Table 2.
Comparison of goodness-of-fit and predictive power of models for SWP. Models were cross-validated by retaining one plot at a time as a validation dataset.
| No. | Models | AIC | BIC | r2 | RMSE training | RMSE cross validated |
|---|---|---|---|---|---|---|
| 1 | SWP = 1.243 · e4.011 · PLWP − 1.616 | −249.42 | −229.55 | 0.659 | 0.214 | 0.216 ± 0.038 |
| 2 | SWP = 1.518 · e4.063 · PLWP · VPDmax −0.233 − 1.614 | −339.94 | −315.10 | 0.688 | 0.205 | 0.210 ± 0.038 |
| 3 | SWP = 20.164 · e3.890 · PLWP · Tmax −0.819 − 1.628 | −388.95 | −363.76 | 0.702 | 0.201 | 0.204 ± 0.037 |
| 4 | SWP = 1.479 · e2.304 · PLWP · VPDmax −0.318 − 0.00580 · DOY − 0.444 | −792.71 | −762.91 | 0.796 | 0.166 | 0.165 ± 0.035 |
| 5 | SWP = 24.789 · e2.144 · PLWP · Tmax −0.896 − 0.00543 · DOY − 0.579 | −809.20 | −779.40 | 0.800 | 0.164 | 0.165 ± 0.040 |
SWP, stem water potential; PLWP, predawn leaf water potential; VPDmax, maximum vapor pressure deficit at the day of measurement; Tmax, maximum air temperature at the day of measurement; DOY, day of the year of measurement; AIC, Akaike’s Information Criterion; BIC, Bayesian Information Criterion; RMSE, root-mean-square error (MPa), ± standard deviation.