3 ms·
But they aren't Poisson processes either. (A Poisson processes has an equal probability per unit time, regardless of how long it has been since the last one.) W
by tlb 9y ago
But they aren't Poisson processes either. (A Poisson processes has an equal probability per unit time, regardless of how long it has been since the last one.) With earthquakes, the probability increases over time as strains build up across faults, until they are suddenly released.
- cfadvan 9y agoAdditionally, the longer the strain builds up, the greater the eventual release will be, although it’s possible in some cases for the energy to be released in a series of smaller events. In the case of this fault however, on geological time scales, it’s more or less like clockwork. There is also the possibility of such an event triggering, perhaps st some temporal remove, the release of other faults such as the San Andreas.
- InclinedPlane 9y agoYup, but you do not know whether the strain will be released in a big quake or many smaller quakes. You could be at a state where a lot of strain has built up and a megaquake could seem immanent but then a smaller quake relieves some of the stress and "resets the clock". And then are you still "overdue" or not? The way things work over very long periods of time is that some faults exhibit some degree of periodicity in terms of large quakes. But simplifying that periodicity into the concept of being "overdue" for a big quake is intellectually irresponsible.