JP7624128B2 - Earthquake prediction methods - Google Patents

Description

The present invention relates to an earthquake prediction method using a solar power generation system.

There are two types of earthquake prediction: electromagnetic earthquake prediction and mechanical earthquake prediction. The following Patent Document 1 and Non-Patent Document 1 are electromagnetic earthquake predictions. Patent Document 1 clarifies the relationship between electromagnetic anomalies and earthquakes, and measures electromagnetic noise to show that there is a high possibility of short-term earthquakes occurring near communication base stations with poor communication conditions. Considering the cost and difficulty of manufacturing the equipment, building the system, and analyzing the data, earthquake prediction using this method is not something that can be easily carried out by non-experts.

Non-Patent Document 1 states that before a huge earthquake, a disturbance occurred in the ionosphere, causing an abnormality in the total electron content of the ionosphere. It is said that an abnormality occurred in the total electron content of the ionosphere 40 minutes before the occurrence of the Tohoku Pacific Ocean Earthquake, so in the case of this earthquake, the abnormality occurred just before the earthquake. In addition, the occurrence of an abnormality in the total electron content of the ionosphere as a precursor to an earthquake may be limited to before the occurrence of a huge earthquake, and it is not clear whether it is possible to predict a large earthquake that is not as large as a huge earthquake but is expected to cause damage. In this specification, a huge earthquake refers to an earthquake with a magnitude (hereinafter referred to as M) of 8.0 or more, a large earthquake refers to an earthquake with a magnitude of 7.0 or more but less than M8.0, and a small or medium earthquake refers to an earthquake with a magnitude of less than M7.0.

Non-Patent Document 2 is a mechanical earthquake research, which predicts earthquakes by analyzing strain. It is an effective method for identifying areas where earthquakes occur in the medium to long term, but it is considered that short-term earthquake prediction is not easy.

Patent 5920810

Kosuke Heki, "Increasing ionospheric electrons just before a major earthquake" Earthquake Journal No. 53, pp. 19-25, Earthquake Prediction Research Promotion Association, June 2012 Takuya Nishimura, "Strain accumulation in the Nankai Trough earthquake source region," Earthquake Journal No. 62, pp. 1-7, Earthquake Prediction Research Promotion Association, December 2016

The problem to be solved by the present invention is to provide a method for easily predicting short-term earthquakes even by non-experts. If a solar power generation system is available, new investment is not required, and even non-experts can easily predict short-term earthquakes using numerical values. In the case of an abnormal value, it is estimated that there is a high probability that an earthquake of a certain magnitude or more will occur near Japan within two days. In addition, even non-experts can predict short-term earthquakes using numerical values based on data from solar power generation systems (hereinafter referred to as multiple distributed systems) that are installed in multiple locations with clear solar cell capacity and can accurately measure solar radiation intensity and generated power. If the measurement results are determined to be in an abnormal state, it is possible to estimate that there is a high probability that an earthquake will occur in the short term in an area close to the abnormal value measurement location. Although both claims are short-term earthquake predictions, earthquake prediction is not possible until just before the occurrence of an earthquake, and in many cases prediction is possible before that. In addition, even if it is not a huge earthquake, it is often possible to predict large earthquakes and small and medium-sized earthquakes of M4.5 or more but less than M7. The accuracy of some of the information about when, where, and how large an earthquake will occur cannot be said to be high, and it is considered to be a prediction rather than a forecast, but there is a possibility that it can be improved to the level of prediction in terms of claims 2 and 3. Solar cell capacity refers to the total value of the nominal maximum output of the solar cell modules that make up a solar power generation system.

Photovoltaic power generation is a method of generating electricity by converting solar energy into electricity, and the nominal maximum output of a solar cell module is the value specified in JIS C8918 at an air mass of 1.5, an irradiance (solar radiation intensity) of 1 kW/ ^m2 , and a module temperature of 25°C. According to the product pamphlet of the manufacturer of a home solar power generation system (hereinafter referred to as a home system), the generated power is at most about 70% to 80% of the solar cell capacity due to solar cell losses, i.e. losses due to rise in the surface temperature of the solar cell module, power conditioner losses, and other losses, and for a home system with a solar cell capacity of 3.7 kW, the maximum is about 2.59 kW to 2.96 kW.

In a home system, the power generation has been recorded to be much higher than 2.96kW, which is 80% of the solar cell capacity, and although there are factors that affect the power generation, such as the intensity of solar radiation, the temperature and degree of dirt on the surface of the solar cell module, and the angle at which sunlight hits the solar cell module, it is unlikely that the power generation will be much higher than that stated in the manufacturer's product brochure, even if favorable conditions are met. In solar power generation systems from other manufacturers, the maximum power generation is said to be around 70% to 80% of the solar cell capacity, and it is unlikely that the power generation will be much higher than this.

