Exploration of earthquake prediction method

(1) Statistical prediction reflecting time series of seismic activity

In statistical prediction, earthquakes are regarded as random events. That is, the probability distribution of earthquakes per unit time requires independence, sequence, stationarity and Poisson distribution:

Introduction to solid geophysics

Where: f(n) is the probability of n occurrences per unit time; λ is the average number of earthquakes per unit time. For example, as early as 1937, Wenner counted the global earthquakes of 1925 ~ 1930. In order to ensure independence, aftershocks within one month after the earthquake should be removed. There are 2585 independent earthquakes in the range of 2 19 1d, and the average number of earthquakes per day is λ = 2585/2191=1.65438 (times/day). Substituting into equation (6-8), we can get

Introduction to solid geophysics

Multiply the obtained f(n) by the total number of days 2 19 1 to get the number of days that occur n times a day. This is the number of calculation days given in Table 6- 1, and the statistical results of observation data are also given. This table shows that earthquakes can be regarded as random events.

Table 6- 1 seismic frequency distribution table (1925 ~ 1930)

Because earthquake time series can be treated as random events, some statistical laws can be used to predict the future. For example, the extreme value theory is used to predict the medium and long-term earthquakes in China, and the correlation analysis between global earthquakes and Chinese mainland earthquakes is used to study the recurrence period and magnitude distribution of continental earthquakes.

Statistical prediction can only make a rough estimate of a large area and a long time. Because this prediction is given in a certain probability sense, there is no certainty, so it is not uncommon that although the probability is high, the earthquake does not actually happen.

(b) B-value prediction reflecting earthquake scale imbalance

Gutenberg and Li Xite first discovered the exponential attenuation relationship between earthquake magnitude m and earthquake times n(M);

lgn(M)=a—bM

It is not difficult to write the expression of b value:

Introduction to solid geophysics

B represents the magnitude value of the ratio of the number of times of a certain magnitude to the number of times of a higher level earthquake, commonly known as the proportional relationship between large and small earthquakes. If the B value deviates from the normal value seriously, that is, the size ratio is out of balance, when the B value is low, a major earthquake may occur.

Seismologists in China have done a lot of research and data collation on B value. Figure 6- 1 shows the change of B value near the epicenter before the four earthquakes in Haicheng, Longling, Tangshan and Songpan. This is a Zhang Yue chart, which indicates the monthly B value.

Fig. 6- 1 curve of b value change year by year before several major earthquakes

The change of B value, which reflects the proportional relationship between large and small earthquakes in time and space, belongs to the change of energy distribution. Maintaining a certain energy distribution is the inherent requirement and attribute of solid media. The monitoring of B value is an effective means to track whether the energy distribution is unbalanced.

(3) H value prediction reflecting earthquake frequency attenuation

In a short time, in a small area, a group of earthquakes occurred. Taking the largest earthquake as the starting point, the change of earthquake frequency with time after the largest earthquake can generally be written as

n()t=n 1t—h (6- 10)

Mao Mu formula (6- 10) is the aftershock attenuation formula. Liu Zhengrong used this formula to analyze the Yunnan earthquake data of 1979, and obtained that: h≤ 1, the earthquake swarm is foreshock type; H≥ 1, and the earthquake swarm is aftershock type. Combined with the size of b value to predict. Figure 6-2 is an H-B diagram for predicting future earthquakes according to the H and B values of a group of earthquakes. If it falls in area A, a main earthquake larger than the largest earthquake in this group will occur; If it falls in area B, a strong aftershock lower than the largest earthquake in this group will occur; It is relatively safe to fall in area C and D. Of course, the attenuation of aftershocks in different areas is quite different and cannot be regarded as the same.

Figure 6-2 h-b space diagram

(4) Prediction of σ value reflecting the total area of fault plane of earthquake source.

Although seismic frequency and seismic energy can be used to indicate the intensity of seismic activity respectively, because the number of small earthquakes is much larger than that of large earthquakes, the change of frequency actually reflects the activity degree of small earthquakes. Because the energy of a large earthquake is much greater than that of a small earthquake or several small earthquakes, the energy change actually reflects the activity of a large earthquake, so the use of frequency and energy alone is not enough to represent the whole earthquake activity, thus introducing a quantity ∑ that takes both into account, which is defined as

Introduction to solid geophysics

Where N(K) is the time from t to t+? The number of earthquakes with energy level K in the interval T, K=lgE (of course, the so-called energy level K refers to a range from K- to K+). According to the size of L, it can be divided into three situations:

(1)L= 1, degenerating into frequency;

(2)L= 10, degenerating into energy;

(3)L=4.5, which represents the sum of the areas of seismic fault planes with different energy levels.

In order to illustrate the third situation, the following analysis can be made: Because the seismic wave energy E is proportional to the earthquake release energy E0, that is, E∞E0, and E0 is proportional to the focal volume V, or is proportional to the 3/2 power of the focal fault plane S, that is, E0∞S3/2, there is the following relationship:

Introduction to solid geophysics

That is to say, if LK in the formula (6- 1 1) is replaced by 4.5K, then

∑(t)=∑ N(K)S(K) (6- 12)

It shows that ∑(t) is the sum of fault plane areas with different energy levels, which is the physical meaning of ∑(t).

