Q: I was wondering how we are able to have such accurate predictions of how a P or S wave would travel from a specific point of the crust through the rest of the layers. Do we only monitor the waves when they happen or do we know where they will likely show up before they are picked up?
A: There are a variety of ways that we characterize seismic velocities (P- and S- waves, even surface waves) in the Earth. Earthquakes generate seismic waves, and if we use both an accurate time of the earthquake and observations of the time when seismic waves (from that earthquake) are observed at different locations on earth, it gives us information on the time it took for those waves to travel through the earth – which tells us the velocity of the material they traveled through.
Q: I hear a lot of people say “we are expecting and earthquake” but I’m curious to what degree we can predict where that will happen and when?
A: You have to be careful about vocabulary when you talk about “earthquake prediction” – scientists don’t claim to be able to predict an earthquake on timeframes that would be useful for- for example- evacuating cities. However, we can anticipate where earthquakes are likely to occur in space (usually the big ones are clustered around plate boundaries, or shallow active fault zones), and can estimate ‘recurrence times’ or repeat times for a given fault. That’s why in the Pacific Northwest, we expect there to be a large earthquake on the Cascadia subduction zone roughly every ~500 years (plus or minus about 200 years). It’s been ~400 years since the last one, so you might say we’re in the danger zone. But we can’t predict exactly when it will be, at least not yet.