Solar radiation intensity is the strength of sunlight, and its unit is kW/ ^m2 . Solar radiation intensity on a clear day is said to be about 1kW/ ^m2 , and the power generated in a solar power generation system is approximately proportional to the solar radiation intensity. In a solar power generation system, the power generated can be measured using an attached monitor, and in a multiple distributed installation system, in addition to the power generated, the solar radiation intensity can be measured using an actinometer installed at the same angle as the solar cell module installation angle. The nominal maximum output of a solar cell module is a value at an irradiance of 1kW/ ^m2 , the power generated is approximately proportional to the solar radiation intensity, and the power generated is said to be 80% or less of the solar cell capacity at most. Therefore, in a multiple distributed installation system, the ratio of the value of the power generated proportionally converted to the solar cell capacity when the solar radiation intensity is 1kW/ ^m2 is usually 80% or less.

The appearance of power generation that far exceeds the contents of the manufacturer's pamphlet in a solar power generation system means that there are cases where a larger power generation occurs even with the same solar radiation intensity. In solar power generation systems in commercial facilities equipped with a solar pyranometer, and in solar power generation systems equipped with a solar pyranometer whose power generation data is published online, it is thought that there is a mismatch in the measurement timing of the power generation and the solar radiation intensity, and uneven solar radiation intensity occurs due to the influence of clouds, and the measurement results of the power generation and solar radiation intensity may not be used directly for earthquake prediction. However, looking at the data in stores and posted online, it can be seen that the change in the ratio of the power generation to the solar radiation intensity is gradual, but the range of fluctuation is large. It is theoretically impossible to explain the large difference in power generation even with the ^same solar radiation intensity, and the large fluctuation in the ratio of the value of the power generation proportionally converted to the solar cell capacity when the solar radiation intensity is 1 kW/m2, and indicates the existence of some unexplained cause that increases the power generation.

The device required in the present invention is a general solar power generation system equipped with a monitor that displays the generated power, as described in claim 1. When the ratio of generated power to the capacity of the solar cell is equal to or greater than a preset first reference value, it is estimated that an earthquake of a certain magnitude or greater will occur in the short term.

In the multiple distributed installation system, multiple photovoltaic power generation systems with a function for measuring solar radiation intensity and generated power, a system for transmitting data from each solar power generation system to a collection location, and a system for predicting earthquakes by collecting and calculating the data at the collection location are required. Even in the case of solar radiation that appears to be the same to the naked eye, the maximum value of generated power in the same time period on one day and the next day may differ greatly. Since the probability of short-term earthquake occurrence is high when abnormal values appear, in the multiple distributed installation system, it is estimated that an earthquake of a certain scale or ^larger will occur in the short term if there are i or more photovoltaic power generation systems in which the ratio of the value of the generated power proportionally converted to the solar cell capacity when the solar radiation intensity is 1 kW/m2 is equal to or greater than a preset second reference value. Depending on how the second reference value is set, it is possible to predict earthquakes of a certain scale or larger, including small and medium earthquakes, or earthquakes limited to earthquakes of a large scale or larger. In addition, it is estimated that an earthquake will occur in the short term when the ratio of the number of measurement points with values equal to or greater than a preset third reference value to the total number of measurement points is equal to or greater than a preset fourth reference value. The third reference value is set lower than the second reference value, and if the distribution of measurement points at or above the third reference value is large and widespread, it is estimated that an earthquake will occur in the short term. Depending on how the third reference value is set, it is possible to predict earthquakes of a certain scale or above, including small and medium-sized earthquakes, or to predict earthquakes limited to large earthquakes or larger. In earthquake predictions using multiple distributed installation systems, short-term often refers to up to two days after the earthquake, but the precursors of an earthquake of a large scale or larger may appear earlier.

The present invention enables even non-experts to easily predict short-term earthquakes, and also enables public institutions to issue earthquake prediction information.