The practical work shows that the peak value of ∑(t) curve in a certain seismic zone often has a certain corresponding relationship with the subsequent strong earthquakes. The larger the peak value, the longer the time interval from the peak value to the strong earthquake, and the greater the magnitude of the earthquake. Relationship. Generally speaking, ∑(t)

(5) Forecast according to the change of wave velocity.

When the local seismic wave passes through the future source area, its propagation speed should change with the change of structure (fault distribution, etc.). ) and physical state (elastic modulus, etc. The focal region of). Generally, the following methods are used for research:

1. Wadachi Kiyoo mapping method for recent earthquakes.

Let K=vP/vS, then there is

Introduction to solid geophysics

Where: refers to the arrival of direct P wave and S wave; O is the time when it happened. For the same earthquake, k and o can be calculated from the sum data recorded by several stations.

2. Apparent velocity method of two teleseismic stations

Also using K=vP/vS, we can get

(S—P)2—(S—P) 1 =(K— 1)(P2—P 1)(6- 14)

Where 1, 2 is the serial number of two workstations. The condition is that the two stations are in the same epicenter direction, and P and S are direct P waves and S waves.

3. teleseismic residual method

When determining the basic parameters of earthquakes, we always get a residual average of P-wave travel time:

Introduction to solid geophysics

Where Ri is the residual of the ith station, and ri = p 'i-Pi (p 'i is the observed value and Pi is the calculated value) is the normal value of the residual.

4. Fixed point blasting method

The blasting point and the receiving point are fixed, and the blasting is carried out regularly to obtain the reliable travel time between the blasting point and the receiving point, and the travel speed is obtained by dividing the distance by the travel time.

The first two are based on the wave velocity ratio K, and the last two are based on the wave velocity vP. In some areas, K-value has been used to predict after earthquakes or to analyze possible precursors. Figure 6-3 is a typical curve of Beijing's K value: it first drops, reaches the lowest point, and then rises. This time is called anomaly duration (t), and the time from the end of anomaly to the occurrence of earthquake is called occurrence delay time (? T). It is generally believed that the duration of the anomaly is related to the magnitude.

Figure 6-Schematic diagram of 6-3K value curve

Using the change of wave velocity to predict earthquakes brings hope to people. However, with the increase of earthquake cases, this method has not completely stood the test. Whether it is the problem of observation accuracy, physical model or both, it is impossible to give a clear conclusion at present.

(six) according to the ground deformation prediction.

Japan is the first country to study the relationship between ground deformation and earthquakes. It is generally believed that there are three stages of earthquake deformation, namely α, β and γ, from * * to final release. α: The crust in the danger zone is abnormal on the basis of long-term slow deformation; β: the crustal deformation speed in the dangerous area increases sharply and changes direction; γ: finally, it breaks and releases a lot of energy in the form of elastic rebound, with the largest amplitude and the opposite direction. Fig. 6-4 is the schematic diagram of its abnormal terrain change.

Figure 6-4 Schematic Diagram of Abnormal Terrain Change

Geodetic survey is generally used for topographic deformation measurement, including triangulation and leveling. Geodetic survey can only find α deformation, which is not suitable for monitoring β deformation. β deformation time is short and changes quickly, and only continuous recording can be effective. Because β deformation is an impending abnormal stage of elastic deformation accumulation from stable process to unstable process, it is particularly noticeable.

Internationally recognized examples of earthquake prediction or earthquake precursors using topographic changes are Niigata earthquake in Japan 1964 (M=7.5) and Tashkent earthquake in the Soviet Union (M=6). Some people think that before the Haicheng earthquake, the short-level deformation data of Jinxian Station showed to some extent.

(7) Forecast according to geomagnetic and geoelectric changes.

Geomagnetic prediction attracts many geophysicists and earthquake prediction workers, because it has reliable scientific basis (piezomagnetic effect) and accurate measuring instruments (theoretically, the additional magnetic field generated by rock compression in the source area can reach 5nT, while the instrument can measure10-1~10-3 nt). For a time, there were many researchers, but they didn't get the expected results. The most famous is the geomagnetic survey with the accuracy of 10-3 nt carried out by Branna of the United States on the San Andreas fault 120km long experimental base. During the test, there were many earthquakes (larger M=4) in and around the test area, but no expected geomagnetic field changes were found. 1980, based on model calculation and some earthquake examples, China put forward the concept that component measurement is equivalent to spatially dense measuring points, and put forward and demonstrated the effectiveness of total intensity measurement in seismomagnetic observation. Under the guidance of this idea, a mobile survey network based on total intensity observation has been set up in key earthquake areas in China.

