Ecological Archives A025079A8
Nicolas Bircher, Maxime Cailleret, and Harald Bugmann. 2015. The agony of choice: different empirical mortality models lead to sharply different future forest dynamics. Ecological Applications 25:1303–1318. http://dx.doi.org/10.1890/141462.1
Appendix H. Performance values for the different model versions at Sigriswil and Scatlè.
Table H1. Performance values for the sites Sigriswil and Scatlè for the different model versions. Loglikelihoods were calculated using a linear regression between simulated and observed basal area and stem numbers respectively while (n) denotes the number of sampling points (i.e., number of inventories) available. For Sigriswil, basal area and stem numbers before and after harvest was taken into account. For calculating corrected AIC values (AICc), delta AIC (ΔAIC), and Akaike weights (ω), the number of parameters included in the different mortality functions (npar) was considered. AIC values were calculated only for Sigriswil as sampling size was too small at Scatlè. Relative bias and relative root mean squared error were calculate for both sites, but, again, note small sampling size at Scatlè. Highest performance for basal area and stem numbers at both sites are highlighted in gray.



basal area 
stem numbers 

Site (n) 
model version 
npar 
loglikelihood 
AICc 
ΔAIC 
ω 
bias % 
RMSE% 
loglikelihood 
AICc 
ΔAIC 
ω 
bias% 
RMSE% 
Sigriswil (18) 
ForClim.3.0 
2 
31.88 
68.56 
1.80 
0.20 
2.81 
12.25 
90.88 
186.55 
0.16 
0.48 
5.72 
24.96 
IM_randNr 
6 
28.48 
76.59 
9.84 
0.00 
2.30 
10.04 
83.38 
186.39 
0.00 
0.51 
3.57 
15.57 

IM_threshold 
6 
37.78 
95.19 
28.44 
0.00 
4.22 
18.43 
116.51 
252.66 
66.27 
0.00 
40.63 
177.35 

TRM_randNr 
4 
28.36 
67.79 
1.04 
0.30 
2.01 
8.77 
91.54 
194.16 
7.77 
0.01 
5.86 
25.58 

TRM_threshold 
4 
27.84 
66.75 
0.00 
0.50 
1.88 
8.22 
95.83 
202.74 
16.34 
0.00 
8.13 
35.50 

Scatlè (3) 
ForClim.3.0 
2 
8.27 
 
 
 
1.19 
2.52 
16.11 
 
 
 
4.76 
10.10 
IM_randNr 
6 
9.04 
 
 
 
4.02 
8.53 
15.64 
 
 
 
0.30 
0.64 

IM_threshold 
6 
9.53 
 
 
 
3.94 
8.35 
18.08 
 
 
 
24.24 
51.42 

TRM_randNr 
4 
9.00 
 
 
 
4.48 
9.51 
16.20 
 
 
 
0.54 
1.13 

TRM_threshold 
4 
9.89 
 
 
 
9.04 
19.17 
16.91 
 
 
 
2.34 
4.97 