Exploratory Statistical Analysis of Precursors to Moderate Earthquakes in Japan

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please check the attached file.

Comments for author File: Comments.pdf

Author Response

Thank you for your thoughtful review. In response to your questions, I have revised the main text and added further information.
I shall answer your questions in a question-and-answer format below.

>The 11 March 2011 Tohoku megathrust 9.0 earthquake presented clear precursory patterns: shortened foreshock intervals, increased magnitudes and a pronounced event swarm immediately before the main event. In this paper, the author attempts to find a possible usual law for earthquake prediction by a statistical method of the frequency and magnitude of foreshocks increasing preceding the major event mainly in Japan mainland. Several strong to large events have been evaluated utilizing this method and the results show that almost all earthquakes with a magnitude equal to or more than 7.0 are preceding this pattern of increased magnitude and occurrence frequency of foreshocks in the near region and most of them are with a swarm shocks, which indicate a potential apply of seismic risk assessment and even earthquake prediction.
>However, the author also finds that:
>1. The swarm pattern does not occur before all considered events and the time interval of this phenomenon from the main shock has not confirmed yet. The installments or the detectability in an area put a heavy influence on the completeness of the earthquake, which gives a subsequent effect on the statistical results.
> So the questions rise:
>1.The author gives the study area in Figure 1. While, for each considered event in this paper, the considered region, where so-called foreshocks occur, has not clearly specified, along the main fault? or a fixed area calculated in a way? The preparation area is usually large if plate relative motion is considered. There are also some ways to specify the affected area for a large event.

This is a perfectly reasonable question. In the present study, the scope has been defined using the regional classification employed by the Japan Meteorological Agency (JMA), with the sole exception of the Iburi classification, which combines adjacent regions. We have now added an explicit explanation of this in the revised text.  

In addition, I am currently submitting a paper that introduces a methodology based on a mesh with one-degree latitude and longitude resolution. In this framework, the scope corresponds to an approximately 100 km square isosceles trapezoid, which has proven effective [15]. Furthermore, I shall soon submit a paper detailing a method for focusing observation on a single plate, which is likely the most logical approach, albeit somewhat laborious. This will be available shortly.

>2. Except for monitoring ability, the question that whether the focal mechanism of the event could affect the precursory pattern or not has not been mentioned in the paper. As we all known, there are not clear foreshock activities preceding the 2008 Wenchuan 8.0 earthquake in China.

This point has been noted in the discussion. Unfortunately, I have been unable to obtain sufficiently detailed data concerning the 2008 Wenchuan 8.0 earthquake in China. To the best of my knowledge, no organisation publishes such data as comprehensively as the Japan Meteorological Agency (JMA). For this reason, I cannot provide a meaningful analysis of that event within the scope of the present study.  

>3.It is necessary for the author to give a simple specified statistical result on the swarm time interval variations as a function of all considered events in this paper. To some sense, a big foreshock could be considered as a main shock in a specified time period.
>The author needs to make some clear statements or discussion on these questions.

We acknowledge the reviewer’s question. In the present study, I have followed the designation of the Japan Meteorological Agency (JMA) regarding which event is considered the main shock, and this has been noted in the Materials section.  

The forthcoming paper that I am preparing addresses the statistical behaviour of aftershock decay, as the duration of aftershocks and the possibility of subsequent major earthquakes remain critical concerns. I believe that any deviation from the expected decay pattern is likely to indicate that the danger has not yet passed. This will be discussed in detail in that work, which I hope the reviewer will have the opportunity to consult.

Reviewer 2 Report

Comments and Suggestions for Authors

As a researcher in geophysics, I acknowledge that the issue of precursory earthquakes is highly controversial in the academic community. I agree that the statistical approach proposed by the author has merit and was well applied to the 2011 Tohoku earthquake. However, to enhance the robustness of the method, I believe the author needs to address at least the following three questions.

1. Aside from the 2011 Tohoku earthquake (M9), are there other earthquakes of M8 or higher that show similarly clear statistical correlations? For example, the recently occurred Kamchatka earthquakes. Focusing on Japan, why does the 2003 Tokachi earthquake not show records of precursory swarms? By 2003, records of precursory swarms should have been reliably available.

2. The second question is: how can one distinguish between precursory swarms and aftershocks? Often, if an earthquake of around M6–7 occurs and is followed by a cluster of earthquakes, we would typically consider this as aftershocks rather than precursory swarms (unless a larger earthquake actually occurs later). However, this approach can seem somewhat retrospective. How do the authors view this issue?

