Earthquake prediction method

Abstract

Provided is an earthquake prediction method that predicts a specific period in which there is a high possibility of the occurrence of an earthquake in a specific region on the earth, on the basis of information pertaining to a prediction target seismic intensity or magnitude.　An earthquake prediction method according to the present invention: narrows a prediction target region; uses, for example, specific coordinates within the field of view in the prediction target region, so as to define, as a range in which an earthquake may occur, a range which includes an error of a regression formula and a gravitational-related numerical value at the time of the occurrence of a past earthquake for each prediction target seismic intensity; and determines that, in a future year, date, and time in which the calculation result obtained by a similar method corresponds to the definition, an earthquake may occur in the prediction target region at the prediction target seismic intensity with the probability indicated in the result.

Description

Earthquake prediction methods

[0001] The present invention relates to an earthquake prediction method.

[0002] The Great East Japan Earthquake that occurred on March 11, 2011, is still fresh in our memories, and the horror of a major earthquake and tsunami is etched in our minds. If such earthquakes could be predicted to some extent, it might be possible to reduce human damage. For this reason, there is a need for an earthquake prediction method that can predict earthquakes of a certain magnitude or larger and warn residents in affected areas before they occur.

[0003] However, even now, when there are concerns about the Nankai Trough earthquake, the Tokai earthquake, and an earthquake directly beneath the capital, earthquake predictions remain limited to vague predictions such as "there is a 70% chance of an earthquake occurring within the next 30 years," and it is difficult to say that earthquake prediction has made any progress.

[0004] There is also a known hypothesis that earthquakes occur primarily when plates bounce up when they continue to move in a certain direction, collide with each other, and the force has nowhere to escape (see, for example, Patent Document 1).

[0005] In addition to inland earthquakes, there are also subduction-zone earthquakes that occur when the lower plate sinks beneath the upper plate, causing the connection between the upper and lower plates to break away.

[0006] There is also a method for predicting earthquakes by analyzing data and magnitude from 1926 to 1986 (see Non-Patent Document 1).

[0007] Japanese Patent Application Publication No. 08-220247

[0008] Shigeo Konno, Junko Ohashi, and Noboru Kojima, "Correlation between Earthquake Occurrence in the Japanese Archipelago and the Positions of the Moon and the Sun," Oyama Technical College Research Bulletin No. 31 (1999), 123-132

[0009] However, it is unclear whether earthquakes are actually caused by plates made of soil and rock bouncing up like springs, as described in Patent Document 1.

[0010] Furthermore, there are two types of seismic waves: P waves and S waves, with P waves traveling faster than S waves. On the other hand, it is the S waves that arrive later that mainly cause damage due to strong shaking. For this reason, earthquake predictions are carried out by taking advantage of the difference in the speed at which seismic waves travel, and alerting people to impending danger when the P waves, which travel first, are detected, before the S waves arrive. However, there is only a few seconds difference in the arrival time of P waves and S waves, which means that this time is insufficient to take measures against earthquakes.

[0011] An object of the present invention is to provide an earthquake prediction method that predicts a specific period during which an earthquake of desired strength (seismic intensity or magnitude) is likely to occur in a specific region on Earth.

[0012] In order to solve the above problem, the earthquake prediction method of the present invention narrows down the prediction target area, and using specific coordinates or the like in the field of view of the prediction target area, defines the range that includes errors in the gravitational force-related values ​​and regression equations from past earthquakes as the range in which an earthquake may occur, based on the predicted seismic intensity. It then determines that there is a possibility that an earthquake may occur in the prediction target area, at the predicted seismic intensity, in the year and date when the calculation results obtained in the same way in the future fall within the above definition.

[0013] According to the present invention, it is possible to provide an earthquake prediction method that predicts a specific period during which an earthquake of desired strength (seismic intensity or magnitude) is likely to occur in a specific region on the earth.

[0014] 1 shows the results of investigating the relative positions of the sun, earth, and moon when an earthquake of seismic intensity 6 or higher occurs in Japan. This figure shows the position of the earth relative to the sun when an earthquake of seismic intensity 6 or higher occurs in Japan, as shown in FIG. 1, is grouped. This figure shows the range of the moon's position relative to the earth when the earth shown in FIG. 1 is moved in the X and Y directions without rotating and aligned in one position. This figure shows the basic concept of the present invention. This figure explains the correlation between the lunar tidal force (X) and the solar tidal force (Y). This figure explains forces on the vertical axis. This figure explains forces on the horizontal axis. This is a flowchart showing the earthquake prediction method of the present invention. This is a graph explaining decision trees 1 and 2. This figure explains decision trees 3, 4, 5, 6, 7, and 8. This shows FIG. 10 expressed in absolute values. This figure explains decision trees 9, 10, 11, and 12. This figure explains decision tree 13. This figure explains decision tree 14. This figure explains decision trees 15 and 16. This is an explanation of decision tree 17 (dedicated to the regression equation for seismic intensity 7). This figure shows, using point cloud data, the correlation between the direction of the moon and the sun at the time of an earthquake of seismic intensity 5 or higher, by seismic intensity. The directional trajectories of the moon and the sun by lunar phase. This figure shows the regression equation for seismic intensity 6-low. This figure explains decision tree 18 (6.5). This figure shows the regression equation for seismic intensity 7. This figure explains decision trees 19 and 20. This figure explains the regression equations for seismic intensity 6+ and seismic intensity 7. This figure explains decision tree 21, and also explains the regression equations Y10, Y11, and Y12 for seismic intensity 6+, and is different from Figure 23. This figure shows how to find the regression equation for seismic intensity 7. This figure explains decision trees 22 to 25, where (a) is decision tree 22, (b) is decision tree 23, (c) is decision tree 24, and (d) is decision tree 25. 1 is a diagram explaining decision trees 26 to 30, where (a) is decision tree 25, (b) is decision tree 27A, (c) is decision tree 28A, (d) is decision tree 29A, and (e) is decision tree 30. FIG. 2 is a diagram explaining decision trees 31 to 34, where (a) is decision tree 31, (b) is decision tree 32, (c) is decision tree 33, and (d) is decision tree 34. FIG. 3 is a diagram explaining decision trees 35 to 38, where (a) is decision tree 35, (b) is decision tree 36, (c) is decision tree 37, and (d) is decision tree 38. FIG. 4 is a diagram explaining decision tree 39.1A and 1B are diagrams illustrating decision trees 40 to 43, where (a) is decision tree 40, (b) is decision tree 41, (c) is decision tree 42, and (d) is decision tree 43. A diagram illustrating decision tree 44. A diagram illustrating decision trees 45 to 48, where (a) is decision tree 45, (b) is decision tree 46, (c) is decision tree 47, and (d) is decision tree 48. A diagram illustrating decision trees 50 and 51, where (a) is decision tree 50 and (b) is decision tree 51. A diagram illustrating decision tree 53. A diagram illustrating decision tree 55. A diagram illustrating decision tree 57. A diagram illustrating decision tree 59. A diagram illustrating decision tree 63. A diagram illustrating decision tree 64. A diagram illustrating decision tree 66. A diagram illustrating decision tree 68. A diagram illustrating decision tree 72. 1 is a diagram explaining decision trees 73 to 76, where (a) is decision tree 73, (b) is decision tree 74, (c) is decision tree 75, and (d) is decision tree 76. A diagram explaining decision tree 77. A diagram explaining decision tree 78. A diagram explaining decision tree 79. A diagram explaining decision tree 80. A diagram explaining decision tree 81. A diagram explaining decision tree 82. A diagram explaining decision tree 83. A diagram explaining decision tree 84. An image of the earthquake prediction process. A table explaining the prediction and results of a seismic intensity 6-. A table explaining the prediction and results of a seismic intensity 6-. A diagram explaining the prediction and results of a seismic intensity 7. A diagram comparing scalars based on different statistical methods for seismic intensity and magnitude. A diagram comparing the gravitational force of the vertical axis based on different statistical methods for seismic intensity and magnitude. A diagram comparing the gravitational force of the horizontal axis based on different statistical methods for seismic intensity and magnitude. 61(a) is a table showing statistics on the number of earthquakes according to the direction of the sun and the moon when they occurred, with Fig. 61(a) being for earthquakes with a seismic intensity of lower 5 or more, and Fig. 61(b) being for earthquakes with a magnitude of 5 or more. Fig. 61(b) is a diagram showing how the present application and Non-Patent Document 1 are divided, with (a) being the present application and (b) being Non-Patent Document 1. Fig. 61(b) is a diagram explaining bar graphs of Non-Patent Documents.

[0015] 1. Basic Concept of the Invention The inventor of the present invention hypothesized that, based on the law of inertia, which states that "an object will maintain a constant motion or remain stationary unless acted upon by an external force," and the existence of tides due to the tidal force of the moon, forces external to the Earth, i.e., the gravitational force and tidal forces of celestial bodies other than the Earth, are related to earthquakes that occur on Earth.

[0016] First, we investigated the relative positions of the Sun, Earth, and Moon when earthquakes of seismic intensity 6 or higher occurred in Japan between January 1, 1995, and December 31, 2022. Figure 1 shows the results, with a line drawn through the Earth's aphelion (July 2) and perihelion (January 2), with the aphelion side designated as the positive direction of the X-axis of the XY coordinate system, and the perihelion (January 2) side designated as the negative direction of the X-axis of the XY coordinate system. Furthermore, with the Sun at the center, July 2 is designated as 0°, and January 2 is designated as 180°.

[0017] This allows the astronomical directions of potentially related items to be converted into mathematical directions (angles), and by using fixed coordinates, trigonometric functions can be used as coefficients, making it possible to compare the strength of forces numerically.

[0018] The symbols attached to the Earth in the figure are explained as follows: 1 is the Earth's position at the time of the Great Hanshin-Awaji Earthquake on January 17, 1995; 2 is the Earth's position at the time of the Tottori Prefecture Western Earthquake on October 6, 2000; 3 is the Earth's position at the time of the Miyagi Prefecture Northern Earthquake on July 26, 2003; 4-6 is the Earth's position at the time of the Niigata Prefecture Chuetsu Earthquake on October 23, 2004; 7 is the Earth's position at the time of the Noto Peninsula Earthquake on March 25, 2007; 8 is the Earth's position at the time of the Niigata-Johchuetsu Offshore Earthquake on July 16, 2007; 9 is the Earth's position at the time of the Iwate-Miyagi Inland Earthquake on June 14, 2008; 10-13 are the Earth's position at the time of the Great East Japan Earthquake on March 11, 12, and 15, 2011; 14 is the Earth's position at the time of the Miyagi Prefecture Offshore Earthquake on April 7, 2011. 15-18 is the position of the Earth during the Kumamoto earthquake on April 14, 15, and 16, 2016. 19 is the position of the Earth during the Hokkaido Eastern Iburi earthquake on September 6, 2018. 20 is the position of the Earth during the Yamagata Prefecture offshore earthquake on June 18, 2019. 21 is the position of the Earth during the Fukushima offshore earthquake on February 13, 2021. 22 is the position of the Earth during the Fukushima offshore earthquake on March 16, 2022.

[0019] The position (direction) of the moon at the time the earthquake occurred is shown on the outer periphery of the Earth at each marked date and time. The black circle indicates the position of the moon when an earthquake of seismic intensity 7 occurred, and the white circle indicates the position of the moon when an earthquake of seismic intensity 6+ occurred.

[0020] (1) The relationship between the Earth's position relative to the sun and earthquakes Figure 2 is a diagram showing groups of the Earth's position relative to the sun when earthquakes of seismic intensity 6 or higher occur in Japan, as shown in Figure 1. As shown in Figure 2, there is a tendency for many earthquakes of seismic intensity 6 or higher to occur in Japan around 0° (group A enclosed by diagonal lines in the figure), around 90° (group B), around 180° (group C), and especially around 270° (group D).

[0021] (2) Relationship between the Position of the Moon Relative to the Earth and Earthquakes Figure 3 is a diagram showing the range of the position of the Moon relative to the Earth when the Earth shown in Figure 1 is moved in the X and Y directions without rotating and aligned in one place. The X1 axis in Figure 3 is parallel to the X axis in Figure 1, and the Y1 axis is parallel to the Y axis in Figure 1. Figure 3 shows that earthquakes with a seismic intensity of 6 or higher that have occurred in Japan tend to occur when the Moon's position is in the + / - direction of the X axis, especially in the -X direction.

[0022] (3) Relationship with celestial bodies in the solar system As mentioned above, when the Earth is in a certain relationship with the Sun and the Moon, there is a tendency for the possibility of earthquakes occurring to be higher. From this, the hypothesis mentioned above that forces external to the Earth, that is, the gravitational force of celestial bodies other than the Earth, are related to earthquakes occurring on Earth, seems to be correct.

[0023] The basic concept of the present invention is shown in Figure 4. In Figure 4, the globe is a butter roll, and the force acting on the globe is a pulling force by hand. For example, it is impossible to tear a butter roll floating in the air by pulling it with a force acting in only one direction, as in Figure 4(a). To tear a butter roll floating in the air, two or more forces equivalent to action and reaction, as in Figure 4(b), are required. When forces acting in multiple directions, as in Figure 4(c), are acting on the butter roll, it is thought that the butter roll will be more likely to tear under certain conditions, rather than being as simple as when two forces are acting on it.

[0024] Therefore, in order to find these specific conditions, we extracted the conditions under which the gravitational forces of the Sun, Moon, Mercury, Venus, Mars, Jupiter, Saturn, and other planets that are thought to have a large gravitational influence on the Earth are more likely to break apart (making earthquakes more likely to occur) from the conditions when past earthquakes have occurred.

[0025] He then hypothesized that when the gravitational force (≒ tidal force) exerted on the Earth by celestial bodies in the solar system, which changes more slowly than tidal forces, decreases, the force of the Earth's gravitational contraction becomes stronger, causing the pressure inside the Earth to increase and become unstable, and when this condition is met with tidal forces that change from moment to moment due to the Earth's rotation, earthquakes are more likely to occur.

[0026] 2. Points to note for understanding this application Figure 5 shows the correlation between the lunar tidal force (X-axis) and the solar tidal force (Y-axis) for each seismic intensity when an earthquake occurs. In Figure 5, since it is tidal force, it is a vector only on the vertical axis of the Earth's surface near the epicenter (Note: The unit is x10^18, which is 1 / 10,000 of the solar gravitational force). The meaning of the graph is that, like gravitational force, it is obvious that tidal force occurs when it is weak (near the origin of the graph), and this tendency is stronger for earthquakes with higher seismic intensity.

[0027] 2-1. Differences between Gravitational Forces and Tidal Forces (Regarding Forces on the Vertical Axis) Figure 6 is a diagram explaining forces on the vertical axis. On the vertical axis shown, A is the gravitational force A from the center of the Earth to the center of celestial body A, and B is the gravitational force B from the Earth's surface to the center of celestial body A. There is a difference in the gravitational forces between gravitational force A from the center of the Earth to the center of celestial body A and gravitational force B from the Earth's surface to the center of celestial body A due to the difference in distance. The gravitational force on the Earth's surface is stronger the closer the object is, so a force acts that pulls it away from the Earth's center, causing phenomena such as high tides. This is tidal force.

[0028] However, in this specification, the state when an earthquake of intensity 6 or higher occurs is limited to when the gravitational and tidal forces of other celestial bodies are weak, as shown in Figure 5, so it can be seen that the force that increases the influence on the Earth's surface on the vertical axis is not tidal force, but the Earth's own gravitational force, that is, "gravity" (a force in the opposite direction to tidal force). (Strictly speaking, it is due to the Earth's own gravity, centrifugal force at different latitudes, and the gravitational forces of other celestial bodies.) Reference: https: / / www.s-yamaga.jp / nanimono / taikitoumi / choseki.html

[0029] (About the force on the horizontal axis) Figure 7 is a diagram explaining the force on the horizontal axis. Typically, the epicenter is overwhelmingly shallower, between 0 and 50 km below the surface of the Earth. In other words, the epicenter can be said to be almost a point on the surface of the Earth, and since there is no difference in distance between calculation points, it can be seen that the force the Earth receives from celestial bodies is not a tidal force but a "gravitational force" (e.g., low tide).

[0030] (Note) However, in the specification, the amount of movement due to an earthquake, i.e., the cross-sectional area and volume of the plate, is unknown before the earthquake occurs, so calculations are made for the entire Earth, and the results of the gravitational force calculation do not directly affect the part of the plate at the epicenter. Therefore, the gravitational force-related numerical values ​​described below are environmental numerical values ​​that apply to the entire Earth, and are considered to correspond to a coefficient of, for example, 100 to 10,000 times the numerical values ​​that actually affect the cross-sectional area of ​​the plate.

[0031] 2-2. Decision Tree (DT) Generally, we hear the word "parameter" when referring to options for making a decision, but in statistics, parameters refer to the coefficients and intercepts of a formula that includes variables, so here we will follow statistics and use the term "decision tree." As an example of a decision tree, if we want to determine the gender of a passerby, we need to consider (1) whether they are wearing makeup, (2) whether they are wearing skinny shoes, (3) whether they are wearing a skirt, etc. Obviously, a single decision tree cannot determine the gender, but if multiple criteria are met, the likelihood of them being female increases. Each of these decision items (1), (2), and (3) is called a decision tree, and if we can find many decision trees with better content, we can make the right decision with a certain degree of probability.

[0032] The earthquake prediction method of the present invention predicts earthquakes using multiple decision trees based on astronomical information showing the relationship between the Earth and other celestial bodies. In this embodiment, decision tree 103 is used. For ease of explanation, the decision trees are appropriately represented as "DT." For example, decision tree 1 is represented as "DT1," and decision tree 2 is represented as "DT2."

[0033] Random Forest is a method of creating many decision trees and creating many decision criteria. In this application, a maximum of 103 trees were created.

[0034] 2-4. Support Vector Machine This is a method of creating boundaries in a decision tree when a cluster of statistical point cloud data can be observed, such as whether the calculation result is less than 10 or more than 10, or whether it is more or less than Y = 3X + 6, and determining one answer as correct and the other as incorrect.

[0035] 2-5. Difference between seismic intensity and magnitude Magnitude is an estimate (integral) of the energy used when the earth's crust moves due to an earthquake, while seismic intensity is the distance traveled in a certain period of time, and indicates acceleration and velocity (differential), so they are not the same thing. In addition, the inaccuracy of magnitude is generally recognized.