The necessary matters for implementing the earthquake prediction method described in claim 1 are as follows.
First, the solar power generation system must have a monitor that displays the generated power. Second, a power conditioner with a rated output power equal to or greater than the solar cell capacity must be used. A power conditioner with a sufficient rated output power is required to predict earthquakes based on the magnitude of the generated power relative to the solar cell capacity. A high-rated power conditioner is preferable, but since this is mainly intended for home solar power generation systems, the minimum realistic conditions are shown. Third, the solar cell module must be installed facing south or close to south. The further away from the south the installation direction is, the smaller the maximum value of the generated power in a day becomes, and it becomes difficult to determine whether the generated power is abnormal when comparing the generated power with the solar cell capacity. Fourth, the installation angle must be around 30°. In the vicinity of Japan, extreme angles such as 0° and 90° mean that there are seasons when the angle between the sunlight and the solar cell module is not suitable for optimal power generation, and are not suitable for earthquake prediction throughout the year. In addition, if the angle is shallow, dirt on the surface of the solar cell module is not easily washed away by rain, and there is a possibility that the generated power will decrease due to surface dirt. Fifth, all solar cell modules must be installed in the same direction and at the same inclination angle. If the installation direction or angle differs depending on the solar cell module, power generation data under different conditions will be mixed, and it will be difficult to accurately determine whether the generated power is abnormal. Sixth, the maximum generated power on the day can be grasped and measured to one decimal place. Some products have a function to record the maximum generated power on the day, but even if this function is not available, visual confirmation on the monitor is also possible. In addition, in order to obtain highly accurate data, it is necessary to be able to measure to one decimal place. Seventh, there is no output control by the power company. Eighth, the solar cell module must not be shaded. Although uneven solar irradiance on the solar cell module due to the influence of cloud shadows is unavoidable, shadows caused by nearby buildings, utility poles, large trees, etc. are not suitable for generating power measurement. Ninth, there must be little dirt on the surface of the solar cell module. Dirt is usually washed away by rain, and although snow accumulation is unavoidable, fallen leaves and bird droppings must be removed if they are found. Tenth, in order to obtain highly accurate data, the solar cell capacity must be 3 kW or more. Eleventh, the solar power generation system must be within the last 15 years of installation and not deteriorated with time. As mentioned above, the 11th condition must be met.

The outline of the home system is that the solar cell module is installed facing south-southeast, with an inclination angle of 24.2°, the solar cell capacity is 3.7kW, the power conditioner rated output power is 4.0kW (actual value 4.1kW), and it is a polycrystalline silicon type, and it meets the necessary requirements. As a specific example, on the day of the M9.0 Tohoku Pacific Ocean Earthquake that occurred at 14:46 on March 11, 2011, a power generation of 4.1kW was recorded around noon before the earthquake. This is about 111% of the output of the solar cell capacity, and is the maximum value due to the constraints imposed by the rated output power of the power conditioner of the home system, and is an abnormally high power generation that is not normally possible. If there were no constraints on the power conditioner, it is possible that the power generation would have been even higher. Measurements using the home system were only started on March 11, 2011, and it is possible that the abnormally high power generation had occurred before the day before. In addition, many large aftershocks occurred after that, and a power generation of about 3.8kW was also recorded during those times. We hypothesized that abnormally high power generation is related to earthquakes of a certain size or larger, and investigated the relationship between power generation and earthquake occurrence. According to the manufacturer's pamphlet, the maximum power generation is 80% or less of the solar cell capacity, and even if favorable conditions such as the angle of sunlight irradiation to the solar cell module being close to a right angle, low temperature, and little dirt on the solar cell module surface after rain are met, it is considered that the generation of power of 111% of the solar cell capacity is not possible under normal conditions. Power generation is almost proportional to the solar radiation intensity, but the solar radiation intensity on a clear day is said to be about 1kW/ ^m2 , and the solar radiation intensity does not greatly exceed 1kW/ ^m2 as a precursor to an earthquake, and various losses also occur, so the generation of power of 111% of the solar cell capacity does not usually occur.

Therefore, the power generation of 90% of the solar cell capacity of the home system, i.e., 3.4 kW, was set as the first reference value in advance, and a power generation of 3.4 kW or more was defined as an abnormal value, and the relationship with earthquakes was examined. In addition, a case where the power generation is less than 80% of the solar cell capacity, 2.9 kW or less, even though the weather is fine and the measurement conditions are good near the home, was defined as a normal value, and the relationship with earthquakes was examined. The earthquakes to be examined are earthquakes of M4.5 or more that occurred near Japan, including the four northern islands, Sakhalin, and Taiwan. The reason for targeting earthquakes of M4.5 or more is that there are many earthquakes near Japan, and if smaller earthquakes are included, the number of target earthquakes will increase, making it difficult to understand the relationship between abnormal values and earthquakes. In addition, if the scale of the target earthquake is increased, the number of target earthquakes will decrease, and earthquakes corresponding to abnormal values are not necessarily large earthquakes, so in this case too, the relationship with earthquakes will be difficult to understand. For this reason, the scale of the target earthquakes for the examination was set to M4.5 or more.

