Anything is possible, but I’m not saying we can’t or shouldn’t fight it or take steps to prevent it or mitigate damage from it. I’m saying no point in being worried about it.
(sorry there is no reply option for lots of pulp).
Isn't worry the one human emotion that actually makes us ready for things. If we had an earthquake 1000 years ago that killed half of society and the people just said, damn oh well. They don't do things the same way after that do they? They change design, learn more about the planet, adjust adjust adjust. All of that is driven by worry.
Then prediction research became the black death of academic careers with little progress.
This century a milder version called 'earthquake forecasting' has gained acceptance. That is more big data and statistical odds.
Plus 'early warning' seeks to detect quakes and warn people at the speed of light, seconds before damaging waves arrive at the speed of sound.
It seems to me like we've recently (~15-30yrs) had a lot of new technology deployments (and accessibility to the underlying data that's produced) which might considerably help us with gaining an understanding over time. Things like high-resolution, high-precision aerial & satellite radar/lidar/photography and the like.
I tend to think it's just a function of time at this point, but on geological time scales that may still take quite a while I suppose.
1. Hold your hands together, palms open, as if you were praying.
2. Press them together very hard, and at the same time try to slide them past one another.
3. Predict exactly when they will slip, by how much, and how loud the sound will be. To be roughly to scale, it must be with millisecond [0] and millimeter [1] accuracy.
[0] Assuming avg. 100 years between quakes, prediction accurate to a week, avg. 5 seconds before hands slip, then 100 yrs * 52 weeks : 1 week ~ 5,000:1 ~ 5 sec : 1 msec.
[1] A 150 KM fault with a prediction of slip accurate to 0.5 M movement is 300,000:1. For a hand span of 150 mm that would equate to 0.5 um. On the other hand (pun intended), we could compare maximum slippages; assuming an earthquake movement max of 20 M and a hand slippage max of 0.2 M, a real-world resolution of 0.5 M would equate to a hand-slip prediction accurate to 2 mm.