[0036] 2-6. Differences in units of lunar and solar power This application makes extensive use of graphs to make them easier to understand visually, but in most cases the X axis is the moon and the Y axis is the sun, and the units for the moon's gravitational force are 10^20 and the sun's gravitational force are 10^22, so please note that the graph's X:Y ratio is 1:100. If the ratio is the same, the result will be a straight line, and the correlation will not be apparent.

[0037] Omission of units The formula for calculating gravitational force F is F=GMm / R^2, gravitational constant: G=6.67・10^-11・m^3・kg^-1・s^-2, celestial body 1: Mkg, celestial body 2: mkg, distance between the two points: Rm, and units will be omitted from here on.

[0038] 3. Overview of the Application This application relates to earthquake prediction, and in particular to the possibility of predicting large-scale earthquakes of seismic intensity 6-low, 6-up, and 7-high in Japan. The prediction algorithm creates a random forest (many decision trees), and for each decision tree, uses a machine learning technique called a support vector machine (boundaries are set, one is correct, the other is incorrect). However, because the number of samples (number of earthquakes) was small, AI was unable to obtain results, so the data was processed manually using spreadsheet software.

[0039] Regarding the contents of the decision tree: The authors were inspired by the law of inertia, which states that "an object will remain at rest or continue in the same motion unless some kind of force is applied," and wondered if the "some kind of force" in question was the force of changes in the gravitational forces of celestial bodies in the solar system. When they calculated the gravitational force-related (≠ tidal force) values ​​for past earthquakes, they were able to identify the following three categories of decision trees (a total of 103 decision trees) when considering the vertical axis (altitude) and horizontal axis (plane = two dimensions) of the surface of the epicenter:

[0040] 3-1. Category 1 Main Decision Tree - Fields related to the gravitational forces of the sun and the moon (Converting astronomical coordinates from the National Astronomical Observatory to mathematical coordinates in hourly units) (1) When the gravitational scalars of celestial bodies in the solar system that affect the Earth, especially the sun and the moon, are 3.5 to 15% smaller than their respective maximum values. (2) When the gravitational forces of the sun and the moon that affect the Earth are decreasing on the vertical axis of the epicenter's surface. (3) When, inevitably, most of the gravitational forces of the sun and the moon that affect the Earth are acting on the horizontal axis.

[0041] 3-2. Category 2 Fields of regression equations from the main decision tree (hourly units) (regression equations only) (4) Correlation between the vertical axis of the moon and the sun (5) Correlation between the direction of the moon and the direction of the sun (6) Correlation between the angle on the Earth's orbit (= position = date) and the direction of the moon as seen from the Earth (age of the moon) (7) Correlation between the interior angles of the moon and the sun and the gravitational force on the horizontal axis From the main decision tree, regression equations of correlation were discovered for items (4) to (7), and it was discovered that the greater the magnitude of the earthquake, the smaller the residuals (errors) and the more regular they were.

[0042] 3-3. Category 3: Fields related to the gravitational forces of solar system bodies (preferably at least daily and hourly units) (A: mathematical coordinates with the Earth as the origin and the Sun fixed at 90°; AbDg: mathematical coordinates with the Sun as the origin, July 2nd at 0°, and January 2nd at 180°) (8) Gravitational forces related to the vertical and horizontal axes of the Sun, Moon, Jupiter, and the planets as a whole (9) Although the causal relationship is unclear, statistical values ​​have boundaries, and data related to gravitational forces, coefficients, and angles to explore the possibility of a correlation with earthquakes (Note: All start with (9), change form, and move from (1) to (8))

[0043] As mentioned above, a total of 103 decision trees were created (the target differs depending on the seismic intensity), and a support vector machine was used to determine whether the event was within the boundary range of past earthquakes of each seismic intensity [1] or not [0]. All years and dates for which all decision trees determined the event to be within the range [1] were extracted and evaluated, and the following probabilities of success were achieved. However, even if all decision trees are missing one or two items, there is still a chance that an earthquake will occur.

[0044] For earthquakes of magnitude 7, the hit rate was 100%, the miss rate was 0%, and the capture rate was 100%. For earthquakes of magnitude 6+ only, the hit rate was 59%, the miss rate was 41%, and the capture rate was 100%. For earthquakes of magnitude 6- only, the hit rate was 33%, the miss rate was 67%, and the capture rate was 100%.

[0045] In other words, if a period of time with similar conditions (1) is extracted from a future period, the algorithm makes it possible to predict earthquakes, including the time, with the above-mentioned probability.

[0046] At that time, they had already successfully performed several test earthquake predictions for the future, and the data showed that "earthquakes occur when the gravitational scalar from solar system bodies acting on the Earth is small, and the vertical axis gravitational force decreases," and "earthquakes are triggered by gravitational contraction caused by the decrease in vertical axis gravitational force, and horizontal gravitational forces represented by the sun and moon, and these numerical values ​​are correlated and can be expressed by various regression equations," which can be said to be the discovery of a new mechanism for earthquake occurrence that differs from the "plate stress limit theory."

[0047] 3-4. Earthquake Prediction These decision trees relate to the direction and magnitude of the gravitational force exerted on the Earth by a given celestial body. The earthquake prediction method of the present invention predicts that an earthquake will occur in a given area if these decision trees are within the earthquake occurrence range of past earthquakes.

[0048] When multiple decision trees are used, the earthquake prediction method of the present invention predicts that an earthquake will occur in a specified area if at least one, preferably more than half, and more preferably all of the multiple decision trees are within the earthquake occurrence range.Incidentally, the above-mentioned probabilities were 49 or more for a seismic intensity of 7, 97 or more for a seismic intensity of 6+, and 89 or more for a seismic intensity of 6-.

[0049] The more decision trees among the multiple decision trees used for earthquake occurrence prediction, the better, as this improves the accuracy of earthquake prediction. In the embodiment, earthquakes are predicted using all of the multiple decision trees described below. However, this is not limited to this, and some of the decision trees may be excluded. In addition, decision trees for new celestial body information, such as those related to tidal forces, may be added.

[0050] 3-5. Earthquake Prediction Method The flow of the earthquake prediction method of the present invention will now be described. Figure 8 is a flowchart showing the earthquake prediction method of the present invention. The earthquake prediction method of the present invention includes an astronomical information collection step S1, a decision tree acquisition step S2, an earthquake range identification step S3, and an earthquake prediction step S4.

[0051] (Astronomical Information Collection Step S1) (Setting of Earthquake Prediction Target Area) In the astronomical information collection step S1, the earthquake prediction target area is first set. In this embodiment, the earthquake prediction target area is Japan. Setting the earthquake prediction target area determines the earthquake prediction target area for which earthquakes are predicted. The earthquake prediction target area can be any region on Earth, and may be, for example, Japan only, or the Asian region (multiple countries such as Indonesia and the Philippines), and can be determined at the forecaster's discretion, such as a specified latitude and longitude range or limited to the West Coast of the United States. However, if the range of the earthquake prediction target area is expanded too much, there may be too much information and it may become difficult to make a prediction, or if the earthquake prediction target area is narrowed too much, there may be too little information and the accuracy may decrease, so an appropriate range must be determined while observing the results.

[0052] (Setting the data collection period) Next, the period for collecting data for earthquake prediction is set. The period for collecting data for earthquake prediction is a specified period in the past (past data collection period) in the earthquake prediction target area and a future target period for earthquake prediction (prediction period). The past data collection period and prediction period can be set at will by the researcher and forecaster, but are usually limited to the period during which earthquake data for that area and astronomical information for that period can be obtained. However, it is important to note that the open data that can be collected may differ depending on the source, and may be corrected at a later date even for the same site. There are surprisingly many different data and corrections, so it may be necessary to check every other year.

[0053] (Collection of Astronomical Information) Then, astronomical information is collected for both the past data collection period and the prediction period for the earthquake prediction target area. In this embodiment, astronomical information is collected for Japan, which is the earthquake prediction target area. The astronomical information includes, for example, the gravitational force, distance, position, angle, mass, etc. of each celestial body, but may also include other characteristics related to the celestial body, such as the state of the celestial body, such as sunspots, tides, gravitational waves, etc. Furthermore, the celestial body information may also include atmospheric pressure on Earth and the Coriolis force.

[0054] (Celestial bodies from which astronomical information is acquired) In the present embodiment, the celestial bodies used for prediction are the Earth, the Sun, the Moon, Mercury, Venus, Mars, Jupiter, and Saturn. However, the present embodiment is not limited to these and may include other celestial bodies. Furthermore, other celestial bodies may be added to the celestial bodies used for prediction in the present embodiment, or some celestial bodies may be removed from the celestial bodies used for prediction in the present embodiment. Furthermore, the other celestial bodies are not limited to planets in the solar system, but may also be celestial bodies belonging to a galaxy other than the solar system or extragalactic bodies, as long as they may have an impact on the Earth. Furthermore, the other celestial bodies are not limited to planets, but may also be nebulae, stars, dwarf planets, satellites, comets, black holes, etc. Furthermore, the other celestial bodies may be fictitious celestial bodies (celestial body X) that may exist far away on the x-axis in geocentric coordinates. The fictitious celestial body (celestial body X) may also be a black hole. In other words, in order to improve the accuracy rate, the number of decision trees may continue to increase.

[0055] In addition, in this embodiment, astronomical information is shown using x-y coordinates (geocentric coordinates) in which the Earth is at the center and the Sun is fixed at a 90° angle, or x-y coordinates (heliocentric coordinates) centered on the Sun. However, this is not limited to this. Astronomical information may also be shown without using coordinates. Furthermore, the center of coordinates is not limited to the Earth or the Sun, but may be the center of the galaxy, a black hole, or the like. Furthermore, in the embodiment, in geocentric coordinates, the x-y axes are lines passing through the perihelion and aphelion, but this is not limited to this, and the x-y axes may be directions passing through the winter solstice, vernal equinox, summer solstice, and autumnal equinox.

[0056] (Target period for obtaining astronomical information) The time unit for obtaining astronomical information is at least one day for gravitational force, but a shorter time of one hour is preferable, such as 10 minutes or one minute. Also, for tidal force, an hour or less is preferable, but a shorter time of 10 minutes or one minute is also acceptable. The period for predicting earthquake occurrence depends on the unit of the collected astronomical information. For example, if celestial information is collected in units of hours or minutes, earthquakes can also be predicted in units of hours or minutes. This information is stored on a storage medium. A computer is preferable as the storage medium, but documents, etc. can also be used. Note that accuracy is improved by setting the survey unit to one hour. However, this slows down the PC and takes time, so a high-performance PC is preferable.

[0057] (Sources of astronomical information) The preferred sources of astronomical information are the Japan Meteorological Agency in Japan and the USGS in the United States outside of Japan, but other specialized organizations are also acceptable. However, when obtaining astronomical information, attention must be paid to differences in the calculation methods for time of occurrence and magnitude due to time differences. The accuracy of astronomical information may vary depending on the source. For example, in the case of the age of the moon, the values ​​given on the Internet may vary slightly depending on the source, and there may be discrepancies between predicted values ​​and actual observed values ​​for the path of celestial bodies, etc. Therefore, a certain degree of ambiguity is acceptable.

[0058] (Decision Tree Acquisition Step S2) Based on the above astronomical information, 103 decision trees were obtained either directly from the astronomical information or by calculation from the astronomical information. The various decision trees will be described later.

[0059] (Earthquake range identification step S3) (Setting the magnitude of the earthquake to be predicted) In the earthquake range identification step S3, first, the date and time of an earthquake of a predetermined magnitude that occurred in the past data collection period in the earthquake prediction target area is identified. The criteria for determining an earthquake of a predetermined magnitude may be a single criteria such as seismic intensity or magnitude, or a combination of multiple criteria.

[0060] (Identifying the earthquake occurrence range) Next, the numerical range of each decision tree (individual earthquake occurrence range) is identified at the date and time when an earthquake of a predetermined magnitude or greater actually occurred within the past data collection period. Here, if the maximum value of the decision tree at the time of the earthquake occurrence is 5.8, the earthquake occurrence range may be any range selected by the forecaster, such as 5.8 or less, 5.81 or less, or 6.0 or less, as long as it includes the range where the earthquake actually occurred.

[0061] In addition, in the embodiment, the maximum value of the astronomical information when an earthquake occurs is calculated, and the range from 0 to a value slightly larger than the maximum value by a value α is set as the earthquake occurrence range. However, this is not limited to this, and the correlation between the minimum or average value of the range of astronomical information when an earthquake occurs and the occurrence of an earthquake may also be used. Furthermore, the correlation coefficient between the angle on the coordinates of a specified celestial body or some state of a specified celestial body (for example, the number of sunspots on the sun) and the occurrence of an earthquake may also be used. Furthermore, the value α can be changed each time a new earthquake occurs, etc.

[0062] In this embodiment, the earthquake occurrence range during the past data collection period and the entire range during the past data collection period were calculated for each decision tree. At this time, the decision tree when an earthquake of a predetermined seismic intensity or greater occurred in Japan was determined as the earthquake occurrence range.

[0063] The hit rate etc. in this embodiment are calculated as follows: Details will be explained from

[0248] onwards When an earthquake is predicted to occur and does occur, it is abbreviated as TP When an earthquake is predicted to occur and does not occur, it is abbreviated as FN When an earthquake is predicted not to occur and does not occur, it is abbreviated as TN When an earthquake is predicted not to occur and does occur, it is abbreviated as FP TP + FN + TN + FP is abbreviated as ALL Overall hit rate Accuracy: (TP + TN) / ALL x 100 Hit rate for earthquake occurrence Recall: TP / (TP + FN) x 100 Capture rate Precision: TP / (TP + FP) x 100

[0064] Note that a seismic intensity of 6+ is just one example; it is also possible for seismic intensity 7 or M7, and although the accuracy rate is significantly lower, it is also possible for weaker earthquakes such as a seismic intensity of 5+.

[0065] Earthquakes may occur several times in a single day, or may occur several days apart. In this embodiment, if the time interval between earthquake occurrence dates and times is within 30 days, which is one lunar cycle, multiple earthquake occurrences are treated as one period. For example, if earthquakes occur on March 1st and March 31st, they are considered to be the same earthquake and are treated as one period, while if earthquakes occur on March 1st and April 1st, they are considered to be different earthquakes and are treated as two periods.

[0066] (Earthquake Prediction Step S4) (Prediction of Future Earthquake Occurrence) Next, from the entire prediction period for which prediction is performed, in this embodiment, all decision trees from DT1 to DT103 or (Option 1: for each seismic intensity specified in Figure 64; Option 2: for the decision tree required for each seismic intensity) are selected to extract the period that falls within the earthquake occurrence range, and a specific future date is obtained, which becomes the predicted date for the possible occurrence of an earthquake. This period corresponds to the past TP + FN, and the probability of a possible earthquake occurrence is estimated as the earthquake occurrence hit rate: TP / (TP + FN) x 100. However, the hit / miss results are updated after the date has passed. Furthermore, while it is preferable for all decision trees from DT1 to DT103 and 58P to fall within the earthquake occurrence range, the number of decision trees can be determined at the discretion of the forecaster, for example, to expand the prediction period to ensure reliable predictions.

[0067] In this embodiment, from the entire prediction period for which predictions are made, a period is selected in which one or more decision trees required for each seismic intensity do not fall within the earthquake occurrence range. This period corresponds to the no-earthquake prediction period, which is the past hit rate for no-earthquake occurrences: TN / (TN+FP) x 100, but the hit / miss results are updated after the deadline, so the hit rate will be revised. Note that while it is preferable to have one or more decision trees that do not fall within the earthquake occurrence range, the number can be determined at the discretion of the forecaster, for purposes such as extending the prediction deadline to ensure reliable predictions.

[0068] If the above excerpts are made on a computer, it is preferable to extract the excerpts for the relevant period in various languages ​​such as Excel or Phython. Also, astronomical information on celestial bodies used for prediction during the past data collection period and the prediction period can be used as a data set for AI machine learning or deep learning, and if there are a large number of samples (number of earthquake occurrences), it may be possible to extract the data using AI, but in this embodiment, the excerpts were made using Excel.

[0069] 4. Explanation of decision trees The contents of each decision tree, calculation results, correlation equations, etc. will be explained below.

[0070] 4-1 Category 1 Main Decision Tree: Areas related to the gravitational forces of the sun and the moon

[0071] [Decision Trees 1 and 2] (DT1 and DT2) Correlation of lunar and solar gravitational scalars by seismic intensity (earthquakes are unrelated in all periods) FIG. 9 is a graph explaining decision trees 1 and 2.

[0072] <Meaning of the graph> The graph in Figure 9 shows the lunar gravitational scalar (X axis) and solar gravitational scalar (Y axis) for each seismic intensity when earthquakes occurred between January 1, 1995 and September 2024, and the gravitational scalar for the entire period is unrelated to the presence or absence of earthquakes. (Example: Seismic Intensity 6.5: Seismic Intensity 6+ for Japan)

[0073] <Characteristics> At a seismic intensity of 6+, the range of the lunar gravitational scalar (X-axis) is approximately 30% smaller in area than normal, and at a seismic intensity of 7, the range of the solar gravitational scalar (Y-axis) is approximately 20% smaller than normal. Also, the smaller the seismic intensity, the more similar it becomes to the scalar over the entire period. Also, the solar gravitational scalar is approximately 200 times the lunar gravitational scalar, and it can be said that the smaller the scalar, the stronger the earthquake shaking. It is important to note that this is not the case as is often assumed: "The larger the scalar, the stronger the earthquake."

[0074] <Numerical ranges for the Decision Tree for prediction> 100% of earthquakes of each seismic intensity are within the specific ranges below, and are defined as having a high probability of occurring under the same conditions in the future. Because there are few samples, it is wise to leave a slight margin for the numerical values ​​when using them for prediction, but if the margin is too large, predictions will become impossible, so finding the right balance is an issue.

