Resource: Earthquake Prediction

Print Background Essay

Earthquakes can have devastating consequences, especially when they strike highly populated areas. Unfortunately, predicting when, where, and with what magnitude a quake will strike is almost impossible. In 1923, with the science of seismology in its infancy, two Japanese scientists squared off about the probability of an earthquake striking the Tokyo area in the immediate future. Their debate illustrates the quandary seismologists faced then and still face today: whether or not to risk alarming citizens with earthquake predictions that are based on limited information. In this case, choosing not to put Tokyo and Yokohama on alert or to strengthen building codes in the cities had tragic consequences.

By 1923, Akitune Imamura, a seismologist at the University of Tokyo, had been predicting a major earthquake in Tokyo for nearly 20 years. His predictions were based on seismic readings and historical records of earthquake events in the area, both of which suggested that Tokyo was long overdue for a major quake. Imamura's superior at the University, Professor Omori, insisted that there was not enough evidence to sound the alarm that an earthquake in the city was imminent. As a result, when the Great Kanto Earthquake struck Tokyo and Yokohama late in the morning of September 1st, 1923, few citizens were prepared. The 7.1-magnitude quake and the fires that erupted in its wake killed approximately 140,000 people in the two cities.

In all fairness, Imamura's prediction was probably based as much on intuition as it was on scientific evidence. Even in the twenty-first century, seismologists remain unable to predict the timing and locations of earthquakes. Instead, they use a variety of indirect measures to estimate the probability that an earthquake of a particular size will strike an area. First, they look at earthquake history. Earthquake-prone areas typically experience quakes at regular intervals. Depending how long it has been since the last quake, they estimate the probability that a future quake will happen in a given time period. Second, they measure the accumulation of strain in the rocks surrounding a fault relative to the amount of strain released in the last quake. This provides a measure of how much more strain the rocks can withstand before another quake. Seismologists also look for precursor events, such as changes in groundwater levels, uplifting or tilting of the landscape, or small tremors called foreshocks. All of these events increase the probability of an earthquake occurring.

To learn more about the instruments scientists use to detect earthquakes, check out Earthquakes: The Seismograph.

To learn more about the effects of earthquakes, check out Earthquakes: San Francisco and Earthquakes: Los Angeles.

To learn more about interactions between tectonic plates, check out Mountain Maker, Earth Shaker and Tectonic Plates, Earthquakes, and Volcanoes.

Print Questions for Discussion

• The two Japanese scientists had differing ideas on earthquake predictions. Whose ideas prevailed and why?
• Scientists are trained to be objective and find the best explanations for the data that they study. They are also human with emotions and prejudices. How is this demonstrated in the story about the two Japanese scientists?
• What can you say about the process of science from viewing this video?
• In the Japanese earthquake, fire caused more death than collapsing buildings. Why?
• The video states that San Francisco has "a 67% probability of a large earthquake." Discuss what this means.