Because the piezoelectric effect and piezomagnetic effect of rocks are equally obvious. In order to measure the resistivity change of underground rocks, many countries have established a large number of geoelectric observation stations and obtained some data. Some people think that terrestrial light is also formed by piezoelectric effect. Ground light is an abnormal flame that appears in the local sky during an earthquake. There are many historical records and current observations about terrestrial light. For example, before the Haicheng 1975 earthquake in Liaoning, local residents generally saw various colors of ground light, covering a wide range. It is difficult to introduce forecasting practice at present because the internal relationship between geolight and seismoelectricity is not clear.

(eight) according to the groundwater level and groundwater radon change forecast.

As mentioned above, the water content in the source is an important material condition for the occurrence of earthquakes, and radioactive radon in water is an important index to distinguish the medium in the source area from that in the general area.

In view of earthquake prediction, China has systematically observed and studied the groundwater level, which is an important content of monitoring impending earthquake anomalies by group monitoring and group prevention network. The change of groundwater level is generally considered to be caused by crustal deformation. However, season, weather, irrigation, etc. It will also cause water level changes, so it is still difficult to distinguish interference from seismic information. In some areas, some special wells sensitive to earthquakes are specially selected as "earthquake windows" for continuous automatic recording.

The measurement and study of radon in groundwater in China is a part of hydrogeochemistry. In addition to radioactive radon, there are many kinds of ions (nitrate ion, chloride ion, etc. ) can be used as an earthquake precursor to participate in prediction. Moreover, it is found that these hydration precursor reactions do not move outward from the epicenter, but have the characteristics of moving from the periphery to the epicenter in large earthquakes with magnitude above 7 (such as Tangshan earthquake). The Soviet Union discussed in detail the relationship between groundwater radon (especially radon in deep well springs) and earthquakes. Figure 6-5 is a record of radon content in spring water obtained by Tashkent Sanatorium (unit: Bq/m3). The change of chemical composition such as radon in water is caused by the migration of gas in aquifer caused by crustal deformation.

Figure 6-5 Variation curve of radon content in groundwater with time in Tashkent sanatorium, Soviet Union

(9) Comprehensive forecast

The above items belong to a single forecast, which is more favorable when combined. Table 6-2 gives a comprehensive table of precursory anomalies of impending earthquakes. According to the time scale and control range of various precursor reactions, the reliability of data and its development in the process of focal physics, comprehensive medium-and long-term and imminent earthquake prediction is usually carried out, as shown in Figure 6-6. Among them, the seismic situation analysis is based on the seismic activity obtained by the seismic measurement system, which can give the overall change trend. Inspired by this main sign, the geophysical fields in dangerous area and dangerous period are compared in depth. When the earthquake is not imminent, the danger zone is analyzed and the magnitude is estimated by using a wider range of topographic deformation data and geological structure background data. When the earthquake approaches, we should not only continue to monitor the earthquake precursors and geophysical field changes at professional observation points, but also pay attention to macro phenomena (such as animal anomalies, well water level and radon changes in water). ) At the group observation point, the macro phenomenon is more obvious and concentrated at this time, thus determining the epicenter danger zone. The influence of astronomical factors and meteorological factors should also be considered in the whole forecast analysis.

Table 6-2 Comprehensive Table of Several Precursor Anomalies

It should be emphasized that earthquake prediction is a probabilistic, empirical and comprehensive catastrophe prediction. The opposite of probability is certainty. Because the occurrence of earthquakes is a probabilistic event determined by many factors, it is impossible to give the only definite prediction like astronomical phenomena (such as solar eclipse and lunar eclipse), only the possibility of earthquakes can be given. Contrary to experience, it is a theoretical or modeled prediction. The success of Haicheng earthquake prediction in China depends largely on Xingtai earthquake experience, while the success of Songpan earthquake prediction in Sichuan depends to some extent on Haicheng earthquake experience. Because Tangshan earthquake is different from Xingtai and Haicheng, there is no experience to learn from, so it is omitted. In this case, people tend to go to the extreme of ignoring experience and relying on theory unilaterally. In this way, it will also cause serious social impact. The counterexample is 1980. Brady of the United States provided the proportional relationship between experimental micro-fracture and actual fault according to the infrared continuous photos of rock fracture process in the laboratory and the similarity law, and applied it to earthquake prediction in Peru and other parts of South America. It is predicted that there will be an earthquake of magnitude 8 in August 198 1 day. As a result, the earthquake did not happen, but it caused "public panic and social chaos." Therefore, too simple in theory and too one-sided in experience will have serious social consequences. In this case, the forecasting principles adopted should be comprehensive. The synthesis mentioned here includes both the synthesis of various means and the synthesis of experience and theory.

Figure 6-6 Comprehensive Forecast Flowchart

There are only two internationally recognized successful predictions with scientific significance: China Haicheng earthquake M = 1975 and Soviet Pamir earthquake M = 198 1. Obviously, the success rate of earthquake prediction in the world is very low at present, and these two successful predictions are accidental. In other words, at present, the understanding of the essence of earthquakes has not made substantial progress, and it is still in the development stage of accumulating data and experience.