[0075] [Decision tree 1] DT1 Coordinates: Centers of celestial bodies The lunar gravitational scalar range when a seismic intensity of 7 occurs is 1.79x10^20 or more and 2.23x10^20 or less The lunar gravitational scalar range when a seismic intensity of 6+ occurs is 1.79x10^20 or more and 2.12x10^20 or less The lunar gravitational scalar range when a seismic intensity of 6- occurs is 1.76x10^20 or more and 2.29x10^20 or less

[0076] [Decision tree 2] DT2 Coordinates: Centers of celestial bodies The gravitational scalar range of the sun when a seismic intensity of 7 occurs is 3.66x10^22 or more and 4.05x10^22 or less The gravitational scalar range of the sun when a seismic intensity of 6+ occurs is 3.72x10^22 or more and 4.19x10^22 or less The gravitational scalar range of the sun when a seismic intensity of 6- occurs is 3.66x10^22 or more and 4.20x10^22 or less

[0077] [Decision tree 3, 4, 5, 6, 7, 8] (DT3, 4, 5, 6, 7, 8) Sum of gravitational vectors of solar system bodies (Sun, Moon, Mercury to Saturn) by seismic intensity (X axis: horizontal axis, Y axis: vertical axis) *The horizontal axis is directionless

[0078] <Meaning of the graph> Figure 10 is a diagram explaining decision trees 3, 4, 5, 6, 7, and 8. The graph in Figure 10 shows the horizontal gravitational vector (X-axis) and vertical gravitational vector (Y-axis) of solar system bodies (the Sun, Moon, Mercury, Venus, Mars, Jupiter, and Saturn) affected by seismic intensity on Earth when an earthquake occurs, based on Tokyo. Both the X-axis and Y-axis are in units of 10^22, so both are approximately the gravitational force of the sun. Therefore, top dead center is solar noon (just before midnight), bottom dead center is just before midnight, the right end is around sunrise, and the left end is around sunset.

[0079] <Characteristics> The greater the seismic intensity, the smaller the gravitational vector on the vertical axis, and the greater the seismic intensity, the larger the gravitational vector on the horizontal axis.

[0080] Additionally, no earthquakes of magnitude 6 or higher have occurred in the circled areas, and even for magnitudes of 5 or higher, there are fewer earthquakes. In other words, when the sun is positioned between the point directly behind the earth and sunrise (which we will call the peculiar time period), there tends to be fewer earthquakes, or no earthquakes at all. We speculate that the reason for this is that the circled peculiar time period is when the following (1) and (2) overlap.

[0081] (1) As the sun's gravitational force moves in the negative direction (from the left edge to the bottom dead center), the strength of gravity tends to increase relative to the Earth's surface at the epicenter, so based on the gravitational contraction hypothesis, it can be assumed that this will promote the occurrence of earthquakes. On the other hand, as the sun's gravitational force moves in the positive direction (from the bottom dead center to the right edge), the strength of gravity tends to decrease relative to the Earth's surface, so it can be assumed that this will suppress the occurrence of earthquakes. In this case, this can be explained by the fact that peculiar time periods are times when earthquakes are suppressed. As an analogy, it can be thought of as being the same as "when you stretch a rubber band until it breaks, the moment it breaks is always while it is being stretched, and it never breaks while it is being contracted." However, an explanation is needed as to why peculiar time periods disappear after 3:00 AM.

[0082] (2) The reason why the peculiar time period disappears after 3:00 AM is that it begins about halfway down the hypotenuse of the fourth quadrant. This could be explained as a bifurcation point where the horizontal axis gravitational vector, one of the key points that causes earthquakes, becomes larger in trigonometric function terms than the cosine (horizontal axis force) force (vertical axis force).

[0083] <Numerical ranges for the Decision Tree for prediction> 100% of earthquakes of each seismic intensity are included within the specific ranges below, and it can be said that there is a high probability of an earthquake occurring under the same conditions in the future. When using the values ​​for prediction, it is wise to leave a slight margin of error, but if the margin is too large, predictions will become impossible, so finding the right balance is key.

[0084] [Decision tree 3] DT3 Coordinates: Centers of celestial bodies The range of the vertical axis plus gravitational vector of all solar system celestial bodies when a seismic intensity of 7 occurs is 1.63 x 10^22 or less The range of the vertical axis plus gravitational vector of all solar system celestial bodies when a seismic intensity of 6+ occurs is 3.32 x 10^22 or less The range of the vertical axis plus gravitational vector of all solar system celestial bodies when a seismic intensity of 6- occurs is 3.5 x 10^22 or less

[0085] [Decision tree 4] DT4 Coordinates: Centers of celestial bodies The range of the negative vertical axis gravitational vector of the total solar system celestial bodies when a seismic intensity of 7 occurs is 1.7 x 10^22 or less The range of the negative vertical axis gravitational vector of the total solar system celestial bodies when a seismic intensity of 6+ occurs is 3.04 x 10^22 or less The range of the positive vertical axis gravitational vector of the total solar system celestial bodies when a seismic intensity of 6- occurs is 3.38 x 10^22 or less

[0086] [Decision Tree 5] DT5 Coordinates: Centers of celestial bodies When a seismic intensity of 7 occurs, the range of the horizontal axis plus the gravitational vector of the total solar system celestial bodies is 5.4 x 10^19 or more When a seismic intensity of 6+ occurs, the range of the horizontal axis plus the gravitational vector of the total solar system celestial bodies is 2.53 x 10^19 or more When a seismic intensity of 6- occurs, the range of the horizontal axis plus the gravitational vector of the total solar system celestial bodies is 2.1 x 10^19 or more

[0087] [Decision tree 6] DT6 Coordinates: Centers of celestial bodies The range of the horizontal axis minus gravity vector of the total solar system celestial bodies when a seismic intensity of 7 occurs is 1.6 x 10^20 or more The range of the horizontal axis minus gravity vector of the total solar system celestial bodies when a seismic intensity of 6+ occurs is 1.52 x 10^20 or more The range of the horizontal axis minus gravity vector of the total solar system celestial bodies when a seismic intensity of 6- occurs is 1.29 x 10^19 or more

[0088] [Decision Tree 7] DT7 Coordinates: Center to center of celestial bodies The total scalar of the vertical and horizontal axes of solar system celestial bodies when a seismic intensity of 7 occurs is 4.08 x 10^22 or less The total scalar of the vertical and horizontal axes of solar system celestial bodies when a seismic intensity of 6+ occurs is 4.21 x 10^22 or less The total scalar of the vertical and horizontal axes of solar system celestial bodies when a seismic intensity of 6- occurs is 4.22 x 10^22 or less

[0089] [Decision Tree 8] DT8 Coordinates: Centers of celestial bodies. Seismic intensity 7, seismic intensity 6+, seismic intensity 6-. When these two events occur together, the gravitational vector range of solar system celestial bodies is as follows: - Vertical axis gravitational vector of solar system celestial bodies is 0.0 or less, set to -A. - Horizontal axis gravitational vector of solar system celestial bodies is 0.0 or more, 2.0x10^22 or less, set to -B. When these events occur together, the range outside A∩B (≒outside the circle) is

[0090] Because this information overlaps with DT3, 4, 5, 6, 7, and 8, the total absolute values ​​(X, Y) of the gravitational vectors of the solar system planets, the sun, and the moon by seismic intensity are shown in Figure 11 for reference.

[0091] <Meaning of the Graph> FIG. 11 shows FIG. 10 expressed in absolute values.

[0092] <Characteristics> The smaller the gravitational vertical vector absolute value (scalar) (Y), the stronger the earthquake shaking. Earthquakes of magnitude 7 only occur when the gravitational horizontal vector absolute value (scalar) (X) is large, while earthquakes of magnitude 5+ or less have the characteristic that the gravitational horizontal vector absolute value (scalar) occurs across the entire area.

[0093] [Decision Trees 9, 10, 11, 12] DT9, 10, 11, 12 By seismic intensity, absolute value of the lunar vertical axis gravitational force vector (X), absolute value of the solar vertical axis gravitational force vector (Y) FIG. 12 is a diagram for explaining decision trees 9, 10, 11, 12.

[0094] <Meaning of the graph> This graph shows the absolute value of the vertical axis gravitational vector of the moon (X) and the absolute value of the vertical axis gravitational vector of the sun (Y) for each seismic intensity.

[0095] <Characteristics> Earthquakes tend to occur more easily when the gravitational force on the vertical axis is weak, not only for earthquakes with strong shaking, but also for earthquakes with weak shaking on the vertical axis of the epicenter. This tendency is particularly noticeable for earthquakes with a seismic intensity of 7.

[0096] <Numerical range of the Decision Tree for prediction> 100% of earthquakes with the target seismic intensity are within the specific range below, and it can be said that there is a high probability of them occurring under the same conditions in the future. When using the numbers for prediction, it is wise to leave a slight margin of error, but if the margin is too large, it will become impossible to make predictions, so it is important to find the right balance.

[0097] [Decision Tree 9] DT9 Coordinates: Centers of celestial bodies When a magnitude 7 earthquake occurs, the range of the moon's vertical axis plus gravitational vector is 1.34 x 10^20 or less When a magnitude 6+ earthquake occurs, the range of the moon's vertical axis plus gravitational vector is 1.67 x 10^20 or less When a magnitude 6- earthquake occurs, the range of the moon's vertical axis plus gravitational vector is 1.82 x 10^20 or less

[0098] [Decision tree 10] DT10 Coordinates: Centers of celestial bodies The range of the moon's vertical axis minus gravitational vector when a magnitude 7 earthquake occurs is 8.6 x 10^19 or less The range of the moon's vertical axis minus gravitational vector when a magnitude 6+ earthquake occurs is 1.61 x 10^20 or less The range of the moon's vertical axis minus gravitational vector when a magnitude 6- earthquake occurs is 1.66 x 10^20 or less

[0099] [Decision tree 11] DT11 Coordinates: Centers of celestial bodies When a magnitude 7 earthquake occurs, the range of the sun's vertical axis plus gravitational vector is 1.7 x 10^22 or less When a magnitude 6+ earthquake occurs, the range of the sun's vertical axis plus gravitational vector is 3.31 x 10^22 or less When a magnitude 6- earthquake occurs, the range of the sun's vertical axis plus gravitational vector is 3.50 x 10^22 or less

[0100] [Decision tree 12] DT12 Coordinates: Centers of celestial bodies When a magnitude 7 earthquake occurs, the range of the sun's vertical axis minus gravitational vector is 1.69 x 10^22 or less When a magnitude 6+ earthquake occurs, the range of the sun's vertical axis minus gravitational vector is 3.03 x 10^22 or less When a magnitude 6- earthquake occurs, the range of the sun's vertical axis minus gravitational vector is 3.37 x 10^22 or less

[0101] [Decision Tree 13] DT13 Seismic Intensity, Moon's Horizontal Axis Gravitational Vector (X), Sun's Horizontal Axis Gravitational Vector (Y) FIG. 13 is a diagram for explaining the decision tree 13.

[0102] <Meaning of the graph> Figure 15 is a graph showing the correlation between the horizontal axis gravitational force vector of the moon (X) and the horizontal axis gravitational force vector of the sun (Y) for each seismic intensity.

[0103] <Characteristics> When an earthquake of magnitude 7 occurs, it occurs when the absolute value of the sun's horizontal axis gravitational vector (Y) is large, and no earthquakes occur within the wide area enclosed by the rectangle near the origin. When an earthquake of magnitude 6+ occurs, the sun's horizontal axis gravitational vector (Y) is slightly large, and although not as large as an earthquake of magnitude 7, no earthquakes occur within the area enclosed by the rectangle near the origin. When an earthquake of magnitude 6- occurs, this tends to occur when the moon and sun's horizontal axis gravitational vector (Y) is slightly large. Although not as large as an earthquake of magnitude 6-, no earthquakes occur within the area enclosed by the rectangle with the origin. Looking at all earthquakes of magnitude 5- or larger, it is common that the gravitational forces (absolute values) of the moon and sun are both large on the horizontal axis, and the larger the earthquake, the stronger this tendency becomes. (For earthquake prediction, we are not looking for the number of occurrences, but rather for "no earthquakes occurring even once under those conditions.")

[0104] <Numerical ranges for the Decision Tree for prediction> 100% of earthquakes of each seismic intensity are included within the specific ranges below, and it can be said that there is a high probability of an earthquake occurring under the same conditions in the future. When using the values ​​for prediction, it is wise to leave a slight margin of error, but if the margin is too large, predictions will become impossible, so finding the right balance is key.

[0105] [Decision Tree 13A(7)] DT13A(7) (Coordinates are NAOJ astronomical coordinates → mathematical coordinates) When a magnitude 7 earthquake occurs, the range of the moon's horizontal axis plus gravitational vector is 1.7x10^20 or more, and when a magnitude 7 earthquake occurs, the range of the moon's horizontal axis minus gravitational vector is 1.7x10^20 or more, and when a magnitude 7 earthquake occurs, the range of the sun's horizontal axis plus gravitational vector is 2.0x10^22 or more, and when a magnitude 7 earthquake occurs, the range of the sun's horizontal axis minus gravitational vector is 2.0x10^22 or more (outside the box)

[0106] [Decision Tree 13B (6.5)] DT13B (6.5) Coordinates are National Astronomical Observatory astronomical coordinates → mathematical coordinates. When a seismic intensity of 6+ occurs, the range of the moon's horizontal axis plus gravitational vector is 8.0x10^19 or more. And when a seismic intensity of 6+ occurs, the range of the moon's horizontal axis minus gravitational vector is 8.0x10^19 or more. And when a seismic intensity of 6+ occurs, the range of the sun's horizontal axis plus gravitational vector is 9.0x10^21 or more. And when a seismic intensity of 6+ occurs, the range of the sun's horizontal axis minus gravitational vector is 9.0x10^21 or more (outside the box).

[0107] [Decision Tree 13C (6.0)] DT13C (6.0) Coordinates are National Astronomical Observatory astronomical coordinates → mathematical coordinates The range of the moon's horizontal axis plus gravitational vector when a seismic intensity of 6 lower occurs is 1.0x10^20 or more And the range of the moon's horizontal axis minus gravitational vector when a seismic intensity of 6 lower occurs is 1.0x10^20 or more And the range of the sun's horizontal axis plus gravitational vector when a seismic intensity of 6 lower occurs is 2.5x10^21 or more And the range of the sun's horizontal axis gravitational vector when a seismic intensity of 6 lower occurs is 2.5x10^21 or more (outside the box)

[0108] [Decision tree 14] DT14 Seismic intensity, absolute value of the moon's horizontal axis vector (X), absolute value of the sun's horizontal axis gravitational force vector (Y)

[0109] <Meaning of the graph> Figure 14 is a diagram for explaining the decision tree 14, and shows the graph in Figure 13 as absolute values. The features are easy to understand.

[0110] <Characteristics> The cluster (area circled) occurs above the diagonal line shown in the diagram below.

[0111] <Numerical range of the prediction decision tree>

[0112] [Decision Tree 14] DT14 Coordinates are astronomical coordinates of the National Astronomical Observatory → mathematical coordinates. The range of the horizontal axis gravitational force vector of the moon or sun when a seismic intensity of 7 occurs (scalar). When the horizontal axis vector of the sun when the earthquake occurs is Yh7 and the boundary formula DT14(7) = 2.0 x 10^22, Yh7 > boundary formula DT14(7)

[0113] The range of the horizontal axis gravitational vector of the moon or the sun when a seismic intensity of 6 or higher occurs. When the horizontal axis vector of the sun at the time of the earthquake is Yh6.5 and the horizontal axis vector of the moon is X, the calculation result of the boundary formula DT14(6.5) = 133.33 x X - 6.66 x 10^21 is Yh6.5 > boundary formula DT14(6.5).

[0114] The range of the gravitational vector of the moon or sun on the horizontal axis when a seismic intensity of 6-low occurs. When the horizontal axis vector of the sun at the time of the earthquake is Yh6 and the horizontal axis vector of the moon is X, the calculation result of the boundary formula DT14(6) = 166.66 x (X) - 183.32 x 10^22 is Yh6 > DT14(6).

[0115] [Decision Trees 15 and 16] DT15 and DT16 Correlation between Moon Altitude and Sun Altitude by Seismic Intensity FIG. 15 is a diagram for explaining decision trees 15 and 16.

[0116] <Meaning of the graph> This shows the altitude of the moon (X) and sun (Y) at the time of the earthquake, depending on the seismic intensity.

[0117] <Characteristics> Based on past trends, it can be assumed that the stronger the shaking of an earthquake, the lower its altitude (if the COS is large, the SIN is small), and this was indeed the result.

[0118] <Numerical range of the prediction decision tree>

[0119] [Decision tree 15] DT15 Coordinates are astronomical coordinates of the National Astronomical Observatory → mathematical coordinates. The range of the moon's altitude when a seismic intensity of 7 occurs is -60° to 60°. The range of the moon's altitude when a seismic intensity of 6+ occurs is -72° to 72°. The range of the moon's altitude when a seismic intensity of 6- occurs is -75° to 75°.

[0120] [Decision tree 16] DT16 Coordinates are astronomical coordinates of the National Astronomical Observatory → mathematical coordinates. The range of the sun's altitude when a seismic intensity of 7 occurs is -45° to 45°. The range of the sun's altitude when a seismic intensity of 6+ occurs is -66° to 66°. The range of the sun's altitude when a seismic intensity of 6- occurs is -75° to 70°.

[0121] 4-2 Category 2 Field of regression equation from the main decision tree Correlation between seismic intensity, vertical axis gravitational vector of the moon (X), and vertical axis gravitational vector of the sun (Y)

[0122] <Meaning of the graph> Figure 16 is an explanatory text for the decision tree 17 (dedicated to the regression equation for seismic intensity 7).

[0123] <Characteristics> The greater the seismic intensity, the smaller the solar vertical axis gravitational vector and the lunar vertical axis gravitational vector become. For seismic intensity 7, this becomes extreme, and it becomes possible to create a regression equation like the one shown by the dotted line in the lower right figure. Regression equation Ydt17 = 46.1168X 3.6089 x 10^21 (maximum residual ±1.35 x 10^22) *Adjusted so that the ± of the residual (intercept) is the same value.

[0124] <Numerical range of the Decision Tree for prediction> 100% of earthquakes with a seismic intensity of 7 are within the specific range below, and it can be said that there is a high probability of an earthquake occurring under the same conditions in the future. When using the values ​​for prediction, it is wise to leave a slight margin of error, but if the margin is too large, predictions will become impossible, so finding the right balance is key.