The relationship between the occurrence of abnormal values of generated power in a home system and earthquakes of a certain scale or more is verified by three methods. When the generated power reaches 90% or more of the solar cell capacity, i.e., 3.4kW or more, which is the first reference value set in advance, there is a high probability that an earthquake of M4.5 or more will occur in the vicinity of Japan by the day after the next day. The first method is a comparison between the occurrence rate of earthquakes of M4.5 or more after the occurrence of an abnormal value and the occurrence rate of earthquakes of M4.5 or more after the occurrence of a normal value without abnormal values. The second method is a case study of individual correspondences between abnormal values and earthquakes of M4.5 or more. The third method is an analysis of the relationship between the occurrence frequency of earthquakes of M4.5 or more and the occurrence frequency of abnormal values. According to the Japan Meteorological Agency, earthquake prediction is a prediction that accurately limits the three elements of time, place, and size of an earthquake. Claim 1 of the present invention does not fully satisfy this in terms of location and size, so it is earthquake forecast rather than earthquake prediction. However, since it is clear that there is a relationship between the occurrence of abnormal values in a solar power generation system and the occurrence of earthquakes of a certain scale or more, it is significant for the realization of future earthquake prediction.

First, we will verify the occurrence of earthquakes of M4.5 or more after the appearance of abnormal values in the home system. In 2020, measurements were taken on the home system for 253 days, of which 170 were sunny days with good measurement conditions, of which 60 were days when abnormal values appeared, and 57 were days when the power generation was 2.9 kW or less despite good weather and good measurement conditions. The number of days with earthquakes of M4.5 or more occurring on the day, the day after, and the day after the day after the appearance of abnormal values was 15 days on the day, 31 days until the day after (15 on the day, 16 the day after), and 42 days until the day after (15 on the day, 16 the day after, 11 the day after), respectively, and the probability of an earthquake of M4.5 or more occurring on the day was 25.0% (15 days out of 60 days), the probability of occurrence by the day after was 51.7% (31 days out of 60 days), and the probability of occurrence by the day after was 70.0% (42 days out of 60 days). In contrast, in the case of a home system where the weather is good and the measurement conditions are good but the power generation is 2.9kW or less, the number of days when earthquakes of M4.5 or more occur is 7 days on the day, 20 days until the next day (7 on the day, 13 the next day), and 30 days until the day after (7 on the day, 13 the next day, 10 the day after), and the occurrence probability is 12.3% (7 days out of 57 days), 35.1% (20 days out of 57 days), and 52.6% (30 days out of 57 days), respectively. In comparison, 25.0% > 12.3%, 51.7% > 35.1%, and 70.0% > 52.6%, and in all cases, the occurrence rate of earthquakes of M4.5 or more is higher when abnormal values appear in the home system than when normal values appear. Note that if multiple earthquakes of M4.5 or more occur by the day after the abnormal value appears, only the first earthquake of M4.5 or more is counted.

Similar verification is being continued for 2021. Measurements were taken on the home system from January to October on 204 days, of which 131 were good measurement conditions, 49 of which were abnormal values, and 53 were good measurement conditions but the power generation was 2.9 kW or less. The number of earthquakes of magnitude 4.5 or more on the day, the day after, and the day after the day after the abnormal value appeared was 18 on the day, 38 on the day (18 on the day, 20 on the day after), and 42 on the day after the day after (18 on the day, 20 on the day after, and 4 on the day after), respectively, and the probability of an earthquake of magnitude 4.5 or more occurring on the day was 36.7% (18 out of 49 days), the probability of occurrence by the next day was 77.6% (38 out of 49 days), and the probability of occurrence by the day after the day was 85.7% (42 out of 49 days). In contrast, when the measurement conditions are good but the power generation is 2.9kW or less, the number of days when earthquakes of magnitude 4.5 or greater occur is 16 on the day, 32 days until the next day (16 on the day, 16 the next day), and 39 days until the day after (16 on the day, 16 the next day, 7 the day after), with occurrence probabilities of 30.2% (16 out of 53 days), 60.4% (32 out of 53 days), and 73.6% (39 out of 53 days), respectively. In comparison, the results are 36.7% > 30.2%, 77.6% > 60.4%, and 85.7% > 73.6%, and in all cases the occurrence rate of earthquakes of magnitude 4.5 or greater is higher when abnormal values appear in the home system than when normal values appear.

