The forecasting problem
Tectonically active zones are the most heavily studied and monitored regions on the planet. Predicting earthquakes and volcanic eruptions has been a scientific goal for a very long time, because even a short warning can give people time to move away from danger. Seismologists at the United States Geological Survey (USGS) publish their own response to a question the public asks again and again, whether earthquakes can be predicted.
Two different kinds of answer are possible, and it helps to separate them from the start.
- A forecast gives the probability that an event of a certain size will happen in a place over a period of time. It comes from statistics about past events.
- A prediction in the strict sense states when, where and how big a particular event will be. It depends on instruments that pick up warning signs before the event.
Geographers, insurers and emergency planners use both. Statistics tell them how often to expect a disaster of a given size. Monitoring tells them whether one may be about to happen. Judging how well each works, and whether either works equally well in every part of the world, is the central question for tectonic hazards.
Statistical ideas such as return periods are easiest to see with floods, because rivers have long and detailed records. The same thinking is applied to earthquakes and eruptions.
9 more sections in this topic
Pick this up in your Library: it holds the whole topic, notes, cheatsheet and questions.