The devastating magnitude 7.8 earthquake and its aftershocks that hit Turkey and Syria earlier this week has now claimed substantially more than 20,000 lives.
This reminds me that one reason that major earthquakes tend to cause more deaths than hurricanes is that, even with all the fancy data and science and AI we have in 2023, they remain essentially unpredictable. We’ve made massive strides in recent decades in terms of improving our ability to predict hurricanes or the weather, but almost none in terms of getting to accurate and specific predictions of earthquakes.
In case that seems unlikely, the US Geological Survey themselves do not hold back on answering that question:
Can you predict earthquakes? No. Neither the USGS nor any other scientists have ever predicted a major earthquake. We do not know how, and we do not expect to know how any time in the foreseeable future.
To be fair seismologists do have a pretty specific definition of “predict”. We can certainly say something about rates and magnitudes, for instance that in a particular region of the world there is a 40% probability that we’ll see an earthquake with a magnitude of 6 within the next 10 years. This means that we know that some regions have many more earthquakes than others. But no-one yet can say anything remotely close to “tomorrow there will be a massive earthquake in your city”.
The “within the next 10 years” in the above statement should be thought of as a rate. It doesn’t mean that if you haven’t had an earthquake in 9 years then you’ll probably have one next year. But rather that for any given year you should act as though there is a 10% chance of such an event. Don’t fall into a geological version of the gambler’s fallacy!
Why is it so hard for us to predict earthquakes? In his book “The Signal and the Noise”, Nate Silver dedicates most of a chapter to the topic. Overall, it seems like we just don’t know enough about what causes them and can’t measure every one of their constituent parts at a granular enough level. They can be thought of as a complex system in which fairly simple-to-understand things interact with each other in ways that are essentially mysterious.
Complex systems are systems whose behavior is intrinsically difficult to model due to the dependencies, competitions, relationships, or other types of interactions between their parts or between a given system and its environment
These types of systems exhibit non-linear properties and have steep feedback loops that mean tiny changes to one part of the model can end up wreaking absolute havoc to another of its aspect. These systems tend to remain static for most of the time, but then occasionally fail in a catastrophic way.
As large earthquakes are rare, our inability to do anything beyond forecasting long term probabilities of suffering from a major earthquake is particularly dangerous for countries without access to great wealth. It’s hard or impossible for poorer locales to prioritise taking very expensive precautions against these devastating events when it may be very unlikely that the event will happen in any particular year.
Of course the wealth side of that is a solvable problem should the world decide that protecting the population of its most vulnerable countries is something that should become a moral priority, whether or not we ever come to fully understand the physics of earthquakes.