3. This issue may be beyond the scope of the authors’ study. As the authors noted, the method used in this paper relies on relatively simple measurements. What underlying physical mechanisms exist, and whether they can provide a corresponding explanation, remains an open question.

Author Response

> Suggestions for Authors
>As a researcher in geophysics, I acknowledge that the issue of precursory earthquakes is highly controversial in the academic community. I agree that the statistical approach proposed by the author has merit and was well applied to the 2011 Tohoku earthquake. However, to enhance the robustness of the method, I believe the author needs to address at least the following three questions.

Thank you for your thoughtful and constructive review. While revising the main text, I have addressed your points in a question-and-answer format below.

>1. Aside from the 2011 Tohoku earthquake (M9), are there other earthquakes of M8 or higher that show similarly clear statistical correlations? For example, the recently occurred Kamchatka earthquakes. Focusing on Japan, why does the 2003 Tokachi earthquake not show records of precursory swarms? By 2003, records of precursory swarms should have been reliably available.

Unfortunately, such data are unavailable. While the Japan Meteorological Agency collects data up to that magnitude, Kamchatka appears to be somewhat beyond their coverage. Few organisations publish data as comprehensively as the Japan Meteorological Agency, likely because Japan experiences a very high frequency of earthquakes.

Regarding the Tokachi offshore event, no records of precursory swarms were identified, possibly due to limitations in measurement capabilities. Similarly, the 2022 Fukushima Offshore earthquake (M7.3) could not be predicted, which I believe was because the epicentre was distant from land and thus from observation equipment. Ideally, prediction would always be possible, but achieving 100% accuracy appears difficult. This caution has been included in the discussion (L267).

>2. The second question is: how can one distinguish between precursory swarms and aftershocks? Often, if an earthquake of around M6–7 occurs and is followed by a cluster of earthquakes, we would typically consider this as aftershocks rather than precursory swarms (unless a larger earthquake actually occurs later). However, this approach can seem somewhat retrospective. How do the authors view this issue?

This is indeed a difficult issue. While aftershocks are occurring, the anomaly persists, and it is not necessarily clear whether the next major event will arrive immediately. I believe the only course is to remain vigilant. Incidentally, I have outlined how aftershocks decay in my next paper, which I shall be submitting shortly. Any movement deviating from this decay pattern would likely indicate danger.

>3. This issue may be beyond the scope of the authors’ study. As the authors noted, the method used in this paper relies on relatively simple measurements. What underlying physical mechanisms exist, and whether they can provide a corresponding explanation, remains an open question.

Quite right. As I mentioned in the discussion, elucidating the mechanisms behind such diverse measurements and predictions is probably beyond the scope of my study. I have added this point explicitly to L340.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses the relationship between seismic swarms and subsequent main shocks, analyzing Japanese earthquake catalogues for precursor phenomena. While the topic is relevant to earthquake forecasting, the paper does not present novel findings and it is not suitable for publication.

The fact that there is no unequivocal or biunique relationship between seismic swarm activity and the occurrence of major earthquakes is already well-established in the international seismological literature. Numerous studies have shown that both strong earthquakes preceded by swarms, and main shocks with no such precursors, are common in various tectonic environment (e.g. Arnaud 2014, Lippiello et al. 2019, Peng and Lei 2020, Zaccagnino et al 2024). The present work reiterates this point using Japanese data but does not advance existing knowledge with new analytical perspectives or practical forecasting tools or in any other way.

Here I summarize some of the principal weaknesses in Structure and Analysis:

• The manuscript is largely descriptive, lacking statistical rigor and clear hypothesis testing.
• There is no systematic comparison with other regional or global earthquake catalogues, nor does the paper engage critically with the wide body of research on precursor phenomena and seismic hazard assessment

• Patterns in timing and magnitude distributions are presented, but intervals between swarms and main shocks are highly variable and often not temporally proximate, undermining their predictive value.

Here I summarize the main Methodological Limitations:

• The approaches used (regional log-ratio, basic magnitude analysis) are simplistic relative to current standards in earthquake forecasting research. (see al the mentioned paper, but I could have cited many more)

• The results do not quantify the probability or conditional likelihood of main shock occurrence following observed swarms, nor do they suggest actionable statistical thresholds for prediction.
• The absence of any physical explanation or modeling for the observations or the predictive capability.

