However, there were also significant differences between the BIC and DT in simulation I-1.
The accuracy of the BIC and DT in simulations with a tree height of 0.1 did not support the general conclusions very well, being different from the simulations using tree heights of 0.3, 0.5, and 0.7.
The estimated numbers of deaths attributable to daily mean, maximum, and minimum temperatures during June August 2020 2037 based on the 32 DT-corrected simulations are presented in Figure 1, along with estimated numbers of deaths attributable to temperatures during 1990 2007 based on observed and DT-corrected simulated temperatures.
All DT-corrected simulations for June August 2020 2037 suggested an increase in daily mean temperatures in Montreal compared with observed temperatures (Table 2): The average daily mean temperature from the 32 future simulations was 20.9°C (range: 20.3 21.3°C), compared with 20.4°C for the observed temperatures for June August 1990 2007.
The time step of the simulation (dt) was initially taken to be 0.04 s.
In fact, there was an even or an approximately even distribution of model categories recovered by the AIC, BIC, and DT in these simulations; but for I-4 and IV-1, models of base + I + Γ category were recovered much less by the BIC and DT than the other model categories.
The time step used in the simulations (dt) was 0.02 s.
The accuracy of the BIC and DT differed among simulations.
The precision of the hLRT was very significantly different from that of BIC and DT in certain simulations (P < 0.01), but not in others, while the precision of AIC was very significantly different from that of the other three criteria (LSD, P < 0.01) in all 14 simulations.
The DT-CP computational simulations provide useful insight into the microscopic activities in nt microstructures, which can be corroborated by experiments, and underscores the importance of discrete twin and TBAZ effects that should be accounted for in developing their corresponding homogenized descriptions.