[0125] [Decision tree 17] DT17 Coordinates are astronomical coordinates of the National Astronomical Observatory → mathematical coordinates. The Y-axis vector range of the gravitational force of the sun when a magnitude 7 earthquake occurs is the regression formula Ydt17 = 46.1168X・3.6089 x 10^21 (±1.36 x 10^22) or less. No magnitude 6-weak or magnitude 6-strong earthquakes.

[0126] [Decision Tree 18] DT18 Correlation between the lunar and solar directions at the time of earthquake occurrence, by seismic intensity

[0127] <Meaning of the graph> Figure 17 uses data from the National Astronomical Observatory in hourly increments for the Tokyo field of view from January 1, 1995 to June 30, 2024, to show the correlation between the lunar and solar directions at the time of earthquakes, broken down by seismic intensity and those with a seismic intensity of lower 5 or higher, as point cloud data. The directions are converted from astronomical coordinates to mathematical coordinates, with 90° being north, 0° being east, 270° being south, and 180° being west.

[0128] <Characteristics> It appears that a regression equation is likely to be required for seismic intensities of 6+ and 7. For seismic intensities of 5-, the points are concentrated in the center (west) of the graph, but this is not the case for seismic intensities of 5+ or higher.

[0129] <Meaning of the graph> (Reason for using a quartic equation in the regression equation) Figure 18 shows the directional trajectories of the moon and sun for each lunar age. 360° was divided into four 90° periods and the trajectories were drawn. Period S is for lunar ages 26 or greater, and from 0 to less than 3.8. Period E is for lunar ages 3.8 to less than 11.3. Period N is for lunar ages 11.3 to less than 18.7. Period W is for lunar ages 18.7 to less than 26.

[0130] For example, when the moon's age is 0, the sun and moon face in almost the same direction, and the interior angle between the two celestial bodies is almost 0°, which is what is known as a new moon. When the moon's age is 14.8, the sun and moon face in opposite directions, and the interior angle between the two celestial bodies is almost 180°, which is what is known as a full moon. In other words, the moon's age can be said to be the interior angle between the directions of the sun and moon as seen from the sky, and because of the existence of the four seasons due to the angle of the Earth's rotation, the transition in the direction of the two celestial bodies as the moon ages is depicted as a quartic curve, as shown in Figure 18.

[0131] [Decision tree 18 (6.0)] DT18 (6.0) Correlation between the direction of the moon and the sun at the time of earthquake occurrence by seismic intensity Regression equation for seismic intensity 6-weak Based on Figure 18, which visualizes the trajectories of the directions of the sun and the moon by lunar phase, we considered the possibility of drawing a quartic curve, and performed a regression analysis of the angle difference between the moon and the sun at the time of seismic intensity 6-weak, seismic intensity 6-up, and seismic intensity 7 occurrence from the point clouds in Figure 17.

[0132] When attempting to create regression equations for the occurrence of earthquakes with seismic intensities from 6-low to 7-high, seven quartic equations (15 equations) were created: Y1, Y2, Y3, (Y1', Y2', Y3', Y3'') for seismic intensity 6-low, Y4, Y5, (Y4', Y5') for seismic intensity 6-upper, and Y6, Y7, (Y6', Y7') for seismic intensity 7, with maximum residuals ranging from 0 to 47. In other words, it is highly likely that the quartic equations described below hold true for the correlation between the directions of the sun and the moon at the time of earthquakes with seismic intensities of at least 6-upper and 7-high.

[0133] <Meaning of the graph> Figure 19 is the regression equation for a seismic intensity of 6 lower. The X axis is the direction of the moon in mathematical coordinates, and the Y axis is the direction of the sun in mathematical coordinates. Assuming that the variable X (moon direction) in the quartic equation can be changed to X' = (X ± 360), all of the point cloud data for a seismic intensity of 6 lower in Figure 17 was clustered (grouped) as shown in the top row of Figure 13, Y1, Y2, and Y3, and divided into three regression equations. When the data was returned to the area between 0° and 360° in the X direction, the direction of actual celestial bodies, the curves Y1 to Y3 shown in the bottom row of Figure 13 were obtained.

[0134] <Characteristics> When -6 is added to the intercepts of Y1, 4.5 to Y2, and 3.6 to Y3, and adjustments are made so that the maximum positive residual and the maximum negative residual are the same number of ±, the maximum residuals of each equation become ±35, ±35, and ±30 in the same order, and all calculation results for the correlation between the direction of the moon and the sun when a seismic intensity of 6-weak occurred are 100% contained within the range of ±30 to ±35 of the calculation results of each equation.

[0135] The gravitational force-related values ​​in Figures 9 to 19 are calculated based on the distance of celestial bodies and the angle of celestial bodies relative to the epicenter on the ground, so it is not surprising that the characteristics of this regression equation were seen in the angle difference between the moon and the sun when a seismic intensity of 6-weak occurred. We should be mindful of the possibility of overfitting (coincidence), but since the regression equation was also found for the later seismic intensity of 6-up and 7, the possibility of overfitting is not zero, but I think it is low.

[0136] Overfitting here refers to the phenomenon where the sample and the regression equation match because there are few samples (earthquakes), but when the number of samples increases, they no longer match the regression equation. This will actually become clear in the future, but since three pattern regression equations are required for seismic intensities from 6-lower to 7, we think the possibility is low.

[0137] <Numerical range of the prediction decision tree> [Decision tree 18 (6.0)] DT18 (6.0) If X is the direction of the moon's mathematical coordinates and Y is the direction of the sun's mathematical coordinates, when the result of calculation of any of the following seven formulas falls within this range, it is defined as the range in which an earthquake may occur. Y1=-1.1980x10^-06 x X^4 +0.001285 x X^3 -0.47 x X^2 +66.9637 x X -2846.8984 -6 Calculation result: ±35 Y1'=-1.1980x(X+360)^4 +0.001285x(X+360)^3 -0.47x(X+360)^2 +66.9637x(X+360) +2846.8984 -6 Calculation result: ±35 Y2=8.0401x10^-8 x X^4 -3.6323x10^-5 x X^3 -2.4456 x 10^-5 x X^2 +1.6746 x X +5.8107 +4.5 Calculation result: ±35 Y2'=8.0401x10^-8 x (X-360)^4 -3.6323 x 10^-5 x (X-360)^3 -2.4456 x 10^-5 x (X-360)^2 +1.6746 x (X-360) +5.8107 +4.5 Calculation result: ±35 Y3=-1.3025 x 10^-7 x X^4 +4.4282 x 10^-5 x X^3 +0.009139 x X^2 -2.3619 x X +109.4183 +3.6 Calculation result: ±30 Y3'=-1.3025x10^-7 x (X-360)^4 +4.4282 x 10^-5 x (X-360)^3 +0.009139 x (X-360)^2 -2.3619 x (X -360)+109.4183 +3.6 Calculation result: ±30 Y3''=-1.3025x10^-7 x (X+360)^4 +4.4282 x 10^-5 x (X+360)^3 +0.009139 x (X+360)^2 -2.36190 x (X+360)+109.4183 +3.6 Calculation result: ±30

[0138] [Decision Tree 18 (6.5)] DT18 (6.5) Correlation between the lunar and solar directions at the time of earthquake occurrence by seismic intensity Finding a regression equation for seismic intensity 6+ Figure 20 is a diagram explaining decision tree 18 (6.5).

[0139] <Meaning of the graph> The X axis is the direction of the moon in mathematical coordinates, and the Y axis is the direction of the sun in mathematical coordinates. We assumed that the variable X (moon direction) in the quartic equation could be changed to X' = (X ± 360), and all the point cloud data for seismic intensity 6+ in Figure 17 was clustered as shown in Y4 and Y5 in Figure 20, and separated into two regression equations.

[0140] Since there are no negative directions for celestial bodies in reality, if we force the data in the negative area in the X direction back to positive, we get the curves Y4 to Y5 shown on the right in Figure 20. *Note 1: There is also a graph near X=345. *Note 2: The dotted line in graph Y3 is not accurate.

[0141] <Characteristics> When 0 is added to the intercept for Y3 and -6 for Y4 and adjustments are made so that the maximum positive residual and the maximum negative residual are the same number of ± values, the maximum residuals for each equation become ±0 and ±46 in the same order, and all calculation results for the correlation between the direction of the moon and the sun when a seismic intensity of 6+ occurs are 100% consistent with the range of ±0 to 46 calculated for each equation. For the gravitational force-related figures in Figures 9 to 20, the distance of the celestial body and the angle of the celestial body above the epicenter are important, and it is not surprising that the characteristics of this regression equation were seen in the angle difference between the moon and the sun when a seismic intensity of 6+ occurs.

[0142] <Numerical range of the prediction decision tree> [Decision tree 18 (6.5)] DT18 (6.5) If X is the direction of the moon's mathematical coordinates and Y is the direction of the sun's mathematical coordinates, when the result of the calculation of any of the four formulas below falls within this range, it is defined as the range in which an earthquake may occur.

[0143] Y4=-0.000653524*X^4 + 0.018232321*X^3 +0.578158213*X^2 -32.11073221 Calculation result: ±10 Maximum residual: 0.000 Y5=(4.199*10^-7)X^4 + (-3.032*10^-4)X^3 + (7.316*10^-2)X^2 + (-5.959)X + 140.0038844 -6 *-6 is the calculation result of the adjustment value to make the residuals ±47 Maximum residual: ±46 Y4'=-0.000653524*(X-360)^4 +0.018232321*(X-360)^3 +0.578158213*(X-360)^2 -32.11073221 Calculation result: ±10 Y5'=(4.199*10^-7)*X^4 + (-3.032*10^-4)*X^3 + (7.316*10^-2)*X^2 + (-5.959)X + 140.0038844 -6 +360 Calculation result: ±47

[0144] [Decision tree 18 (7.0)] DT18 (7.0) Correlation between seismic intensity, moon and solar direction at the time of earthquake occurrence Regression equation for seismic intensity 7

[0145] <Meaning of the graph> Figure 21 is the regression equation for a seismic intensity of 7. The X axis is the direction of the moon's mathematical coordinates, and the Y axis is the direction of the sun's mathematical coordinates. Point clouds are marked when a seismic intensity of 7 occurs. Assuming that the variable X in the quartic equation can be changed to X' = (X ± 360), all point cloud data for seismic intensity of 7 in Figure 17 was clustered as shown at Y6 and Y7 in Figure 21, and separated into two regression equations. (It was not possible to interpolate the approximation equation using the dotted lines for Y6 and Y7.) When the direction of actual celestial bodies was returned to within the range of 0 to 360°, the curves from Y6 to Y7 were as shown on the far right of Figure 21.

[0146] <Characteristics> The maximum residual for each equation was ±4.0 x 10^-5, and all calculation results for the correlation between the direction of the moon and the sun when a magnitude 7 earthquake occurred were in 100% agreement within the range of ±4.0 x 10^-5 for each equation. The gravitational force-related values ​​in Figures 6 to 19 are important in terms of the distance of the celestial body and the angle of the celestial body above the epicenter, and it is not surprising that the characteristics of this regression equation were seen in the angle difference between the moon and the sun when a magnitude 7 earthquake occurred.

[0147] <Numerical range of the prediction decision tree> [Decision tree 18 (7.0)] DT18 (7.0) If X is the direction of the moon's mathematical coordinates and Y is the direction of the sun's mathematical coordinates, when the result of the calculation of any of the following four formulas falls within this range, it is defined as the range in which an earthquake may occur.

[0148] Y6=(4.26327*10^-7)*X^4 + (-3.19176*10^-4)*X^3 + (6.6742389*10^-2)*X^2 + (4.26327*10^-7)*X -651.454 Calculation result ±10 Maximum residual ±0.00004 Y7=(-2.58557*10^-6)*X^4 + (5.08942*10^-4)*X^3 + 31.04594536 Calculation result ±10 Maximum residual ±0.00000 Y6' =(4.26327*10^-7)*(X+360)^4 + (-3.19176*10^-4)*(X+360)^3 + (6.6742389*10^-2)*(X+360)^2 + (4.26327*10^-7)*(X+360) -651.454, the result is ±10. Y7'=(-2.58557*10^-6)*X^4 + (5.08942*10^-4)*X^3 + 31.04594536 -360, the result is ±10.

[0149] [Decision Tree 19, 20] (DT19, 20) Correlation between the Earth's four seasons (dates) and the age of the moon by seismic intensity

[0150] <Meaning of the graph> Figure 22 is a diagram explaining decision trees 19 and 20. Using a coordinate system (coordinates AbDg) in which the sun is the origin 1, the perihelion of the Earth is January 2 (mathematical coordinates 180°), and the aphelion is July 2 (mathematical coordinates 0°), the point cloud data shows the number of degrees on the Earth's orbit around the sun (graph X-axis) when an earthquake of each seismic intensity occurs, and at the same time, the number of degrees in the direction of the moon when the Earth is taken as origin 2 (direction AbDg similar to origin 1) (graph Y-axis).

[0151] <Characteristics> Correlation between the position on the Earth's orbit (date) and the position of the Moon when an earthquake occurs as seen from the Earth, by seismic intensity. By deriving the regression equations for seismic intensities 6+ and 7, the conditions for occurrence of seismic intensities 6+ or higher can be narrowed down. Figure 23 is a diagram explaining the regression equations for seismic intensities 6+ and 7. In Figure 22, it appears that a relatively large number of earthquakes with a seismic intensity of 5+ or higher occur around 270° on the X axis, particularly around 180° on the Y axis. This location is shown in the example in Figure 23, and corresponds to "around lunar phase 7 in March" (Great East Japan Earthquake). In other words, there appears to be a correlation between the position on the Earth's orbit and the age of the moon.

[0152] For seismic intensity levels of 6-lower or higher, no earthquakes occurred in the area enclosed by the square in Figure 22. No earthquakes for seismic intensity levels of 6-lower or higher occurred between November and January, and between lunar ages 23 and 29, and between lunar ages 0 and 7. For seismic intensity levels of 6-upper and 7, a regression equation like that shown in Figure 23 could be created.

[0153] <Numerical range of the prediction decision tree> [Decision tree 19] DT19 Seismic intensity 6-weak, 6-upper, and 7 common Coordinates AbDg: "X: less than 120°, more than 210°, Y: more than 180°" (outside the range enclosed by the square in Figure 22)

[0154] [Decision tree 20] DT20 Using a coordinate system (coordinates AbDg) with the sun as origin 1, the earth's perihelion as January 2nd (mathematical coordinates 180°), and the aphelion as July 2nd (mathematical coordinates 0°), when an earthquake of each seismic intensity occurs, the degree to which the earth is in its orbit around the sun is defined as X, and the degree to which the moon is in the direction when the earth is also at origin 2 (same direction AbDg as origin 1), is defined as Y. If each angle at the time of the earthquake falls within the range of the following calculation results, this is defined as the range in which an earthquake may occur.

[0155] Seismic intensity 6+ Regression formula Y8 Y8=5.58321x10^-9 x X^4 -2.7613x10^-5 x X^3 +0.021150215 x X^2 -4.02071314 x X +248.7765525 +11.5 (±69) Seismic intensity 7 Regression formula Y9 Y9= 3.23974 x 10^-6 x X^4 -0.002352004 x X^3 + 0.598601546 x X^2 -61.69503227 x X +2243.124239 -1 (±18)

[0156] [Decision tree 21 (6.5)] DT21 (6.5) Correlation between the total gravitational force of the sun and moon on the X-axis and the angle difference (interior angle) between the sun and moon, by seismic intensity. Find the regression equation for seismic intensity 6+.

[0157] <Meaning of the graph> Figure 24 is a diagram explaining the decision tree 21, and also explains the regression equations Y10, Y11, and Y12 for a seismic intensity of 6+, and is different from Figure 23. It shows the correlation between the gravitational force of the sun |X| (graph X-axis) when a seismic intensity of 6+ occurs and the interior angle (within 180°) between the direction of the sun and the moon (graph Y-axis) when the earthquake occurs.

[0158] <Characteristics> Three regression equations were created. From the left, they are Y10, Y11, and Y12. The equation below is the sum of the gravitational forces of the sun and the moon, |X|. Numerically, it is almost entirely the gravitational force of the sun, |X|, so one might think that the sun is most closely related to earthquakes, but the following three patterns can be seen.

[0159] From the equation for Y10, when the interior angle with the moon is narrow (= when the moon is in almost the same direction as the sun), or when the interior angle with the moon is wide (= when the moon is in the opposite direction to the sun), an earthquake of magnitude 6+ occurs even if the gravitational force |X| is relatively weak. From the equation for Y11, the gravitational force of the sun |X| becomes larger in proportion to the interior angle with the moon, causing an earthquake.

[0160] From the equation for Y12, earthquakes occur when the interior angle is approximately 25°, regardless of the gravitational force of the sun |X|.

[0161] <Numerical range of prediction decision tree> [Decision tree 21 (6.5)] DT21 (6.5) When the gravitational force of the sun |X| at the time of the occurrence of a seismic intensity of 6+ is defined as X, and the interior angle (within 180°) between the directions of the sun and the moon at the time of the earthquake is defined as Y, if each angle at the time of the earthquake is within the range of the calculation results below, this is defined as the range in which an earthquake may occur.

[0162] Y10=1.9067x10^-20xX -107.2275 (±10) Y11=4.0909x10^-21xX-7.2178 +6.5 (±29) Y12=-3.8063x 10^-23xX +25.1579 (±10)

[0163] [Decision Tree 21 (7.0)] DT21 (7.0) Correlation between the total gravitational force of the sun and moon on the X-axis and the angle difference (interior angle) between the sun and moon, by seismic intensity 83B

[0164] <Meaning of the graph> Figure 25 is a diagram for determining the regression equation for a seismic intensity of 7. It shows the correlation between the gravitational force of the sun |X| (graph X-axis) when a seismic intensity of 7 occurs and the angular difference between the directions of the sun and the moon (graph Y-axis) (*limited to within 180°) when the earthquake occurs.

[0165] <Characteristics> Two regression equations were created. From the right, they are Y13 and Y14. The slope has changed completely and is now the opposite of a strong 6. If the gravitational force of the sun, |X|, is strong, at 3x10^22 or more, the interior angle will be other than 90°±30°, and if the gravitational force of the sun, |X|, is weak, the interior angle will be around 120°. The number of samples for seismic intensity 7 is particularly small, so the possibility of overfitting cannot be ruled out, but a nice regression equation has been created and it appears there is a correlation.

