A Proposal to Review How Geophysical Precursors Can Help Predict Earthquakes
Introduction
Throughout the world, devastating earthquakes occur with little or no advance warning. Some of these earthquakes kill hundreds of people. If the times, magnitudes, and locations of these earthquakes could be accurately predicted, many lives could be saved. This document proposes a review of how monitoring geophysical precursors can help in the short-term prediction of earthquakes. The proposed review will discuss the physical principles behind the monitoring of three common precursors and evaluate how accurate each monitoring is in predicting earthquakes. Included in this proposal are my methods for gathering information, a schedule for completing the review, and my qualifications.
Justification of Proposed Review
On the morning of April 18, 1906, the population of San Francisco was awakened by violent shaking and by the roar caused by the writhing and collapsing of buildings [Hodgson, 1964] . The ground appeared to be thrown into waves that twisted railways and broke the pavement into great cracks. Many buildings collapsed, while others were severely damaged. The earthquake caused fires in fifty or more points throughout the city. Fire stations were destroyed, alarms were put out of commission, and water mains were broken. As a result, the fires quickly spread throughout the city and continued for three days. The fires destroyed a 5 square-mile section at the heart of the city [Mileti and Fitzpatrick, 1993] . Even more disastrous was the Kwanto earthquake in Japan that devastated the cities of Yokohama and Tokyo on September 1, 1923 [Hodgson, 1993] . In Yokohama, over 50 percent of the buildings were destroyed [Bolt, 1993] , and as many as 208 fires broke out and spread through the city [Hodgson, 1964] . When the disaster was over, 33,000 people were dead [Bolt, 1993] . In Tokyo, the damage from the earthquake was less, but the resulting fires were more devastating. The fires lasted three days and destroyed 40 percent of the city [Hodgson, 1964] . After the fire, 68,000 people were dead and 1 million people were homeless [Bolt, 1993] .The 1906 San Francisco earthquake and the Kwanto earthquake were two of the most famous and devastating earthquakes of this century. These earthquakes struck without warning and with disastrous results. If earthquakes could be predicted, people would be able to evacuate from buildings, bridges, and overpasses, where most deaths occur.Some earthquakes have been successfully predicted. One of the most famous predictions was the Haicheng Prediction in China. In 1970, Chinese scientists targeted the Liaoning Province as a site with potential for a large earthquake. These scientists felt that an earthquake would occur there in 1974 or 1975. On December 20, 1974, an earthquake warning was issued. Two days later, a magnitude 4.8 earthquake struck the Liaoning Province; however, further monitoring suggested a larger earthquake was imminent [Mileti and others, 1981] . On February 4, 1975, the Chinese issued a warning that an earthquake would strike Haicheng within 24 hours [Bolt, 1993] . The people in Haicheng were evacuated, and about 5.5 hours later, a magnitude 7.3 earthquake shook the city of Haicheng. If the people hadn’t been evacuated, the death toll could have exceeded 100,000.Using geophysical precursors, the Chinese have predicted more than ten earthquakes with magnitudes greater than 5.0 [Meyer, 1977] . For example, the Chinese predicted a pair of earthquakes of magnitude 6.9 that occurred 97 minutes apart in Yunnan on May 19, 1976 [Bolt, 1993] . Despite these successes, the Chinese failed to predict the earthquake that struck the city of Tangshan on July 27, 1976; this earthquake killed 250,000 people and injured 500,000 more [Bolt, 1988] . This earthquake wasn’t completely unexpected, but the Chinese believed it to be a few years away. Other earthquakes have been predicted, but the predictions didn’t have enough precision for warnings to be issued. For example, in 1983, a young geophysicist predicted that an earthquake of magnitude 8 would strike Mexico City within four years [Deshpande, 1987] . Two years later, an earthquake of magnitude 8 did strike Mexico City. Because the prediction was not more precise, no warning was issued and the earthquake took the population of Mexico City by surprise. Other predictions have turned out to be false warnings. For example, an earthquake warning was issued in August 1976 near Hong Kong [Bolt, 1988] . During the earthquake alert, people slept outdoors for two months. No earthquake occurred.
Objectives
I propose to review the available literature on how geophysical precursors can be used for short-term predictions of earthquakes. In this review, I will achieve the following three goals:
- explain three commonly monitored geophysical precursors: ground uplift and tilt, increases in radon emissions, and changes in the electrical resistivity of rocks;
- show what happens to each of these precursors during the five stages of an earthquake; and
- discuss how each of these precursors is used for short-term earthquake predictions.
Geophysical precursors are changes in the physical state of the earth that are precursory to earthquakes. In addition to monitoring geophysical precursors, there are other strategies for predicting earthquakes-in particular, analyzing statistical data on prior earthquakes. Analyzing statistical data on prior earthquakes, however, is solely a long-term prediction technique [Bolt, 1993] . For that reason, I will not consider it.In my review, I will discuss three common geophysical precursors: ground uplift and tilt, increases in radon emissions, and changes in the electrical resistivity of rocks. Earthquakes occur in five stages as there is a build up of elastic strain within faults in the earth, followed by the development of cracks in the rocks, then the influx of water into those cracks. The fourth stage is the actual rupture of the fault and the release of seismic waves. The fifth stage is the sudden drop in stress in the fault. In this stage, aftershocks occur.During these five stages, the geophysical precursors follow distinct patterns. For instance, the ground uplift and tilt increases during the second stage as the volume of rock increases. In my review, I will relate how the three geophysical precursors relate to the five stages of an earthquake and how well this relation can be used to predict the oncoming fault rupture.
