Figure 8.
Plot of probabilities PM( ΔT|TL) given by BPT and Poisson models, for ΔT = 50 yr. For each fault, the probabilities are computed on 106 rupture histories, by randomly choosing μ inside the 90 per cent confidence interval obtained for that fault (listed in Table 3) and, α from 0 to 2, for BPT models. The solid black line marks the median of PM(ΔT|TL) for BPT models, versus α; the grey shadow area shows the respective 90 per cent confidence interval. The solid and dashed red lines mark the median and the 90 per cent confidence bounds, respectively, of PM(ΔT|TL), given by the Poisson model.