Secondly, we will provide examples of the relationship between abnormal values in a home system and earthquakes of a certain magnitude or larger. We have already mentioned that 4.1 kW (about 111% of the solar cell capacity) was recorded around noon on the day before the Tohoku Pacific Ocean Earthquake occurred, but we will provide five other examples.

The first example is the M6.4 Hualien earthquake that occurred in eastern Taiwan at 00:50 on February 7, 2018 (Japan time), followed by a M6.1 earthquake in the early morning of February 8. Measurement results at that time included 3.8kW (approximately 103% of the original) on February 5, two days before the earthquake, 4.0kW (approximately 108% of the original) on February 6, the day before, and 3.8kW (approximately 103% of the original) on February 7, the day before the subsequent M6.1 earthquake, and these significant abnormal values are thought to have been precursors to the earthquake in question.

The second example is a magnitude 5.6 earthquake that occurred at 2:25 a.m. on April 28, 2019, in the southern Tokachi region. An abnormal value of 3.5 kW (approximately 95%) was recorded on the previous day, April 27.

The third example is a magnitude 5.8 earthquake that occurred off the east coast of Hokkaido at 6:23 a.m. on March 3, 2021, recording an abnormal value of 3.9 kW (approximately 105%) on the previous day, March 2.

The fourth example describes an earthquake that was not over M4.5, but was located off the north coast of Nata, Higashi-ku, Fukuoka City, near my home, about 10 km from my home to the epicenter. A M3.8 earthquake occurred at 20:42 on March 16, 2020, and my home system recorded a significant abnormal value of 4.0 kW (about 108%) before the earthquake on that day. This is thought to be an example of an earthquake that is not large in scale, but has a large precursor if the epicenter is close.

The fifth example is a magnitude 6.6 earthquake that occurred at 00:45 on November 11, 2021, southeast of the main island of Okinawa, and recorded an abnormal value of 3.8 kW (approximately 103%) on the previous day, November 10. All five of the above examples indicate that when power generation exceeds the first preset reference value, a magnitude 4.5 or higher earthquake or something similar occurs near Japan within the next two days.

Thirdly, we will discuss the fact that there are many anomalous values in months with many earthquakes of magnitude 4.5 or greater.
In 2020, April and July are months with many earthquakes of a certain scale or greater. Conversely, November is a month with few earthquakes of a certain scale or greater. The situation for these months will be explained. These do not directly indicate the relationship between the occurrence of power generation exceeding the first reference value set in advance and earthquakes of M4.5 or greater, but based on the relationship between the frequency of earthquakes of M4.5 or greater and the frequency of occurrence of abnormal values, they explain the certainty that when power generation exceeding the first reference value occurs, there is a high probability that an earthquake of M4.5 or greater will occur near Japan within two days.

In April 2020, there were 22 earthquakes of magnitude 4.5 or greater. Of the days when measurements were taken at home, there were 17 days with good weather and good measurement conditions, of which 8 days had abnormal values in the home system, which is a high rate of 47.1%. Of the 170 days in 2020 when the weather was good and suitable for measurements near the home, 60 days had abnormal values, which is 35.3%, so the abnormal value occurrence rate of 47.1% can be said to be high.

Even in July 2020, there were 17 earthquakes of magnitude 4.5 or greater, and while there were 9 days when measurement conditions were good, there were 5 days when abnormal values appeared in the home system, resulting in an abnormal value occurrence rate of 55.6%, which can be considered high.

November 2020 corresponds to a month with few earthquakes of a certain magnitude or above, with eight earthquakes of magnitude 4.5 or above, and two days in which abnormal values appeared in the home system compared to 14 days with good measurement conditions, resulting in an abnormal value occurrence rate of only 14.3%, which is a low value. This is an example that shows that there are fewer abnormal values in months with fewer earthquakes of a certain magnitude or above.

The three methods show that there is a relationship between the occurrence of abnormal values in the home system and earthquakes of a certain scale or larger. The above is an earthquake prediction method that estimates that an earthquake will occur when the ratio of generated power to solar cell capacity is equal to or greater than 90%, which is a first reference value set in advance. When generated power equal to or greater than the first reference value appears, it is possible to estimate that there is a high probability that an earthquake of M4.5 or greater will occur near Japan within two days after the occurrence of the first reference value. The high probability is 70.0% in 2020 and 85.7% from January to October 2021. Although this method has limitations in making highly accurate earthquake predictions, the earthquake prediction that there is a high probability of an earthquake of M4.5 or greater will occur near Japan within two days after the occurrence of an abnormal value cannot be said to be an earthquake prediction, but it can be said to be a meaningful earthquake prediction method that will lead to the realization of future earthquake predictions.