Plan of Action
This section presents my plan for obtaining the objectives discussed in the previous section. Because of the recent earthquakes in California and Japan, there has arisen a strong interest to predict earthquakes precisely. As a consequence of that strong interest, many books and journals have been written on earthquakes and earthquake prediction. I have gathered five books and several articles on the subject. In addition, there are dozens of books and articles available in the library. These books and articles should provide sufficient information for me to write my review. The following paragraphs discuss how I will use these sources in my research.The first goal of my research is to explain the physical principles behind monitoring geophysical precursors. For example, why does the electrical resistivity of rocks decrease before an oncoming earthquake? Or, what does a sudden increase in radon emissions reveal about the future likelihood of a massive earthquake? The second goal of my research is to show what happens to each of these precursors during the five stages of an earthquake. To achieve these two goals, I will rely on three books that give an overview to earthquake prediction: Earthquakes [Bolt, 1988] , Earthquakes and Geological Discovery [Bolt, 1993] , and Earthquakes and Earth Structure [Hodgson, 1964] .A third primary goal of the literature review is to cover the accuracy of monitoring each precursor. By accuracy, I mean how well does the method work in predicting the time, place, and size of earthquakes. This discussion will not include many statistics on the predictions of earthquakes, because at present there just haven’t been enough successful predictions to validate these types of statistics. Instead, I intend to evaluate the potential accuracy of monitoring each precursor based on the opinions of experts and preliminary data. To achieve this goal, I will rely on two of my most recent sources: The Great Earthquake Experiment [Mileti and Fitzpatrick, 1993] and Earthquakes and Geological Discovery [Bolt, 1993] .Should I require additional sources other than the ones I have, I will search for them in the library system at the University of Wisconsin. Should I not be able to find that information, I will modify the scope of my research accordingly.Because the primary readers for my proposed literature review are engineering students who are probably not familiar with the theories behind earthquakes, I will have to provide selected background information frommy sources. These engineering students already know that earthquakes are devastating. They also know that if earthquakes could be predicted, people would be able to prepare for them and lives would be saved. However, they may not know the different methods of predicting earthquakes. My intent is to inform these students of three methods of predicting earthquakes.A secondary audience for the review would be non-technical readers who either live in earthquake-prone areas or are affected financially when earthquakes occur. My proposed literature review will provide this group with an unbiased discussion of three methods for earthquake prediction. This discussion, drawing much from overview chapters in Earthquakes, Animals and Man [Deshpande, 1987] and California Quake [Meyer, 1977] , will put into perspective how accurate, or inaccurate, the named methods are and what hurdles face engineers who try to predict earthquakes.
Management Plan
This section presents my schedule, costs, and qualifications for performing the proposed research. The proposed research project culminates in a formal report that will be completed by December 6, 1995. To reach this goal, I will follow the schedule presented in Figure 1. Because I already possess several books and articles on earthquake prediction, most of my time will be spent sifting through the information, finding the key results, and presenting those results to the audience.
Figure 1. Schedule for completion of literature review. The two triangles represent milestones for the project, the first being the formal presentation on November 11, 1996, and the second being the formal report on December 6, 1996.
Given that I can obtain all my sources for the literature review from the library, there is no appreciable cost associated with performing this literature review. The only costs, which will be minor, are for copying articles, printing the review, and spiral binding the review. I estimate that I can do these tasks for under $10.I am a senior in the Geological Engineering Department at the University of Wisconsin at Madison. In my undergraduate courses I have taken rock mechanics, soil mechanics, geophysics, and stratigraphy, all of which have included the principles of seismology and stress-strain relationships. In addition, I have taken field courses on structural geology that have introduced me to subsurface behaviors. I believe that these courses and my hands-on experience will aid me in assimilating the proposed literature review. For further information about my qualifications, see the attached resume (not attached on this web site).
References
Bolt, Bruce A., Earthquakes (New York: W. H. Freeman and Company, 1988).
Bolt, Bruce A., Earthquakes and Geological Discovery (New York: Scientific American Library, 1993).
Deshpande, Prof. B. G., Earthquakes, Animals and Man (Pune, India: The Maharashtra Association for the Cultivation of Science, 1987).
Hodgson, John H., Earthquakes and Earth Structure (Englewood Cliffs, NJ: Prentice-Hall, 1964).
Meyer, Larry L., California Quake (Nashville: Sherbourne Press, 1977).
Mileti, Dennis S., and Colleen Fitzpatrick, The Great Earthquake Experiment(Boulder, Colorado: Westview Press, 1993).