[0166] <Numerical range of the prediction decision tree> [Decision tree 21 (7.0)] DT21 (7.0) When the gravitational force of the sun |X| at the time of a seismic intensity 7 occurrence is X, and the interior angle (within 180°) between the direction of the sun and the moon at the time of the earthquake occurrence is Y, the range of the angles at the time of the earthquake occurrence that are within the range of the following calculation results is defined as the range in which an earthquake may occur. Y13=-8.18308x10^-21xX +433.5072 -1.25 ±10 Y14=-9.0005x10^-21xX +317.0332 +1 ±10

[0167] 4-3 Category 3: Fields related to gravitational forces of solar system bodies Figure 26 is a diagram explaining decision trees 22 to 25.

[0168] [Decision Tree 22] DT22 Figure 26(a) is a diagram explaining decision tree 22. When the northern hemisphere of the Earth is viewed from above, the Earth is set as the coordinate origin, and the position of the sun in the mathematical coordinate system is always fixed at 90 degrees. The "maximum value of the y+ component" of the gravitational force exerted by the moon on the Earth when an earthquake of seismic intensity 7 (JP7), 6+ (JP6.5), or 6- (JP6) occurs in Japan is found, and anything below this is set as (the range when an earthquake occurs).

[0169] (Range at the time of earthquake occurrence) JP7: 7.55x10^19 or less JP6.5: 1.78x10^20 or less JP6: 2.19x10^20 or less *Law of universal gravitation F=GMm / R^2, unit is G=6.67428x10^-11 m^3 Kg^-1 S^-2, mass M, m are kg, distance R is m, units of calculation results are omitted hereafter.

[0170] [Decision Tree 23] DT23 Figure 26(b) is a diagram explaining decision tree 23. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the y-component" of the gravitational force that the moon exerts on the Earth is found, and anything below this is taken as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.85x10^20 or less JP6.5: 2.02x10^20 or less JP6.0: 1.97x10^20 or less

[0171] [Decision Tree 24] DT24 Figure 26(c) is a diagram explaining decision tree 24. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the X+ component" of the gravitational force exerted by the moon on the Earth is found, and anything below this is taken as (the range when an earthquake occurs). (Range when an earthquake occurs): JP7: 1.47x10^20 or less JP6.5: 9.09x10^19 or less JP6: 1.77x10^20 or less

[0172] [Decision Tree 25] DT25 Figure 26(d) is a diagram explaining decision tree 25. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the X-component" of the gravitational force that the moon exerts on the Earth is found, and anything below this is defined as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.95x10^20 or less JP6.5: 1.91x10^20 or less JP6: 2.03x10^20 or less

[0173] [Decision Tree 26A] DT26A Figure 27 is a diagram explaining decision trees 26 to 30. Figure 27(a) is a diagram explaining decision tree 25. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the Y+ component" of the gravitational force exerted on the Earth by the planets in the solar system is found, and anything below this is defined as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.0x10^18 or less JP6.5: 1.13x10^18 or less JP6: 1.3x10^18 or less

[0174] [Decision Tree 27A] DT27A Figure 27(b) is a diagram explaining decision tree 27A. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the Y-component" of the gravitational force exerted on the Earth by the planets in the solar system is found, and anything below this value is defined as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.0x10^18 or less JP6.5: 1.94x10^18 or less JP6: 2.08x10^18 or less

[0175] [Decision Tree 28A] DT28A Figure 27(c) is a diagram explaining decision tree 28A. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the X+ component" of the gravitational force exerted on the Earth by the planets in the solar system is found, and anything below that is defined as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.6x10^16 or more JP6.5: 1.38x10^18 or less JP6: 1.43x10^18 or less

[0176] [Decision Tree 29A] DT29A Figure 27(d) is a diagram explaining decision tree 29A. The coordinates are the same as DT22. When an earthquake of seismic intensity 7 (JP7), seismic intensity 6+ (JP6.5), or seismic intensity 6- (JP6) occurs in Japan, the "maximum value of the X-component" of the gravitational force exerted on the Earth by the planets in the solar system is found, and anything below this is defined as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 1.6x10^16 or more JP6.5: 9.63x10^17 or less JP6: 1.48x10^18 or less

[0177] [Decision Tree 30] DT30 Figure 27(e) is a diagram explaining decision tree 30. The coordinates are the same as DT22. When an earthquake of magnitude 7 (JP7), 6+ (JP6.5), or 6- (JP6) occurs in Japan, calculate the maximum value of the |Y+|+|Y-|+|X+|+|X-| component of the gravitational force exerted on the Earth by the planets in the solar system, that is, the sum of DT26A to DT29A, and determine the value below that as (the range when an earthquake occurs). (Range when an earthquake occurs) JP7: 2.0x10^18 or less JP6.5: 2.31x10^18 or less JP6: 2.36x10^18 or less

[0178] [Decision Tree 31] DT31 Figure 28 is a diagram explaining decision trees 31 to 34. Figure 28(a) is a diagram explaining decision tree 31. The coordinates are the same as DT22. The fraction of the Y+ components is DT22 / DT26A. (Range at the time of earthquake occurrence) JP7: 300 or less JP6.5: 500 or less JP6: 2000 or less

[0179] [Decision Tree 32] DT32 Figure 28(b) is a diagram explaining the decision tree 32. The coordinates are the same as DT22. The fraction of the Y-components is DT23 / DT27A. (Range at the time of earthquake occurrence) JP7: 14,000 or less JP6.5: 4,000 or less JP6: 5,000 or less

[0180] [Decision Tree 33] DT33 Figure 28(c) is a diagram explaining the decision tree 33. The coordinates are the same as DT22. The fraction of the X+ components is DT24 / DT28A. (Range at the time of earthquake occurrence) JP7: 9200 or less JP6.5: 2000 or less JP6: 10000 or less

[0181] [Decision Tree 34] DT34 Figure 28(d) is a diagram explaining the decision tree 34. The coordinates are the same as DT22. The fraction of the X-components is DT25 / DT29A. (Range at the time of earthquake occurrence) JP7: 4900 or less JP6.5: 78000 or less JP6: 457000 or less

[0182] [Decision Tree 35] DT35 Figure 29 is a diagram explaining decision trees 35 to 38. Figure 29(a) is a diagram explaining decision tree 35. The coordinates are the same as DT22. The fraction of the inverse components of Y is DT22 / DT27A. (Range at the time of earthquake occurrence) JP7: 1500 or less JP6.5: 1500 or less JP6: 210000 or less

[0183] [Decision Tree 36] DT36 Figure 29(b) is a diagram explaining the decision tree 36. The coordinates are the same as DT22. The fraction of the inverse components of Y is DT23 / DT26A. (Range at the time of earthquake occurrence) JP7: 1000 or less JP6.5: 4100 or less JP6: 4000 or less

[0184] [Decision Tree 37] DT37 Figure 29(c) is a diagram explaining decision tree 37. The coordinates are the same as DT22. The fraction of the inverse components of X is DT24 / DT29A. (Range at the time of earthquake occurrence) JP7: 300 or less JP6.5: 1000 or less JP6: 16000 or less

[0185] [Decision Tree 38] DT38 Figure 29(d) is a diagram explaining the decision tree 38. The coordinates are the same as DT22. The fraction of the inverse components of X is DT25 / DT28A. (Range at the time of earthquake occurrence) JP7: 3200 or less JP6.5: 4500 or less JP6: 9000 or less

[0186] [Decision Tree 39] DT39 Figure 30 is a diagram explaining decision tree 39. The coordinates are the same as DT22. When the Y± components ((Y+) + (Y-)) of the solar system planets (Mercury, Venus, Mars, Jupiter, Saturn, Uranus, Neptune) are Ysum and the X± components ((X+) + (X-)) are Xsum, the result is (Ysum^2 +Xsum^2)^1 / 2 Ysum = DT26A-DT27A, Xsum = DT28A-DT29A (Range at the time of earthquake occurrence) JP7: 1.5x10^18 or less JP6.5: 1.89x10^18 or less JP6: 2.03x10^18 or less

[0187] [Decision Tree 40] DT40 Figure 31 is a diagram explaining decision trees 40 to 43. Figure 31(a) is a diagram explaining decision tree 40. The coordinates are the same as DT22. This is the "gravitational force Y+ component of the solar system planets acting on the Earth" when the gravitational force Y+ component of the Moon's gravitational force on the Earth is 0 (the Moon is in the third or fourth quadrant), or the "gravitational force Y- component of the solar system planets acting on the Earth" when the gravitational force Y+ component of the Moon's gravitational force on the Earth is + (the Moon is in the first or second quadrant). (Range when an earthquake occurs) JP7: 9.68x10^1 or less JP6.5: 1.13x10^18 or less JP6: 1.04x10^18 or less

[0188] [Decision Tree 41] DT41 Figure 31 (b) is a diagram explaining decision tree 41. The coordinates are the same as DT22. This is the "X+ component of gravitational force exerted by the solar system planets on the Earth" when the X+ component of the gravitational force exerted by the Moon on the Earth is 0 (the Moon is in the second or third quadrant), or the "X- component of gravitational force exerted by the solar system planets on the Earth" when the X+ component of the gravitational force exerted by the Moon on the Earth is + (the Moon is in the first or fourth quadrant). (Range when an earthquake occurs) JP7: 1.5x10^18 or less JP6.5: 1.38x10^18 or less JP6: 1.48x10^18 or less

[0189] [Decision Tree 42] DT42 Figure 31 (c) is a diagram explaining decision tree 42. The coordinates are the same as DT22. The gravitational force of the total solar system planets in the same quadrant as the moon. (Range at the time of earthquake occurrence) JP7: 1.64x10^18 or less JP6.5: 1.97x10^18 or less JP6: 2.12x10^18 or less

[0190] [Decision Tree 43] DT43 Figure 31 (d) is a diagram explaining decision tree 43. The coordinates are the same as DT22. The gravitational force of the moon and the solar system planets in the target quadrant. (Range at the time of earthquake occurrence) JP7: 1.6x10^18 or less JP6.5: 1.44x10^18 or less JP6: 1.7x10^18 or less

[0191] [Decision Tree 44] DT44 Figure 32 is a diagram explaining decision tree 44. The coordinates are the same as DT22. This is the gravitational force (fraction) of the solar system planets on X+ when the moon is at X+, or the gravitational force (fraction) of the solar system planets on X- when the moon is at X-. Figure 32 is an example. (Range when an earthquake occurs) JP7: 9200 or less JP6.5: 78000 or less JP6: Not applicable

[0192] [Decision Tree 45] DT45 Figure 33 is a diagram explaining decision trees 45 to 48. Figure 33(a) is a diagram explaining decision tree 45. The coordinates are the same as DT22. This is the gravitational force of the moon when the moon is in Y+ minus the gravitational force of the solar system planets in Y+ (subtraction). (Range when earthquake occurs) JP7: 7.7x10^19 or less JP6.5: 1.78x10^20 or less JP6: 2.19x10^20 or less

[0193] [Decision Tree 46] DT46 Figure 33(b) is a diagram explaining decision tree 46. The coordinates are the same as DT22. This is the gravitational force of the moon when the moon is in Y- minus the gravitational force of the solar system planets in Y- (subtraction). (Range at the time of earthquake occurrence) JP7: 1.85x10^20 or less JP6.5: 2.02x10^20 or less JP6: 1.97x10^20 or less

[0194] [Decision Tree 47] DT47 Figure 33(c) is a diagram explaining decision tree 47. The coordinates are the same as DT22. This is the gravitational force of the moon when the moon is at X+ minus the gravitational force of the solar system planets at X+ (subtraction). (Range when earthquake occurs) JP7: 1.48x10^20 or less JP6.5: 9.10x10^20 or less JP6: 1.77x10^20 or less

[0195] [Decision Tree 48] DT48 Figure 33(d) is a diagram explaining decision tree 48. The coordinates are the same as DT22. It is the gravitational force of the moon when the moon is at X- minus the gravitational force of the solar system planets at X- (subtraction). (Range when earthquake occurs) JP7: 1.96x10^20 or less JP6.5: 1.92x10^20 or less JP6: 2.03x10^20 or less

[0196] [Decision tree 49] DT49 The coordinates are the same as DT22. It is the sum of DT45 + 46 + 47 + 48. (Range at the time of earthquake occurrence) JP7: 2.23x10^20 or less JP6.5: 2.12x10^20 or less JP6: 2.30x10^20 or less

[0197] [Decision Tree 50] DT50 Figure 34 is a diagram explaining decision trees 50 and 51. Figure 34(a) is a diagram explaining decision tree 50. The coordinates are the same as DT22. When the moon is in Y+, it is DT22 + DT27A. When the moon is in Y-, it is DT23 + DT26A. Figure 34 is a diagram of DT22 + DT27A. (Range at the time of earthquake occurrence) JP7: 1.86x10^20 or less JP6.5: 2.03x10^20 or less JP6: 2.18x10^20 or less

[0198] [Decision Tree 51] DT51 Figure 34(b) is a diagram explaining decision tree 51. The coordinates are the same as DT22. When the moon is at X+, DT24 + DT29A. When the moon is at X-, DT25 + DT28A. Figure 34 is a diagram of DT25 + DT28A. (Range at the time of earthquake occurrence) JP7: 1.95x10^20 or less JP6.5: 1.91x10^20 or less JP6: 2.03x10^20 or less

[0199] [Decision tree 52] DT52 coordinates are the same as DT22. This is the sum of DT50 and DT51. (Range at the time of earthquake occurrence) JP7: 2.24x10^20 or less JP6.5: 2.13x10^20 or less JP6: 2.29x10^20 or less

[0200] [Decision Tree 53] DT53 Figure 35 is a diagram explaining decision trees 50 and 51. Figure 35 is a diagram explaining decision tree 53. The coordinates are the same as DT22. When a straight line is drawn between Jupiter and Earth, if 0°≦angle θ≦180°, this is the Y+ component of Jupiter's gravitational force. (Range when an earthquake occurs) JP7: 9.12x10^17 or less JP6.5: 9.55x10^17 or less JP6: 9.56x10^17 or less

[0201] [Decision tree 54] DT54 DT53 takes into account the latitude of Tokyo. When the altitude of Jupiter as seen from Tokyo is θ, the calculation result of DT54 = DT53 is x Cos(θ)^2. *The Y axis of the solar system plane is the horizontal axis, so it is Cos. (Range at the time of earthquake occurrence) JP7: 6.8x10^17 or less JP6.5: 7.47x10^17 or less JP6: 9.28x10^17 or less

[0202] [Decision Tree 55] DT55 Figure 36 is a diagram explaining decision tree 55. The coordinates are the same as DT22. When a straight line is drawn between Jupiter and Earth, if 180° < angle θ < 360°, this is the Y-component of Jupiter's gravitational force. (Range when earthquake occurs) JP7: 9.12x10^17 or less JP6.5: 1.82x10^18 or less JP6: 2.08x10^18 or less

[0203] [Decision Tree 56] DT56 This is DT55 with the latitude of Tokyo taken into account. When the altitude of Jupiter as seen from Tokyo is θ, the calculation result of DT56 = DT55 is x Cos(θ)^2. Since the Y axis of the solar system plane is the horizontal axis, it is Cos. (Range at the time of earthquake occurrence) JP7: 6.8x10^17 or less JP6.5: 1.37x10^18 or less JP6: 2.07x10^18 or less

[0204] [Decision Tree 57] DT57 Figure 37 is a diagram explaining decision tree 57. The coordinates are the same as DT22. When a straight line is drawn between Jupiter and Earth, if 0°≦angle ω≦90° or 270°≦angle ω<360°, this is the X+ component of Jupiter's gravitational force. (Range when earthquake occurs) JP7: 1.2x10^17 or more JP6.5: 1.3x10^18 or less JP6: 1.38x10^18 or less

[0205] This is DT57 with the latitude of Tokyo taken into account, and when the altitude of Jupiter as seen from Tokyo is θ, the calculation result of DT58 = DT57 x Cos(ω)^2. Since the X axis of the solar system plane is the horizontal axis, it is Cos. (Range at the time of earthquake occurrence) JP7: 4.6x10^16 or more JP6.5: 1.26x10^18 or less JP6: 1.28x10^18 or less

[0206] [Decision Tree 59] DT59 Figure 38 is a diagram explaining decision tree 59. The coordinates are the same as DT22. When a straight line is drawn between Jupiter and Earth, if the angle ω is 90°≦ω<270°, this is the X-component of Jupiter's gravitational force. (Range when an earthquake occurs) JP7: 1.2x10^17 or more JP6.5: 7.6x10^17 or less JP6: 1.43x10^18 or less

[0207] [Decision Tree 60] DT60 is DT59 taking into account the latitude of Tokyo. When the altitude of Jupiter as seen from Tokyo is θ, DT60 = DT59 calculation result x Cos(ω)^2. Since the X axis of the solar system plane is the horizontal axis, it is Cos. (Range at the time of earthquake occurrence) JP7: 4.6x10^16 or more JP6.5: 7.08x10^17 or less JP6: 1.07x10^18 or less

[0208] [Decision tree 61] DT61 coordinates are the same as DT22. DT61 = the sum of the observed values ​​of DT53 + 55 + 57 + 59. (Range at the time of earthquake occurrence) JP7: 8.0x10^17 or more and 1.8x10^18 or less JP6.5: 1.87x10^18 or less JP6: 2.14x10^18 or less

[0209] [Decision tree 62] DT62 coordinates are the same as DT22. DT62 = the sum of the observed values ​​of DT54 + 56 + 58 + 60. (Range at the time of earthquake occurrence) JP7: 1.4x10^17 or more and 1.2x10^18 or less JP6.5: 1.71x10^18 or less JP6: 2.08x10^18 or less

[0210] [Decision Tree 63] DT63 Figure 39 is a diagram explaining decision tree 63. The coordinates are the same as DT22. What is the degree from the position of the Earth at the time of the earthquake on July 2nd? DT63 = - (minus) E position - (minus) 90 E position: The angle of the Earth on coordinates AbDg, with the sun as the origin and the Earth's position as 0° on July 2nd and 180° on January 2nd (Range at the time of the earthquake) JP7: 101° or less, 150° to 216°, 336° or more JP6.5: 58° or less, 150° to 187°, 237° or more JP6: 50° or less, 76° or more

[0211] [Decision Tree 64] DT64 Figure 40 is a diagram explaining decision tree 64. The coordinates are the same as DT22. DT64 = Sin(DT63), which is a continuation of DT63. The following results suggest a trend. (Range at the time of earthquake occurrence) JP7: -0.55 or more and -0.1 or less, 0.23 or more and 0.42 or less, 0.86 or more JP6.5: -0.82 or less, -0.63 or more and 0.47 or less, 0.64 or more JP6: Not applicable

[0212] [Decision tree 65] DT65 coordinates are the same as DT22. DT65 = Sin(DT64)^2. (Range at the time of earthquake occurrence) JP7: 0.3 or less, 0.83 or more JP6.5: 0.65 or less, 0.75 or more JP6: Not applicable

[0213] [Decision Tree 66] DT66 Figure 41 is a diagram illustrating the decision tree 66. The coordinate system is the AbDg coordinate system. The AbDg coordinate system is as follows: Assuming a year has 365 days, the sun is the origin, the aphelion is July 2 (mathematical coordinate 0°), +92 days (mathematical coordinate 90°), the perihelion is January 2 (mathematical coordinate 180°) on +91 days, +90 days (mathematical coordinate 270°), and the Earth returns to July 2 on +92 days. The Earth's orbit around the sun is technically an ellipse, but is represented graphically as a circle. Furthermore, the Earth's angular velocity is (90 / 92)° per day when it is in the first or fourth quadrant, (90 / 91)° per day when it is in the second quadrant, and (90 / 90)° per day when it is in the third quadrant. This coordinate system will be referred to as the AbDg coordinate system. When using these coordinates AbDg, we investigate what angle the Earth was at when the earthquake in question occurred. If the angle at the time of the earthquake was 43°, then the range (range at the time of earthquake occurrence) is 38-48°, plus a tolerance of ±5°. (Range at the time of earthquake occurrence) JP7: 54° to 120°, 169° to 205°, 238° to 293° JP6.5: 33° or less, 82° to 120°, 212° or more JP6: 194° or less, 220° or more *Note: This makes it possible to create a regression equation.