The earthquake prediction method described in claim 1 has several problems. First, because the weather is judged to be suitable for measurement by visual inspection, if this judgment is incorrect, accurate earthquake prediction is not possible. Second, because the solar radiation intensity is not measured, it is difficult to judge whether the generated power is abnormal in a time period when there is insufficient solar radiation or when there is solar radiation but the solar altitude is low and the generated power is small. However, for a time period when there is sufficient solar radiation and the solar altitude is not high, it is not impossible to predict by comparing with the past monthly and hourly generated power. Third, this method is mainly intended for measurement at one location, but the situation in other areas is unknown with measurement at one location, and it is not possible to grasp the precursors of a wide-area earthquake, and it cannot be said that the prediction accuracy is high, especially with regard to the scale and epicenter of the earthquake. Fourth, there are time periods when power is not generated depending on the direction of the sun, and power is not generated after sunset and before sunrise.

Earthquake prediction using a multiple distributed installation system solves many of these problems. By installing solar power generation systems that can measure solar radiation intensity and generated power at the same time over a wide area and at multiple locations, the following effects are achieved. First, earthquake prediction using objective numerical values is possible because values calculated from generated power, solar radiation intensity, and solar cell capacity are used for earthquake prediction. Second, if power is generated by sunlight, measurement is possible regardless of weather, and measurement time can be extended. Third, multi-point measurement makes it possible to determine the presence or absence of earthquake precursors over a wide area, and comprehensive earthquake prediction improves prediction accuracy, improving prediction accuracy for the scale and epicenter of earthquakes. The fourth problem of the earthquake prediction method described in claim 1 is not solved because even with a multiple distributed installation system, no power is generated after sunset and before sunrise, making measurement impossible. In addition, in the case of a solar cell module with a fixed installation angle, there are time periods when power is not generated depending on the direction of the sun even if there is solar radiation, making measurement impossible, but with a sun tracking type, this problem can be solved.

It has been found that in a solar power generation system, the ratio of the power generated on a fine day to the capacity of the solar cell is not almost constant but has a considerable range, and when the power generated is 90% or more of the capacity of the solar cell, there is a high probability of an earthquake of a certain scale or more occurring. In addition, it has been found that the power generated is almost proportional to the solar radiation intensity, and the nominal maximum output of a solar cell module is a value when the irradiance (solar radiation intensity) is 1kW/ ^m2 . For this reason, in a solar power generation system equipped with a solar pyranometer, the ratio of the value proportionally converted to the power generated at a solar radiation intensity of 1kW/ ^m2 to the capacity of the solar cell is not almost constant but fluctuates over a considerable range, and when this value is an abnormal value, there is a high probability of an earthquake of a certain scale or more occurring. If multiple solar power generation systems equipped with solar pyranometers are installed in a wide area by a multiple distributed installation system, the distribution state of the relationship between the solar radiation intensity and the power generated at a certain time can be grasped, making it possible to grasp the precursors of an earthquake.

The requirements for a multiple distributed installation system are as follows. First, in order to obtain reliable data by measuring under the same conditions, each system equipped with a solar photovoltaic power generation system and a solar pyranometer must have the same specifications. Second, the installation direction and inclination angle of the solar photovoltaic power generation system must be the same as that of the solar cell module. Third, in order to cover the earthquake prediction range and perform highly accurate earthquake prediction, as many solar photovoltaic power generation systems as possible must be installed in a wide area, including islands, with at least 20 solar photovoltaic power generation systems installed in Japan. Fourth, in order to generate more power than the solar cell capacity and to respond to the prediction of earthquakes of a magnitude larger than a major earthquake, a power conditioner with a rated output power of 130% or more of the solar cell capacity must be used. Fifth, the solar photovoltaic power generation system can be either a fixed type with a fixed solar cell module installation direction and angle, or a sun tracking type, but it must be unified into one of them. In the case of a fixed solar cell module, the installation angle must be the same inclination angle of about 30° for all solar cell modules, and the installation direction of the solar cell module must face south. Sixth, the solar cell module must not be shaded by buildings, utility poles, large trees, etc. Seventh, the surface of the solar cell module must be kept clean. Eighth, the system must not be subject to output control by the power company. Ninth, a thermometer must be installed. Since there is solar cell loss, i.e. loss due to the rise in the surface temperature of the solar cell module, a thermometer must be installed in preparation for making corrections. Tenth, in order to enable accurate observation according to Japan Standard Time, each solar power generation system must be capable of receiving standard radio waves, and must measure the solar radiation intensity and generated power at exactly the same timing at a frequency of at least once per minute, and transmit the data together with the measurement time and temperature to the collection location. Eleventh, data collection and calculation can be performed at the collection location. Twelfth, in order to obtain highly accurate data, the solar cell capacity must be approximately 50 kW or more. Thirteenth, the generated power (unit: kW) must be measurable to one decimal place, and the solar radiation intensity (unit: kW/ ^m2 ) to two decimal places. Fourteenth, the solar power generation system must be installed within 15 years and not deteriorated with time. Fifteenth, in order to respond to even more advanced earthquake predictions in the future, a database of the sun direction and solar altitude for each day and time at each measurement point will be kept at the collection site.The above condition 15 must be met.

