mean DBH increment	NA	NA	*	*	5.8	6.5
    Mean Residual		-0.06	1.21	0.94	1.07	0.49	0.44
Validation  (PERM1b)
	Number of observations	1178	**	839	791	1748	**
	Mean DBH increment	4.8	**	4.7	5.4	5.4	**
	Mean Residual		-1.05	**	.54	-1.21	1.05	**

*	Old model coefficients used
**	All SP and RF trees are in PERM1a
+	Number of observations for adjustment of potential function; potential 
        from the previous SP model was used.
	
The authors are Professor and Post Graduate Researcher, respectively, Department 
of Forestry and Resource Management, University of California.  The contributions 
of Walter Meerschaert, Pamela Schwartz, and Craig Olson are gratefully acknowledged.
Research conducted with support from the Northern California Forest Yield Cooperative
under University of California Agriculture Experiment Station Project MS-3815.

To complete the coverage of northern California, the north coast redwood and Douglas-
fir forests are simulated by the CRYPTOS model by Wensel, Meerschaert and Krumland 
(1987).	

The data from the Mendocino region was omitted from the tabulation in Table 2  since 
it was not used for fitting the parameters given here.
The growth rates for Region 4 appeared to be different enough from the other regions 
that the data were held aside for a separate analysis.  Additional samples will be 
collected for this analysis.
Research Note No. 23		
Revised parameter estimates for CACTOS growth models