[0214] [Decision tree 67] DT67 At coordinates AbDg, calculate DT67 = CosX^2 as the position of the Earth (angle X) when the earthquake occurs. (Range when earthquake occurs) JP7: 0.3 or less, 0.83 or more JP6.5: 0.65 or less, 0.75 or more JP6: Not applicable

[0215] [Decision Tree 68] DT68 Figure 42 is a diagram explaining decision tree 68. It uses coordinates AbDg. When the earthquake under investigation occurred, what degree was the position of the moon at coordinates AbDg? In earthquakes of magnitude 6 or higher in Japan, regardless of when they occur, the moon occurs at an angle other than 90° or 270°, as shown in Fig*. We investigated whether there is a relationship between this coordinate AbDg and the following parameters. DT68 = Earth position (angle) of AbDg + age of the moon x 12.08 + 180. (Range at time of earthquake occurrence) JP7: 16° ​​to 56°, 146° to 222° JP6.5: 68° or less, 138° or more JP6: Not applicable Developing into regression equations Y8 and Y9 → DT19,20

[0216] [Decision tree 69] DT69 DT68 = COS(DT68). (Range at the time of earthquake occurrence) JP7: -0.55 or less, 0.71 or more JP6.5: -0.44 or less, 0.2 or more JP6: Not applicable

[0217] [Decision Tree 70] DT70 uses coordinates AbDg. DT70 = Cos(90 + (Monthly x 12.0805) - DT63) was used to convert to coordinates AbDg, and the influence of the direction on July 2nd of the month when the earthquake occurred was expressed in COS. The closer the absolute value is to 1 (-1 or 1), the more likely there is a relationship. (Range at the time of earthquake occurrence) JP7: -0.74 or less, -0.08 to 0.12 or less, 0.38 to 0.58 or less, 0.83 or more JP6.5: -0.38 or less, -0.19 or more JP6: -0.44 or less, -0.26 or more

[0218] [Decision tree 71] DT71 Uses coordinates AbDg. DT71 = DT70^2. (Range at the time of earthquake occurrence) JP7: -0.33 or less, 0.61 or more JP6.5: Not applicable JP6: Not applicable

[0219] [Decision Tree 72] DT72 Uses coordinates AbDg. Most of the gravitational influence that Earth receives from the planets in the solar system comes from Jupiter, and we will explore the influence of Jupiter alone. Figure 43 is a diagram explaining decision tree 72. When the angle of Jupiter at coordinate DT22 is Jdeg, the formula SIN (DT63 - Jdeg) is used to convert it to coordinate AbDg, and the degree of influence of Jupiter's direction on July 2nd at the time of the earthquake is expressed in SIN. The closer it is to 0, the more likely there is a relationship. (Range at the time of earthquake occurrence) JP7: -0.8 or less, 0.43 or more JP6.5: -0.08 or less, 0.27 or more JP6: -0.02 or less, 0.25 or more

[0220] [Decision Tree 73] DT73 Figure 44 is a diagram explaining decision trees 73 to 76. Figure 44(a) is a diagram explaining decision tree 73. Coordinates AbDg are used. The gravitational influence that the Earth receives from the planets in the solar system is explored. The formula SIN(DT63) x DT26A was used to examine Y+ on coordinates AbDg. (Range at the time of earthquake occurrence) JP7: 6.0x10^17 or less JP6.5: 7.6x10^17 or less JP6: 8.39x10^17 or less

[0221] [Decision Tree 74] DT74 Figure 44(b) is a diagram explaining decision tree 74. Coordinates AbDg are used. The gravitational influence that the Earth receives from the planets in the solar system is explored. The formula SIN(DT63) x DT27A was used to examine Y+ on coordinates AbDg. (Range at the time of earthquake occurrence) JP7: 3.6x10^17 or less JP6.5: 4.1x10^17 or less JP6: 3.47x10^17 or less

[0222] [Decision Tree 75] DT75 Figure 44 (c) is a diagram explaining decision tree 75. Coordinates AbDg are used. The gravitational influence of the planets in the solar system on the Earth is explored. The formula SIN (DT63) x DT28A was used to examine X+ on coordinates AbDg. (Range at the time of earthquake occurrence) JP7: 5.9x10^17 or less JP6.5: 1.8x10^17 or less JP6: 1.01x10^18 or less

[0223] [Decision Tree 76] DT76 Figure 44 (d) is a diagram explaining decision tree 76. DT76 Old 48 (Sequential 81) Uses coordinates AbDg. Explores the gravitational influence that the Earth receives from the planets in the solar system. Uses the formula SIN (DT63) x DT29A to find X- on coordinates AbDg. (Range at the time of earthquake occurrence) JP7: 1.5x10^18 or less JP6.5: 5.5x10^16 or less JP6: 1.47x10^18 or less

[0224] [Decision tree 26B] DT26B DT26A / DT30. (Range at the time of earthquake occurrence) JP7: 0.28 or less and 0.43 or more JP6.5: 0.27 or less and 0.65 or more JP6: 0.9 or less

[0225] [Decision tree 26C] DT26C is the maximum value of DT26A / DT30. (Range at the time of earthquake occurrence) JP7: 0.33 or less JP6.5: 0.39 or less JP6: 0.45 or less

[0226] [Decision Tree 26D] DT26D coordinates are the same as DT22. If the sum of all the planets in the solar system (Mercury, Venus, Mars, Jupiter, Saturn) is considered to be one fictional planet, and a straight line is drawn between the fictional planet and the Earth, with 0°≦Angle A≦180° and the altitude of each planet as seen from Tokyo being θ, then DT26D = sum of each planet Y + component x Cos(θ)^2. (Range at the time of earthquake occurrence) JP7: 8.1x10^17 or less JP6.5: 8.06x10^17 or less JP6: 9.78x10^17 or less

[0227] [Decision tree 27B] DT27B DT27A / DT30. (Range at the time of earthquake occurrence) JP7: 0.59 or less JP6.5: Not applicable JP6: 0.24 or less, or 0.35 or more

[0228] [Decision tree 27C] DT27C is the maximum value of DT27A / DT30. (Range at the time of earthquake occurrence) JP7: 0.35 or less JP6.5: 0.64 or less JP6: 0.69 or less

[0229] [Decision Tree 27D] DT27D coordinates are the same as DT22. If the sum of all planets in the solar system (Mercury, Venus, Mars, Jupiter, Saturn) is considered to be one fictional planet, and a straight line is drawn between the fictional planet and Earth, with 180° < angle A < 360°, and the altitude of each planet as seen from Tokyo is θ, then DT27D = sum of each planet's Y-component x Cos(θ)^2. (Range at the time of earthquake occurrence) JP7: 7.0x10^17 or less JP6.5: 1.46x10^18 or less JP6: 2.07x10^18 or less

[0230] [Decision tree 28B] DT28B DT28A / DT30. (Range at the time of earthquake occurrence) JP7: 0.19 or less, 0.42 or more JP6.5: 0.27 or less, 0.65 or more JP6: Not applicable

[0231] [Decision tree 28C] DT28C is the maximum value of DT28A / DT30. (Range at the time of earthquake occurrence) JP7: 0.24 or less JP6.5: 0.47 or less JP6: 0.49 or less

[0232] [Decision Tree 28D] DT28D coordinates are the same as DT22. The sum of all the planets in the solar system (Mercury, Venus, Mars, Jupiter, Saturn) is considered to be one fictional planet, and a straight line is drawn between the fictional planet and Earth, with 0°≦Angle A≦90° or 270°≦Angle A≦360°, and the altitude of each planet as seen from Tokyo is θ, then DT28D = the sum of each planet X + component x Cos(θ)^2. (Range at the time of earthquake occurrence) JP7: 5.56x10^17 or less JP6.5: 1.26x10^18 or less JP6: 1.28x10^18 or less

[0233] [Decision tree 29B] DT29B DT29A / DT30. (Range at the time of earthquake occurrence) JP7: 0.17 or less and 0.32 or more JP6.5: 0.47 or less JP6: Not applicable

[0234] [Decision tree 29C] DT29C is the maximum value of DT29A / DT30. (Range at the time of earthquake occurrence) JP7: 0.49 or less JP6.5: 0.35 or less JP6: 0.51 or less

[0235] [Decision Tree 29D] DT29D coordinates are the same as DT22. If the sum of all the planets in the solar system (Mercury, Venus, Mars, Jupiter, Saturn) is considered to be one fictional planet, and a straight line is drawn between the fictional planet and Earth, with 90° < angle A < 90° or 270° < angle A < 360°, and the altitude of each planet as seen from Tokyo is θ, then DT29D = sum of each planet X + component x Cos(θ)^2. (Range at the time of earthquake occurrence) JP7: 8.27x10^17 or less JP6.5: 7.08x10^17 or less JP6: 1.11x10^18 or less

[0236] [Decision Tree 77] DT77 Figure 45 illustrates decision tree 77. Using the same coordinates as DT22 and coordinates AbDg, we finally determined the maximum Y component of the gravitational force of the solar system planets on coordinates AbDg. (Since coordinates AbDg are specific to this specification, they must be converted using a formula from the same coordinates as DT22.) A. Let Ysum be the Y component of the sum of the gravitational forces of the solar system planets at the time of the earthquake on coordinates DT22 (calculated with + being positive and - being negative), and Xsum be the arctan (Ysum / Xsum) of the X component. Let (1) be the arctan (Ysum / Xsum). B. Let (2) be the angle of the Earth at the time of the earthquake on coordinates AbDg. C. The scalar of the sum of the gravitational forces of the solar system planets is (Ysum^2 + Xsum^2)^1 / 2. The angle of the sum of the gravitational forces of the solar system planets on coordinates AbDg is (1) + (2) + 90°. D The Y component to be calculated is (Ysum^2 + Xsum^2)^1 / 2 x Sin((1) + (2) + 90°)^2 (Range at the time of earthquake occurrence) JP7: 1.4x10^18 or less JP6.5: 9.36x10^17 or less JP6: 1.71x10^18 or less

[0237] [Decision Tree 78] DT78 Figure 46 is a diagram for explaining decision tree 78. Using the same coordinates as DT22 and coordinates AbDg, finally, the maximum value of the X component of the gravitational force of the solar system planets on coordinates AbDg is obtained. (Since coordinates AbDg are coordinates dedicated to this specification, it is necessary to convert them from the same ones as coordinates DT22 using a calculation formula) A. When the Y component of the total gravitational force of the solar system planets at the time of earthquake occurrence on coordinates DT22 is set as Ysum and the X component is set as Xsum, arctan(Ysum / Xsum) is defined as (1). B. The angle of the Earth at the time of earthquake occurrence on coordinates AbDg is defined as (2). C. The scalar of the total gravitational force of the solar system planets is (Ysum^2 + Xsum^2)^(1 / 2). The angle of the total gravitational force of the solar system planets on coordinates AbDg is (1) + (2) + 90°. D. The X component to be obtained is (Ysum^2 + Xsum^2)^(1 / 2) x Cos((1) + (2) + 90°)^2 (range at the time of earthquake occurrence) JP7: 6.4x10^17 or less JP6.5: 1.77x10^18 or less JP6: 1.61x10^18 or less

[0238] [Decision Tree 79] DT79 Figure 47 is a diagram for explaining decision tree 79. Using the same coordinates as DT22 and coordinates AbDg, finally, the maximum value of the Y+ component of the gravitational force of the solar system planets on coordinates AbDg is obtained. (Since coordinates AbDg are coordinates dedicated to this specification, it is necessary to convert them from the same ones as coordinates DT22 using a calculation formula) When the angle of the Earth on coordinates AbDg at the time of earthquake occurrence is set as Eω, A. When Eω = 0°, it has the same value as DT28A. B. When 0° < Eω < 90°, (DT27A + DT28A) x Sin(Eω)^2 C. When Eω = 90°, it has the same value as DT27A. D. When 90° < Eω < 180°, (DT27A + DT29A) x Sin(Eω)^2 E. When Eω = 180°, it has the same value as DT29A. F. When 180° < Eω < 270°, (DT26A + DT29A) x Sin(Eω)^2 G. When Eω = 270°, it has the same value as DT26A. H. When 270° < Eω < 360°, (DT26A + DT28A) x Sin(Eω)^2 (range at the time of earthquake occurrence) JP7: 9.13x10^17 or less JP6.5: 1.06x10^18 or less JP6: 2.25x10^18 or less

[0239] [Decision Tree 80] DT80 FIG. 48 is a diagram for explaining the decision tree 80. Using the same coordinates as DT22 and the coordinates AbDg, finally, the maximum value of the Y-component of the gravitational force of the solar system planets on the coordinates AbDg is obtained. (Since the coordinates AbDg are coordinates dedicated to this specification, it is necessary to convert them from the same coordinates as DT22 using a calculation formula) When an earthquake occurs, if the angle of the earth on the coordinates AbDg is Eω, then A When Eω = 0°, it is the same value as DT29A B When 0° < Eω < 90°, (DT26A + DT29A) x Sin(Eω)^2 C When Eω = 90°, it is the same value as DT26A D When 90° < Eω < 180°, (DT26A + DT28A) x Sin(Eω)^2 E When Eω = 180°, it is the same value as DT28A F When 180° < Eω < 270°, (DT27A + DT28A) x Sin(Eω)^2 G When Eω = 270°, it is the same value as DT27A H When 270° < Eω < 360°, (DT27A + DT29A) x Sin(Eω)^2 (range during earthquake occurrence) JP7: 1.58x10^18 or less JP6.5: 1.56x10^18 or less JP6: 1.62x10^18 or less

[0240] [Decision Tree 81] DT81 FIG. 49 is a diagram for explaining the decision tree 81. Using the same coordinates as DT22 and the coordinates AbDg, finally, the maximum value of the X+ component of the gravitational force of the solar system planets on the coordinates AbDg is obtained. (Since the coordinates AbDg are coordinates exclusive to this specification, it is necessary to convert them from the same coordinates as DT22 using a calculation formula) When an earthquake occurs, if the angle of the earth on the coordinates AbDg is Eω, then A When Eω = 0°, it is equivalent to DT27A B When 0° < Eω < 90°, (DT27A + DT29A) x Cos(Eω)^2 C When Eω = 90°, it is equivalent to DT29A D When 90° < Eω < 180°, (DT26A + DT29A) x Cos(Eω)^2 E When Eω = 180°, it is equivalent to DT26A F When 180° < Eω < 270°, (DT26A + DT28A) x Cos(Eω)^2 G When Eω = 270°, it is equivalent to DT28A H When 270° < Eω < 360°, (DT27A + DT28A) x Cos(Eω)^2 (range during earthquake occurrence) JP7: 1.6x10^18 or less JP6.5: 1.85x10^18 or less JP6: 1.79x10^18 or less

[0241] [Decision Tree 82] DT82 Figure 50 is a diagram for explaining decision tree 82. Using the same coordinates as DT22 and coordinate AbDg, finally, the maximum value of the X-component of the gravitational force of the solar system planets on coordinate AbDg is obtained. (Since coordinate AbDg is a coordinate exclusive to this specification, it needs to be converted from the same coordinates as DT22 using a calculation formula.) When an earthquake occurs, if the angle of the Earth on coordinate AbDg is Eω, then A When Eω = 0°, it is the same value as DT26A B When 0° < Eω < 90°, (DT26A + DT28A) x Cos(Eω)^2 C When Eω = 90°, it is the same value as DT28A D When 90° < Eω < 180°, (DT27A + DT28A) x Cos(Eω)^2 E When Eω = 180°, it is the same value as DT27A F When 180° < Eω < 270°, (DT27A + DT29A) x Cos(Eω)^2 G When Eω = 270°, it is the same value as DT29A H When 270° < Eω < 360°, (DT26A + DT29A) x Cos(Eω)^2 (range during earthquake occurrence) JP7: 3.8x10^17 or less JP6.5: 7.9x10^17 or less JP6: 1.69x10^18 or less

[0242] [Decision Tree 83] DT83 Figure 51 is a diagram for explaining decision tree 83. It is the south celestial altitude in Tokyo on the day of the earthquake occurrence. (range during earthquake occurrence) JP7: 42° or less, or 65° or more JP6.5: 42° or less, or 65° or more JP6: Not applicable

[0243] [Decision Tree 84] DT84 Figure 52 is a diagram for explaining decision tree 84. The coordinates are the same as DT22. When the sun is fixed at 90° in mathematical coordinates, what is the angle (inner angle) between the sun and the moon at the time of earthquake occurrence? (range during earthquake occurrence) JP7: 26° or more and 46° or less, 156° or more and 212° or less, 266° or more and 333° or less JP6.5: 25° or more and 65° or less, 98° or more and 283° or less JP6: 17° or more and 347° or less Capture: Developed into a regression equation from this

[0244] [Decision tree 85] DT85 The lunar vertical axis gravitational force scalar, DT9 + |DT10|. (Range at the time of earthquake occurrence) JP7: 0 or less than 1.34x10^20 JP6.5: less than 1.67x10^20 JP6: less than 1.82x10^20

[0245] [Decision tree 30B] DT30B The sum of the gravitational scalar of each solar system planet and the attenuated scalar due to the altitude (Sin) of each planet relative to Tokyo (the sum of the gravitational scalar of each planet x altitude (Sin) of each planet) DT26D + DT27D + DT28D + DT29D (Range at the time of earthquake occurrence) JP7: Not required JP6.5: 1.86 x 10^18 or less JP6: 2.2 x 10^18 or less

[0246] 5. Explanation of earthquake prediction method Figure 53 is an image of the earthquake prediction process.