This section describes a method for calculating the values required for short-term earthquake prediction in a multiple distributed installation system. The solar cell capacity (kW) of each solar power generation system is a, the solar radiation intensity (kW/ ^m2 ) at a certain time is b, and the generated power (kW) is c. Since the generated power is almost proportional to the solar radiation intensity, it is assumed to be completely proportional, and the value of "what percentage of the generated power (kW) converted when the solar radiation intensity is 1kW/ ^m2 is compared to the solar cell capacity a (kW)" is d (%). The value of d can be calculated by d = ((c/b)/a)) x 100.

The nominal maximum output of a solar cell module is said to be the irradiance, i.e. the power generated when the solar radiation intensity is 1kW/ ^m2 , and since the power generated is approximately proportional to the solar radiation intensity, the value of d indicates what percentage of the solar cell capacity the power generated is when converted to a solar radiation intensity of 1kW/ ^m2 . For example, if the solar cell capacity is 50kW, the solar radiation intensity is 0.50kW/ ^m2 , and the power generated is 20.0kW, the value of d is 80.0%. d is calculated to one decimal place. Since the power generated is said to be a maximum of 80% of the solar cell capacity, the value of d is usually 80% or less.

In order to maintain the accuracy of the value of d, the following two steps are to be performed. First, measurement data where the solar radiation intensity is less than 0.10 kW/ ^m2 is to be invalidated and excluded. Second, if the value of d does not change gradually and it is suspected that the solar radiation intensity and generated power are not measured at the same time, or if the solar radiation intensity on the solar cell module is suspected to be uneven due to the influence of cloud shadows, or if accurate data cannot be obtained due to snow accumulation or other reasons, the relevant data is to be invalidated and excluded. The above is necessary.

As for whether or not the value of d needs to be corrected to take into account solar cell losses, i.e. losses due to increases in the surface temperature of the solar cell module, there is insufficient data to be accumulated, so it is not possible to say definitively. A thermometer will be installed in each solar power generation system that constitutes a multiple distributed installation system, and temperature data will be transmitted to the collection location and used as necessary. Data will be accumulated, and if necessary, corrections will be made by setting correction coefficients for each region and month. In addition, since it is believed that there is some cause between the sun and the earth that increases the generated power before an earthquake, a database of the direction and altitude of the sun by date and time at each measurement point will be kept at the collection location, and will be used to correct the value of d in the future.

There are many unknowns, such as whether the distribution of d values when an earthquake is imminent will be a contour line with a single peak or a scattered distribution of high values, how far above the ground the cause of the increase in generated power is, and in which direction from the epicenter the cause of the increase in generated power will be the greatest. In any case, it is estimated that the larger the earthquake, the more locations there will be where the value of d as a precursor is high, and that there will be a greater number of locations showing high values, and that the precursors will appear earlier than for small and medium-sized earthquakes. The second and third reference values can be set according to the scale of the earthquake to be predicted.

