Table 8.
Multivariate model set evaluation for A4
| Model | Formula | k | RSS | Adjusted R 2 | AICc | Model likelihood | w |
|---|---|---|---|---|---|---|---|
| g17 |
Temp + PC1 |
4 |
157.74 |
0.69 |
130.56 |
1.00 |
0.61 |
| g08 |
UV + Temp |
4 |
173.83 |
0.66 |
133.09 |
0.28 |
0.17 |
| g11 |
UV + Silt |
4 |
189.09 |
0.63 |
135.28 |
0.09 |
0.06 |
| g16 |
Temp + Silt |
4 |
195.87 |
0.61 |
136.19 |
0.06 |
0.04 |
| g23 |
AtmP + Silt |
4 |
196.90 |
0.61 |
136.33 |
0.06 |
0.03 |
| g24 |
AtmP + PC1 |
4 |
199.13 |
0.61 |
136.62 |
0.05 |
0.03 |
| g10 |
UV + AtmP |
4 |
206.78 |
0.59 |
137.60 |
0.03 |
0.02 |
| g19 |
RH + AtmP |
4 |
210.71 |
0.58 |
138.09 |
0.02 |
0.01 |
| g01 |
UV |
3 |
250.36 |
0.52 |
139.76 |
0.01 |
0.01 |
| g09 |
UV + RH |
4 |
228.52 |
0.55 |
140.20 |
0.01 |
0.00 |
| g20 |
RH + Silt |
4 |
237.10 |
0.53 |
141.16 |
0.01 |
0.00 |
| g27 |
All |
9 |
112.68 |
0.71 |
141.16 |
0.00 |
0.00 |
| g05 |
Silt |
3 |
269.41 |
0.49 |
141.67 |
0.00 |
0.00 |
| g13 |
UV + PC2 |
4 |
249.97 |
0.50 |
142.53 |
0.00 |
0.00 |
| g12 |
UV + PC1 |
4 |
250.00 |
0.50 |
142.54 |
0.00 |
0.00 |
| g04 |
AtmP |
3 |
311.23 |
0.41 |
145.42 |
0.00 |
0.00 |
| g02 |
Temp |
3 |
323.71 |
0.39 |
146.44 |
0.00 |
0.00 |
| g25 |
AtmP + PC2 |
4 |
297.46 |
0.41 |
147.06 |
0.00 |
0.00 |
| g15 |
Temp + AtmP |
4 |
298.47 |
0.41 |
147.14 |
0.00 |
0.00 |
| g14 |
Temp + RH |
4 |
305.03 |
0.40 |
147.71 |
0.00 |
0.00 |
| g18 |
Temp + PC2 |
4 |
323.17 |
0.36 |
149.21 |
0.00 |
0.00 |
| g21 |
RH + PC1 |
4 |
383.28 |
0.24 |
153.65 |
0.00 |
0.00 |
| g22 |
RH + PC2 |
4 |
383.91 |
0.24 |
153.69 |
0.00 |
0.00 |
| g06 |
PC1 |
3 |
430.20 |
0.18 |
153.84 |
0.00 |
0.00 |
| g26 |
PC1 + PC2 |
4 |
390.20 |
0.23 |
154.11 |
0.00 |
0.00 |
| g03 |
RH |
3 |
454.76 |
0.14 |
155.28 |
0.00 |
0.00 |
| g07 |
PC2 |
3 |
491.51 |
0.07 |
157.30 |
0.00 |
0.00 |
| g28 | naïve | 2 | 548.85 | 0.00 | 157.60 | 0.00 | 0.00 |
UV = UV index, Temp = average maximum temperature, RH = relative humidity, AtmP = atmospheric pressure, PC1 and PC2 = genetic principal components, k = number of parameter estimates, RSS = residual sum of squares, AICc = Akaike information criterion adjusted for the sample size, w = conditional model probability (likelihood of model i divided by the sum of model likelihoods).