[0247] First, we set the target area for prediction, which we will assume is Japan.

[0248] Secondly, based on past records of earthquakes that have occurred in that area (for example, the period from 1995 to the present) (at least in units of one day, and units of one hour or less are recommended), a range is defined so that the calculation results of decision trees 1 to 103 include all values ​​at the time the predicted seismic intensity occurs (seismic intensities from seismic intensity 6-low to seismic intensity 7), and if they fall within this defined range, a 1 is entered, otherwise a 0 is entered. Naturally, a 1 will be entered in all decision trees at the time of the earthquake, and this will be the earthquake occurrence range, which will be called "TP" (true positive), indicating that an earthquake actually occurred (Past in the table).

[0249] On the other hand, most of the other dates and years have fewer than 103 entries for 1, but even for dates and years in which no earthquakes occurred, there are some dates and years in which all 1s are entered or the prescribed decision tree for each seismic intensity has a 1 entered. We will call this an "FN" (false negative), which means that although it is within the earthquake occurrence range, no earthquake actually occurred.

[0250] Thirdly, the decision tree is judged in the same way for future periods and a 1 or 0 is entered. Then, there will be a certain date and time where all decision trees are 1, or all the specified decision trees for each seismic intensity are 1. For example, if there is a seismic intensity of 7 on December 22, 2029 (Future in the table), the algorithm predicts that there is a possibility that an earthquake of that predicted seismic intensity (7) will occur in the set area on this date and time with a probability of TP / (TP+FN).

[0251] Furthermore, if we look at the years and dates that are missing one or two from the total number of decision trees for the target seismic intensity, there are decision trees that do not fall within the defined range values. It is important to note that for these years and dates, although the probability is lower than TP / (TP+FN), there is still a possibility that an earthquake may actually occur due to the small number of earthquake occurrence samples.

[0252] Regarding experimental earthquake predictions, the experimental earthquake predictions will begin in 2021. The first version will be called Ver. Alpha, the second Ver. Beta, and the third Ver. Gamma. Ver. Alpha correctly predicted the seismic intensity 6+ earthquakes on February 13, 2021 (Fukushima), March 16, 2022 (Fukushima), and May 5, 2023 (one day off) (Ishikawa), but failed to predict the seismic intensity 7 earthquake in Ishikawa Prefecture on January 1, 2024. Therefore, the system was improved to Ver. β (changing the scope of the definition), and subsequently predicted a seismic intensity 6+ (actually a 6-) on April 17, 2024 (see the previous application of this application). However, because the probability was 20-50%, the system was improved to Ver. γ of the present invention by changing the unit from one day to one hour, etc., and it became possible to predict a seismic intensity 6-. The accuracy rate for the occurrence of a seismic intensity 6- was 33%, for a seismic intensity 6+, 59%, and for a seismic intensity 7, 100%. It is believed that some of these earthquakes will occur with this probability. Please check it out. The results in the next section show realistic future earthquake predictions for Japan with a seismic intensity of 6- or higher.

[0253] Seismic intensity 6 lower (33% accuracy rate, assuming one out of three predictions will be correct) 2025 (1) March 8th to July 5th (2) October 28th to November 16th 2026 (3) June 11th to October 1st (4) November 9th to November 13th 2027 (5) March 17th to November 30th 2028 (6) April 25th to May 19th (7) June 21st to December 17th

[0254] Seismic intensity 6+ (59% accuracy rate, 3 out of 5 predictions will be correct) 2027 (1) October 7th, 9pm ±1 hour 2028 (2) September 27th, 2-4pm and 8-9pm, ±1 hour each (3) October 26th, 9pm ±1 hour

[0255] Seismic intensity 7 (100% accuracy rate, 5 out of 5 predictions will be correct) None until 2028 *Due to the small number of samples, the probability will be slightly lower in the future. 2029 (1) December 22, 5:00 ± 1 hour

[0256] 6. RESULTS 6-1 Seismic Intensity 6 Lower Prediction and Results Figures 54 and 55 are tables explaining the prediction and results of a seismic intensity 6 lower. The left column shows the predicted date (time omitted) when a seismic intensity 6 lower was predicted to occur, and blank columns indicate the predicted non-occurrence period. The middle column shows the year, date, time, and region where a seismic intensity 6 lower actually occurred. The right column shows the evaluation of each prediction using the confusion matrix (TP, FN, TN, FP). Accuracy rate (overall hit rate): 0.67 Recall rate (occurrence hit rate): 0.33 Precision (capture rate): 1.00 Generally, people pay the most attention to the occurrence hit rate, and the higher the number, the better the performance. However, the inventor is committed to a capture rate of 1.00 (not missing an earthquake) and a high occurrence hit rate.

[0257] Only seismic intensity 6-weak earthquakes were predicted, and all earthquakes that occurred were of seismic intensity 6-weak. For example, it is possible that earthquakes known as aftershocks before and after seismic intensity 7 could also be predicted. In the previous versions, Ver. Alpha and Ver. Beta, when predicting seismic intensity 6-weak, seismic intensity 6-strong and seismic intensity 7 were treated as seismic intensity 6-weak or higher, but Ver. Gamma predicts them separately. For example, an intensity 7 earthquake occurred on January 17, 1995, but no predictions of intensity 6-weak or 6-strong earthquakes were made, and none have occurred. However, for other intensity 7 earthquakes, predictions of intensity 6-weak or 6-strong earthquakes have been made, and they have occurred. It is suspected that the strength of earthquake shaking (seismic intensity) is not the result of chance, but that seismic intensity 6-weak or 6-strong earthquakes are destined to occur, and the regression equation may be the answer.

[0258] Overall accuracy rate (accuracy of both predictions of occurrence and non-occurrence): 67% Accuracy rate of occurrence (occurrence / occurrence + number of predicted occurrences that did not occur): 33% Capture rate (number of predicted occurrences / number of predicted occurrences + number of occurrences that could not be predicted): 100% Captured all earthquakes of magnitude 6 lower Miss rate (number of non-occurrences / number of predicted occurrences): 67% It is imagined that two out of three predictions did not result in an earthquake occurring.

[0259] 6-2 Seismic Intensity 6+ Prediction and Results Figure 56 illustrates the prediction and results of a seismic intensity 6+. The left column shows the predicted year and date of a seismic intensity 6+ earthquake, with blank spaces indicating the predicted non-occurrence period. The middle column shows the actual date and region of the seismic intensity 6+ earthquake. The right column shows the evaluation of each prediction using the confusion matrix (TP, FN, TN, FP). Accuracy (overall accuracy): 0.80. Recall (occurrence accuracy): 0.59. Precision (capture rate): 1.00. Only seismic intensity 6+ earthquakes were predicted, and all but one earthquake that occurred were seismic intensity 6+. For example, it is possible that the system was able to predict earthquakes known as aftershocks before and after a seismic intensity 7 earthquake. In previous versions like Ver. α and Ver. β, when predicting seismic intensity 6+, seismic intensity 7 earthquakes were treated as seismic intensity 6+, but Ver. γ predicts them separately. For example, on January 17, 1995, there was an earthquake with a seismic intensity of 7, but no predictions of seismic intensity 6-weak or 6-upper were made, and none have occurred. However, for other earthquakes with a seismic intensity of 7, predictions of seismic intensity 6-weak or 6-upper have been made, and they have occurred. It is suspected that the strength of earthquake shaking (seismic intensity) is not the result of chance, and that there may be a reason why seismic intensity 6-weak or 6-upper occurs, and the regression equation may be the answer. Two seismic intensity 6-upper earthquakes have been predicted for the future, from October 2027 to 2028. Since the probability of occurrence is 59%, there is a high possibility that one out of two will occur. This is an opportunity to confirm this prediction method.

[0260] Overall hit rate (the hit rate for both occurrence and non-occurrence predictions): 80% Overall results Hit rate for occurrence (occurrence / occurrence + number of predicted occurrences that did not occur): 59% Capture rate (number of predicted occurrences / number of predicted occurrences + number of occurrences that could not be predicted): 100% Captured all seismic intensity 6+ earthquakes Miss rate (number of non-occurrences / number of predicted occurrences): 41%

[0261] 6-3 Seismic Intensity 7 Prediction and Results Figure 57 is a diagram explaining the prediction and results of a seismic intensity 7. The items on the left are the year and date when a seismic intensity 7 earthquake was predicted to occur, and blank spaces are predictions of the non-occurrence period assuming that no earthquake will occur. The items in the middle are the date, time, and region where a seismic intensity 7 earthquake actually occurred. The items on the right are evaluations using the confusion matrix (TP, FN, TN, FP) of each prediction. Accuracy rate (overall hit rate) 1.00 Recall rate (occurrence hit rate) 1.00 Precision (capture rate) 1.00 Only seismic intensity 7 earthquakes were predicted, and all earthquakes that occurred were of seismic intensity 7. When calculations were performed until a future prediction was made for only seismic intensity 7, one seismic intensity 7 earthquake was predicted around 5:00 a.m. on December 22, 2029.

[0262] Overall accuracy rate (accuracy for both predictions of occurrence and non-occurrence): 100% Accuracy rate of occurrence (occurrence / occurrence + number of predicted occurrences that did not occur): 100% This is like 7 earthquakes occurring out of 7 predictions Capture rate (number of predicted occurrences / number of predicted occurrences + number of occurrences that could not be predicted): 100% This is like all magnitude 7 earthquakes were captured Miss rate (number of non-occurrences / number of predicted occurrences): 0% This is like 0 earthquakes occurring out of 7 predictions.

[0263] 7. DISCUSSION Regarding the contents of DT1 to DT16 in Category 1. This not only explains the decision tree criteria, but also explains the mechanism of earthquakes, since the contents of the multiple decision trees can predict earthquakes with the above-mentioned probability. In other words, it can be said that when the global environment and epicenter are (1) to (3), an earthquake with a seismic intensity of 6- or higher will occur.

[0264] Category 1 Main Decision Tree Areas Related to the Gravitational Forces of the Sun and the Moon DT1-16 (1) When the gravitational scalars of the solar system bodies, especially the Sun and the Moon, that affect the Earth are 3.5 to 15% smaller than their respective maximum values. (2) When the gravitational forces of the Sun and the Moon that affect the Earth are decreasing on the vertical axis of the epicenter's surface. (3) When, inevitably, most of the gravitational forces of the Sun and the Moon that affect the Earth are acting on the horizontal axis.

[0265] Regarding Category 2, from DT19 to DT26, a regression equation for correlation was found between items (4) to (7) from the main decision tree, and it was found that the higher the seismic intensity of an earthquake, the smaller the residual (error) and the more regular it was. The study period was 30 years (298,058 hours), but for an example of a magnitude 7 earthquake, the regression decision tree alone narrowed the number of potential earthquake occurrence dates to 251 hours, less than 0.09% of the study period. Adding the Category 1 decision tree further reduced this to 148 hours. Adding Category 3 further reduced the number of potential earthquake occurrence dates for a total of 49 decision trees to 7 hours (= number of earthquakes), achieving a 100% accuracy rate. This is likely because the content of Category 1 is consistent with the earthquake mechanism.

[0266] Category 2: Fields of regression equations from the main decision tree (4) Correlation between the vertical axes of the moon and the sun (5) Correlation between the direction of the moon and the direction of the sun (6) Correlation between the angle on the Earth's orbit (= position = date) and the direction of the moon as seen from the Earth (lunar age) (7) Correlation between the interior angles of the moon and the sun and the gravitational force on the horizontal axis

[0267] Regarding Category 3 earthquakes DT27 to DT85, the probability of a magnitude 7 earthquake cannot be reached 100% unless Category 3 is added. Furthermore, the number of decision trees required to calculate the probability of each outcome is 90 for a magnitude 6-low earthquake and 97 for a magnitude 6-high earthquake, and the weaker the earthquake, the more likely it is that the planets in the solar system, especially Jupiter, are involved.

[0268] Category 3: Gravitational forces of solar system bodies (8) Gravitational forces of the sun, moon, Jupiter, and the planets as a whole on the vertical and horizontal axes (9) Data on gravitational forces, coefficients, and angles to explore the possibility that the statistical values ​​have boundaries and are correlated with earthquakes, even though the causal relationship is unclear

[0269] Regarding the statement that "earthquake prediction is impossible with current science," the content of this specification is based on classical physics, so it is not as difficult as the above description suggests. The strong impression is that the inventor was able to predict earthquakes because he (1) conducted statistical work on seismic intensity, not magnitude, and (2) conducted statistical work on gravitational forces.

[0270] Figure 58 compares the scalars for different statistical methods for seismic intensity and magnitude. For seismic intensity (top), a decision tree can be set for seismic intensity 6+ and seismic intensity 7, but not for magnitude.

[0271] Figure 59 compares the gravitational pull of the vertical axis between seismic intensity and magnitude using different statistical methods. For seismic intensity (top), a regression equation can be set at seismic intensity 7, but not for magnitude.

[0272] Figure 60 compares the gravitational pull of the horizontal axis between seismic intensity and magnitude statistical methods. As shown in the figure above, this prediction uses a support vector machine (boundary setting) method, which requires separating areas where the calculated values ​​for the target seismic intensity are 100% contained in the predicted region from those where they are not. However, this separation is not possible using magnitude. Therefore, it is not possible to find a regression equation that narrows the candidates to 0.09% for magnitude. Furthermore, if predictions are made using magnitude, which does not allow for the use of important decision trees, the prediction period becomes longer and the probability becomes significantly lower, making it impractical. Therefore, I do not see any point in using magnitude instead of seismic intensity. Since statistics using seismic intensity (acceleration) used in Japan are effective, I hope that more Japanese-style seismic intensity meters will be installed around the world, including in marine areas, to enable wider-area predictions.

[0273] 8. Conclusion Of the 103 decision trees DT1 to DT85, which are the main points of (1) to (9) below, the more "years, dates, and times" that fall within the earthquake occurrence definition ranges of seismic intensity 6-low to 7 in the future, the more likely it is that an earthquake will occur in the target prediction area with the probability noted in the results. (The possibility remains until all decision trees required for each seismic intensity have been reached, or until one or two trees are missing.)

[0274] Category 1 Primary Decision Tree Areas related to the gravitational forces of the Sun and the Moon (Hourly Units) (1) When the gravitational scalars of solar system bodies, especially the Sun and the Moon, affecting the Earth are 3.5 to 15% smaller than their respective maximum values. (2) When the gravitational forces of the Sun and the Moon affecting the Earth are decreasing on the vertical axis of the epicenter's surface. (3) When, inevitably, most of the gravitational forces of the Sun and the Moon affecting the Earth are acting on the horizontal axis.

[0275] Category 2 Fields of regression equations from the main decision tree (hourly units) (4) Vertical axis of the moon and sun (5) Direction of the moon and direction of the sun (6) Angle on the Earth's orbit (= position = date) and angle of the moon as seen from the Earth (≒ age of the moon) (7) Interior angle of the moon and sun and gravitational force on the horizontal axis From the main decision tree, we discovered a regression equation of correlation between items (4) to (7), and found that the greater the magnitude of the earthquake, the smaller and more regular the residuals.

[0276] Category 3: Gravitational forces of solar system celestial bodies (daily basis) (8) Gravitational forces of the sun, moon, Jupiter, and the planets as a whole on the vertical and horizontal axes (9) Gravitational and angular forces that are suspected to be correlated with earthquakes, although the causal relationship is unclear and statistical values ​​have boundaries

[0277] 9. Differences from Non-Patent Document 1, "Correlation between Earthquake Occurrence in the Japanese Archipelago and the Configurations of the Moon and the Sun"

[0278] 9-1 Points to Note First, the differences between the above Non-Patent Document 1 and the present application will be explained. Non-Patent Document 1 performs an analysis using data and magnitude from 1926 to 1986, while the present application performs an analysis using data and seismic intensity from 1995 to 2024 (magnitude does not produce good results). Although it was noted that this may be the cause of the discrepancy between the statistical results asserted in Non-Patent Document 1 and the statistical results of the present application, there were many questions, so the cause was investigated.

[0279] 9-2 Difference 1 Non-patent document 1 concludes that "When only one of the moon and the sun is aligned to the east or west, the frequency of earthquakes occurring at that location is not high. When both celestial bodies are aligned to the east or west at the same time, the frequency of earthquakes occurring at that location is high. Therefore, an analysis method that takes into account only the alignment of the moon is insufficient."

[0280] However, while it is advisable to consider the sun and other celestial bodies, this does not match the statistical data presented in this application.

[0281] Figure 61 is a table showing statistics on the number of earthquakes occurring according to the direction of the sun and moon, using data from the National Astronomical Observatory and the Japan Meteorological Agency, with Figure 61(a) showing earthquakes with a seismic intensity of 5 or higher and Figure 61(b) showing earthquakes with a magnitude of 5 or higher. For both seismic intensity and magnitude, the most frequent earthquakes occurred when "only one of the moons was aligned to the east or west," followed by "when both were aligned to the north or south," with extremely few earthquakes occurring when "both were aligned to the north or south."