When predicting earthquakes of a magnitude greater than or equal to a major earthquake, in consideration of the fact that a home system recorded a power generation of approximately 111% of the solar cell capacity before the Tohoku Pacific Ocean Earthquake, as an example of a prediction, if there is a point where the value of d is 110% or more at a certain time, it can be assumed that an earthquake of a magnitude greater than or equal to a major earthquake will occur in a location relatively close to the point where the value of d is the highest in the short term. Also, if the number of points where the maximum value of d does not reach 110% but is 100% or more is 25% or more of the total, it can be assumed that an earthquake of a magnitude greater than or equal to a major earthquake will occur in a location relatively close to the point where the value of d is high in the short term. The 110% given as an example is the second standard value, 100% is the third standard value, and 25% is the fourth standard value. For example, if the value of d is 95.0% in Tokyo, 100.0% in Osaka, 120.0% in Kochi, and 110.0% in Miyazaki around noon on a certain day, there are points where the value of d is 110% or more, so it is predicted that an earthquake of a magnitude greater than a major earthquake will occur in the short term, and the epicenter is likely to be off the southern coast of Shikoku to the southern coast of Kyushu. Also, if the value of d is not 110% or more but 25% or more of the total measurement points have a measurement value of 100% or more, and there is a spread in the value of d, it is predicted that an earthquake of a magnitude greater than a major earthquake will occur in the short term. For example, if there are 20 measurement points, and around noon on a certain day, the value of d is 105% in Kochi and Miyazaki, and 100% in the other three measurement points west of Kinki, the maximum value of d is not 110% or more, but there are five measurement points with a value of 100% or more, which is 25% of the total, and it is also predicted that an earthquake of a magnitude greater than a major earthquake will occur in the short term with the epicenter off the southern coast of Shikoku to the southern coast of Kyushu. Short-term often means up to two days later, but warning signs of a large or larger earthquake may appear earlier.

When predicting small and medium earthquakes of less than M7.0, for example, if the second standard value is set to 95%, it becomes possible to estimate the occurrence of small and medium earthquakes of about M4.5 or more, and if the third standard value is set to 90% and the fourth standard value to 25%, it becomes possible to similarly estimate the occurrence of small and medium earthquakes of about M4.5 or more. The scale of an earthquake is thought to be roughly linked to the maximum value of d, and the epicenter is thought to be close to a point with a high value of d.

Color-coded display according to the value of d is also one way to make earthquake predictions easier to understand. The magnitude of the precursors of earthquakes is defined as follows: 110% or more is extremely abnormal, 100% or more but less than 110% is extremely abnormal, 95% or more but less than 100% is abnormal, 90% or more but less than 95% is slightly abnormal, 80% or more but less than 90% is high, 70% or more but less than 80% is normal, and less than 70% is low. The colors red, orange, yellow, yellow-green, green, light blue, and blue are assigned to the respective values, and the color-coded display can be displayed on computers and smartphones via the Internet, or on television by public organizations and weather information companies, allowing the magnitude of the precursors of earthquakes of a magnitude of large or larger to be communicated. In addition, calculation results at the collection site can be used by public organizations to issue earthquake prediction information, and can be provided to weather information companies to provide earthquake prediction information services. Furthermore, the publication of the data allows even non-experts to easily make short-term earthquake predictions.

Regarding short-term earthquake prediction in multiple distributed installation systems, the use of data from commercial facility storefronts or online data is not permitted, so the concept of earthquake prediction is described. It is not possible to say definitively what the appropriate standard value is, but the standard value will be revised as necessary based on the relationship between measurement results and earthquakes that have occurred, in order to improve the accuracy of earthquake prediction. From the measurement results of the home system, short-term does not mean that it is not possible to predict an earthquake of a magnitude greater than a major earthquake until it occurs, but rather that it is highly likely that it can be predicted before that, and the larger the earthquake, the earlier the precursors appear. Although both claims 2 and 3 are earthquake prediction rather than earthquake prediction, it is possible that the accuracy of earthquake prediction can be improved to earthquake prediction by adding corrections for the value of d and its distribution, as well as temperature, the direction of the sun, and altitude.

Not only will it enable weather information companies to provide earthquake prediction information services, but it will also be possible to provide home solar power generation systems with earthquake prediction functions. If it becomes possible to measure generated power and solar radiation intensity with high precision, it will lead to the commercialization of earthquake prediction devices such as watches, smartphones, and cars with earthquake prediction functions, which could have industrial applications.

Claims (3)

An earthquake prediction method that measures the power generated by a solar power generation system and predicts that an earthquake will occur if the ratio of the power generated to the solar cell capacity of the solar power generation system is equal to or greater than a predetermined first reference value. An earthquake prediction ^method that measures the generated power and solar radiation intensity by solar power generation systems installed in multiple locations, transmits the data of the measurement time, generated power, and solar radiation intensity to a data collection location, and predicts that an earthquake will occur if there is one or more solar power generation systems where the ratio of the generated power proportionally converted to a solar radiation intensity of 1 kW/m2 to the solar cell capacity is equal to or greater than a preset second reference value. The earthquake prediction method described in claim 2, which estimates that an earthquake will occur when the proportion of measurement points with values equal to or greater than a predetermined third standard value to the total number of measurement points is equal to or greater than a predetermined fourth standard value.