[0282] After investigating the causes of the discrepancies, we found that: (1) Non-Patent Document 1 divides the data into ±30° intervals, which means that the east, west, north, and south directions cannot be divided equally; and (2) Non-Patent Document 1 states that "bar graphs are a misleading method of display," which likely means that the data was not counted accurately.

[0283] FIG. 62 is a diagram showing the classification of the present application and the classification of Non-Patent Document 1, where (a) shows the classification of the present application and (b) shows the classification of Non-Patent Document 1.

[0284] A. Division of the Application (See Figure 62(a)) (1) The area of ​​the larger rectangle is 1. (2) Divide each of the east, west, north, and south directions by ±45°, into 4 x 4 parts, for a total of 16 parts in terms of area. (3) Part a is "both east or west" with an area of ​​4 / 16. (4) Part b is "both south or north" with an area of ​​4 / 16. (5) Part c is "one side east or west = one side south or north" with an area of ​​8 / 16. (6) The discussion is about the difference in results, and whether "both east or west = a" is more common or "one side east or west = c" is more common, so the question is "which is more common, a or c?" (7) At first glance, the graph makes it seem like a has more, but c has twice the area, so if you count accurately, c will be slightly more.

[0285] B Division method in Non-Patent Document 1 (see Figure 62(b)) (1) Divided into 144 areas (not equal). (2) Part a is "both east or west" with an area of ​​16 / 144. (3) Part b is "both south or north" with an area of ​​16 / 144. (4) Part c is "one side east or west = one side south or north" with an area of ​​64 / 144. (5) Part d has an area of ​​48 / 144. (6) The discussion is about the difference in results, and whether there are more "both east or west = a" or "one side east or west = c", so the question is "which is more common, a or c?" (7) Even if I was mistaken, it feels like there are more c, but I should count accurately anyway, and the division is not equal.

[0286] Figure 63 is a diagram explaining the bar graph in the non-patent document. If we divide it into 30-degree increments and separate the moon and the sun as in the non-patent document, we should get a bar graph with a mixture of parts a, c, and c, as in the graph in Figure 63. However, the actual graph is colorless, leading to the misunderstanding that all "high-numbered angles" are areas where two types of lines intersect. Furthermore, part d, which does not belong to either category, appears, making it difficult to understand.

[0287] It should be divided into north, south, east, west (each ±45°), or if northeast, southeast, southwest, and northwest are also included, they should all be divided into the same conditions (each ±22.5°); if they are not divided evenly, it will not be a "comparison." For example, if you roll a die twice and try to find the probability of either 1 or 6 both times, or one time including either 1 or 6, it is similar to artificially deforming the shape of the die to make it more difficult to roll 2 to 5.

[0288] Matrix division (combination) is important in the evaluation process because it allows for the discovery of results that humans did not anticipate, and by dividing equally, the possibility of discovering unexpected results, such as "part d was actually the most numerous" (a fictional story), is eliminated before statistics are run due to human assumptions. Statistics is a method for finding the truth, and humans should not manipulate the results.

[0289] The same problem occurs when evaluating earthquake predictions. People unfamiliar with statistics tend to focus on only two things: whether an earthquake occurred during the earthquake prediction period, but in terms of a matrix, there are four types: "predicted, not predicted" and "occurred, not predicted," with a 2x2 matrix. It is also necessary to consider whether an earthquake occurred during the period when no predictions were made.

[0290] We speculate that the reason for this is that Non-Patent Document 1 was biased by looking at the concentrated point clouds of the black and white dispersion graph and did not check the number of matrix combinations. In either case, we conclude that the above is the cause of the difference in the statistical results from Non-Patent Document 1, and the statistical results of this application are correct.

[0291] 9-3 Difference 2 Non-patent document 1 concludes that "The magnitude of an earthquake is unrelated to the configuration of celestial bodies. Whether the magnitude is large or small, the correlation between the frequency of earthquakes and the configuration of celestial bodies is roughly the same. Therefore, it is a mistake to associate only large earthquakes with the configuration of celestial bodies."

[0292] The results of the analysis in this application show that if "magnitude" is changed to "seismic intensity," the expression becomes, "The magnitude of an earthquake's seismic intensity is closely related to the configuration of celestial bodies."

[0293] The pattern trends (somewhat vague) in the two items above certainly have a similar feel regardless of the magnitude of the earthquake, but the story changes when it comes to correlations (regression equations: more precise than trends). Regression analysis is a common method for finding correlations in statistics, and through regression analysis, the present application has been able to discover a regression equation for the correlation between the positions of the moon and the sun for earthquakes with a seismic intensity of 6-low to 7-high. The stronger the earthquake, the more elegant the equation, with residuals (errors) approaching 0. The smaller the seismic intensity, the larger the residuals, and it became impossible to create a regression equation for earthquakes with a seismic intensity of 5-high or lower. In other words, the greater the seismic intensity, the stronger the correlation.

[0294] 9-4 Difference 3 Non-Patent Document 1 concludes, "We narrowed the earthquake occurrence area to try to improve the correlation, but no significant increase was observed. This is thought to be because narrowing the area reduces the number of earthquakes that can be processed, increasing the statistical variance in the histogram and hiding the correlation." In the present study, it is difficult to find a correlation when statistics are taken by magnitude, but when statistics are taken by seismic intensity, a correlation can be confirmed to the extent that a regression equation for the correlation between the positions of the moon and the sun can be found through regression analysis even for earthquakes with a low occurrence frequency of seismic intensity 6-6 to 7. Because there is a large difference in the ease of finding this correlation, we are beginning to wonder if there is a problem with the method of calculating magnitude.

[0295] For example, if you try to pull out a wine cork with your bare hands using a spiral corkscrew, the more difficult the cork is to remove, the more energy it takes to do so.

[0296] Magnitude is an index (integral) of the "amount of energy required to cause an earthquake," and seismic intensity is the acceleration and speed (differential) of the movement of the earth's crust. However, the magnitude of the earthquake, which measures the energy required to pull out something that is "difficult to pull out," is calculated using three inaccurate factors: (1) the collapse area (the contact area between the cork and the bottle), (2) the amount of sliding (the length of the cork that needs to be pulled out), and (3) the rigidity of the area (the friction coefficient of the cork). I'm a little unsure about how accurate (1), (2), and (3) are.

[0297] In contrast, seismic intensity is measured by measuring acceleration and velocity using a machine, which has little inaccuracy. If the object is difficult to pull out, the acceleration and velocity after pulling out will increase, and the "difficulty of pulling out" will be proportional to the magnitude of the earthquake, making it reliable.

[0298] Professor Ide Satoshi of the University of Tokyo writes in his book that "magnitudes are inaccurate," and there are similar statements on the Japan Meteorological Agency website, so this is something we should consider as a possibility. Reference: Japan Meteorological Agency website https: / / www.data.jma.go.jp / eqev / data / joho / info_magnitude.html

[0299] 9-5 Difference 4 The conclusion of Non-Patent Document 1 states, "A: This study found that earthquakes occur frequently when the moon and sun are both aligned east or west at a given location. B: This can be easily explained by the fact that the tidal action of celestial bodies on the Earth's bedrock is at its greatest at such times, causing stress-accumulating bedrock to collapse and resulting in an earthquake." This seems to be the most important conclusion in Non-Patent Document 1, but I disagree with part A. The reason is as stated in item 2, difference 1. Regarding part B: Tidal force is incorrect. Tidal force is a force that only exists when there is a difference in distance between the Earth's surface and its center. As long as it is on the Earth's surface, it acts only vertically. Tidal force is at its greatest when high tide occurs, which occurs when the moon is at its noon, so it is almost in the south (there is a time difference). In contrast, when celestial bodies are aligned east or west, the gravitational force on the horizontal axis is strong, which is close to low tide, which is a contradiction.

[0300] As shown in Figure 5 above, the results of this study show that earthquakes occur when the vertical axis values ​​are small, whether it is gravitational force or tidal force. As shown in Figure 5, this is the exact opposite result to Non-Patent Document 1. This is evidence that the claim in Non-Patent Document 1 that "earthquakes occur when tidal force is at its maximum" is incorrect.

[0301] 9-6 Conclusion Overall, the only common point between the document and the present application is, in a broad sense, that "earthquakes are related to celestial bodies, particularly the moon and the sun." However, in the details, almost all of the key points in Non-Patent Document 1 (Differences 2-5) differ from the results of the present application, and the present invention is not an extension of the content asserted in Non-Patent Document 1 and cannot be considered prior art. The forces that the present application focuses on are particularly "changes in the gravitational force itself, and changes in the vertical and horizontal axes," and the assertion in Non-Patent Document 1 that "tidal forces are the greatest" is incorrect.

[0302] In conclusion, the contents of Non-Patent Document 1 do not constitute "what the present invention normally assumes," and even if we concede and allow that "they have in common the connection between the sun and the moon" and that "tidal forces are mistaken for gravitational forces," it can be said that the present invention is an advanced invention compared to Non-Patent Document 1, given that the accuracy rate of earthquake predictions using the present application's unique earthquake prediction decision tree is 100% for earthquakes with a seismic intensity of 7.

[0303] As described above, the present invention provides the following earthquake prediction method: (I) A method of earthquake prediction that narrows down a target prediction area, and uses specific coordinates in the field of view of the target prediction area, defines a range that includes errors between past earthquake occurrences and the regression equation for each target seismic intensity as a range in which an earthquake may occur, and determines that a future calculation result obtained using a similar method will indicate a possibility of an earthquake occurring in the target prediction area at the target seismic intensity for the year and date that falls within the definition.

[0304] (II) The coordinates include mathematical coordinates that consider the Earth's orbit as an angle, with the Sun as the origin, the perihelion as 180°, and the aphelion as 0°, or mathematical coordinates that consider the solar system as a plane when the Earth is the origin and the Sun is always fixed at 90°, and use the angle of each solar system body as seen from the Earth.

[0305] Supplementary note about (II): This is an explanation of special coordinates. In particular, when using decision trees related to the solar system planets, if the altitude and direction of the astronomical coordinates published by the observatory are used as is, the accuracy rate of occurrence of a magnitude 7 earthquake drops from 100% to 78%. This result suggests that this is not just a convenience for using trigonometric functions, but that coordinates based on perihelion and aphelion seem to be related to earthquakes, and since the solar system planets have almost horizontal axes, this may be evidence that gravitational forces, not tidal forces, are involved.

[0306] (III) The above definition is: (1) when the gravitational scalars of solar system bodies, especially the sun and moon, that affect the Earth are 3.5 to 15% smaller than their respective maximum values; (2) when the gravitational forces of the sun and moon that affect the Earth are reduced on the vertical axis of the surface of the epicenter; (3) when most of the gravitational forces of the sun and moon that affect the Earth inevitably act on the horizontal axis; (4) the correlation between the vertical axes of the moon and the sun; (5) the correlation between the direction of the moon and the direction of the sun; (6) the correlation between the angle on the Earth's orbit and the moon as seen from the Earth; (7) the correlation between the interior angles of the moon and the sun and the gravitational forces on the horizontal axis; (8) the gravitational relations between the vertical and horizontal axes of the sun, moon, Jupiter, and the planet as a whole; (9) the earthquake prediction method of (I) is determined from a decision tree based on gravitational forces, coefficients, and angle-related data to explore possible correlations with earthquakes.

[0307] Supplementary information about (III): Explanation of the calculation method for each decision tree (how to determine the definition, not the specific numerical values). The decision trees in Category 1 can be used not only as definitions but also as explanations of the earthquake mechanism.

[0308] (IV) The earthquake prediction method of (I) uses data exclusive to countries at the same latitude as Japan, and defines the range in which an earthquake may occur as the result of the regression equation, calculated by adding the maximum residual plus or minus any excess.

[0309] Supplementary information about (IV): This is a regression equation (with specific numerical values) based on data specific to Japan. It is thought to be applicable only to countries at the same latitude. China and the United States are the target countries.

[0310] (V) The earthquake prediction method of (I), in which the greater the number that falls within the range of the definition, the higher the probability of occurrence.

[0311] Supplementary information about (V): The survey period was 30 years (298,058 hours), but taking the example of a seismic intensity of 7, the candidate earthquake occurrence dates were narrowed down to 251 hours, less than 0.09% of the survey period, using only the Category 2 regression decision tree, and this was further reduced to 148 hours when the Category 1 decision tree was added. Furthermore, adding some Category 3, the total number of candidate earthquake occurrence dates that fall under the 49 decision trees is 7 hours (= number of earthquakes), achieving a 100% accuracy rate. In other words, with Categories 1 and 2 at the top, "the more applicable numbers, the higher the probability of occurrence." The total number of decision trees required (to reach the announced probability) is 97 for a seismic intensity of 6+ and 90 for a seismic intensity of 6-.

[0312] (VI) If a new earthquake occurs and it is outside the range of the existing definition, the range in which an earthquake may occur is redefined and revised to include the newly occurring earthquake, the range including the error of the regression equation and the occurrence of past earthquakes, according to the seismic intensity to be predicted. (I) Earthquake prediction method.

[0313] Supplementary information about (VI): The number of earthquakes (samples) that have occurred since scientific earthquake data became available is small, and there is currently no guarantee that all future earthquakes will fall within the scope of the current definition. We believe that updating the definition of the decision tree we discovered will enable more accurate earthquake predictions into the future, and the achievement of this research is the discovery of a decision tree and algorithm that makes earthquake prediction possible. It will take another 250 years, at least until the number of samples with a seismic intensity of 7 reaches 50, before more accurate results can be obtained.

[0314] The present invention also provides the following earthquake prediction method: (A) An earthquake prediction method including: an astronomical information collection step of collecting astronomical information for a predetermined period in the past and a predetermined period in the future, a parameter acquisition step of acquiring parameters based on the collected astronomical information, an earthquake occurrence range identification step of identifying an earthquake occurrence range that is within the range of the parameters during an earthquake occurrence period during which an earthquake actually occurred during the predetermined period in the past, and an earthquake prediction step of extracting a period during the predetermined period in the future during which the parameters are within the earthquake occurrence range and predicting that period as a period during which an earthquake is likely to occur.

[0315] (B) An earthquake prediction method according to (A), in which, in the astronomical information collection step, astronomical information is collected for a predetermined period in the past and a predetermined period in the future in a specific region on Earth that is subject to earthquake prediction, and in the earthquake prediction step, it is predicted that this is a period in which an earthquake may occur in the region that is subject to earthquake prediction, thereby enabling the identification of an earthquake-occurrence region.

[0316] (C) In the earthquake occurrence range identification step, the period during which an earthquake of a specific seismic intensity or a specific magnitude or greater occurs is defined as the earthquake occurrence period, thereby limiting the seismic intensity or magnitude to be investigated.

[0317] (D) The earthquake prediction method of (A), wherein the parameters are values ​​calculated based on the gravitational force or tidal force that a celestial body other than the Earth exerts on the Earth, or the distance, angle, or mass of the other celestial body relative to the Earth.

[0318] The earthquake prediction method according to (D), wherein the other celestial body is the moon, the sun, or a planet in the solar system.

[0319] (E) The parameters are: xy coordinates, which are geocentric coordinates with the center of the Earth as the origin and the position of the sun relative to the Earth fixed at 90° and in the y+ direction; or XY coordinates, which are heliocentric coordinates on a plane including the Earth's orbit, with the center of the Sun as the origin, the position of the Earth at aphelion at 0° relative to the origin, the position of the Earth at perihelion at 180°, the X-axis being a line passing through the position of the Earth at aphelion, the origin, and the position of the Earth at perihelion, and the Y-axis being a line passing through the origin and perpendicular to the X-axis; or, when calculating tidal forces, values ​​on coordinates obtained by converting astronomical direction and altitude into mathematical coordinates in which the horizontal plane of the prediction area in a predetermined time unit is on the X-axis, east is X+, west is X-, and altitude, which is a vertical plane, is on the Y-axis.

[0320] S1 Astronomical information collection step S2 Decision tree acquisition step S3 Earthquake range identification step S4 Earthquake prediction step

Claims

1. An earthquake prediction method in which the area to be predicted is narrowed down, and specific coordinates in the field of view in said prediction area are used, and the range including the error of the regression equation from the time of past earthquakes, for each predicted seismic intensity, is defined as the range in which an earthquake may occur, and in the future, calculation results obtained in a similar manner are deemed to indicate the possibility of an earthquake occurring in the prediction area at the predicted seismic intensity in the year and date that fall within the above definition, with the probability stated in the results.

2. The earthquake prediction method according to claim 1, wherein the coordinates include mathematical coordinates in which the sun is the origin, the perihelion is 180°, and the aphelion is 0°, and the angle of the earth's orbit is considered as an angle, or mathematical coordinates in which the earth is the origin, the sun is always fixed at 90°, the solar system is considered to be flat, and the angle of each solar system body as seen from the earth.

3. The earthquake prediction method of claim 1, wherein the definition is: (1) solar system bodies that affected the Earth at the time of past earthquakes, when the gravitational scalars of solar system bodies, particularly the sun and moon, that affected the Earth are 3.5 to 15% smaller than their respective maximum values; (2) when the gravitational forces of the sun and moon that affected the Earth are reduced on the vertical axis of the surface of the epicenter; (3) when, inevitably on the horizontal axis, most of the gravitational forces of the sun and moon that affected the Earth act on the horizontal axis; (4) the correlation between the vertical axes of the moon and the sun; (5) the correlation between the orientation of the moon and the orientation of the sun; (6) the correlation between the angle on the Earth's orbit and the moon as seen from the Earth; (7) the correlation between the interior angles of the moon and the sun and the gravitational forces on the horizontal axis; (8) the gravitational relations between the vertical and horizontal axes of the sun, moon, Jupiter, and the planet as a whole; and (9) the determination is made from a decision tree based on gravitational forces, coefficients, and angle-related data to explore possible correlations with earthquakes.

4. The earthquake prediction method according to claim 1, wherein the range in which an earthquake may occur is defined as the range obtained by adding the maximum residual plus or minus any excess to the results of the regression equation, using dedicated data for countries at the same latitude as Japan.

5. The earthquake prediction method according to claim 1, wherein the greater the number that falls within the range of the definition, the higher the probability of occurrence.

6. The earthquake prediction method described in claim 1, wherein if a new earthquake occurs and it is outside the range of the existing definition, the range in which an earthquake may occur is redefined and revised to include the error in the regression equation and the newly occurring earthquake, based on the seismic intensity to be predicted.

Images