In summary, the manuscript provides a reiteration of a well-known result without significant methodological advancement or practical implications for seismic forecasting. The paper presents nothing new, does not present a critical review of international literature, completely disregards any kind of physical explanation, and adds nothing to the well-established knowledge on the subject.

 

Peng, Zhigang, and Xinglin Lei. "Physical mechanisms of earthquake nucleation and foreshocks: Cascade triggering, aseismic slip, or fluid flows?." Earthquake Research Advances 5.2 (2025): 100349.

Zaccagnino, Davide, et al. "Are foreshocks fore‐shocks?." Journal of Geophysical Research: Solid Earth 129.2 (2024): e2023JB027337.

Mignan, Arnaud. "The debate on the prognostic value of earthquake foreshocks: A meta-analysis." Scientific reports 4.1 (2014): 4099.

Lippiello, Eugenio, Cataldo Godano, and Lucilla De Arcangelis. "The relevance of foreshocks in earthquake triggering: A statistical study." Entropy 21.2 (2019): 173.

Author Response

>The manuscript addresses the relationship between seismic swarms and subsequent main shocks, analyzing Japanese earthquake catalogues for precursor phenomena. While the topic is relevant to earthquake forecasting, the paper does not present novel findings and it is not suitable for publication.

This is a rather harsh opinion, but I believe it is based on a clear misrepresentation of the facts. I should like to briefly point this out here.

>The fact that there is no unequivocal or biunique relationship between seismic swarm activity and the occurrence of major earthquakes is already well-established in the international seismological literature. Numerous studies have shown that both strong earthquakes preceded by swarms, and main shocks with no such precursors, are common in various tectonic environment (e.g. Arnaud 2014, Lippiello et al. 2019, Peng and Lei 2020, Zaccagnino et al 2024). The present work reiterates this point using Japanese data but does not advance existing knowledge with new analytical perspectives or practical forecasting tools or in any other way.

We appreciate the reviewer’s comments. However, we believe the assessment misrepresents the contribution of this work.  

As shown in my previous paper [2], I was the first to apply a rigorous statistical method, demonstrating that the long-accepted GR law did not hold. The present study builds on that foundation by re-examining Japanese earthquake catalogues, thereby contributing to the reassessment of established assumptions.  

While international literature has emphasised the lack of a biunique relationship between swarms and major earthquakes [1], my work shows that prediction was previously deemed impossible largely because appropriate statistical methods were not employed.  

This paper is not intended to criticise earlier limitations, but to demonstrate how correct methodology can yield new insights. I respectfully submit that re-examining conventional wisdom represents a constructive and scientifically sound contribution to earthquake forecasting.

>Here I summarize some of the principal weaknesses in Structure and Analysis:

>The manuscript is largely descriptive, lacking statistical rigor and clear hypothesis testing.
Can this be considered a criticism of not performing hypothesis testing? EDA emerged partly as a reaction against an overly hypothesis-testing approach to statistics. Testing is generally avoided unless unavoidable.
That is all. All the differences shown here are only those that are evident from the data. For example, one could subject the results in Fig. 6C to testing, but frankly, that exercise would be meaningless.
Repeated testing also raises the issue of multiple testing. I believe tests should only be performed when truly necessary, and I suspect most statisticians would agree.
I shall add the reason for not performing tests to L348.

>There is no systematic comparison with other regional or global earthquake catalogues, nor does the paper engage critically with the wide body of research on precursor phenomena and seismic hazard assessmentPatterns in timing and magnitude distributions are presented, but intervals between swarms and main shocks are highly variable and often not temporally proximate, undermining their predictive value.

I acknowledge the reviewer’s observation that intervals between swarms and main shocks are highly variable and not always temporally proximate, which indeed limits predictive value. However, this variability reflects the nature of the data itself and cannot be artificially constrained.  

Importantly, the present study represents the first successful detection of the anomaly in Japanese catalogues using a rigorous statistical approach. While prediction remains inherently difficult, we consider this contribution a meaningful step towards re-examining precursor phenomena with appropriate methodology.

>Here I summarize the main Methodological Limitations:

>The approaches used (regional log-ratio, basic magnitude analysis) are simplistic relative to current standards in earthquake forecasting research. (see al the mentioned paper, but I could have cited many more)

I acknowledge the reviewer’s comments regarding methodological limitations. Our choice to employ relatively simple statistical approaches was deliberate, as these methods ensure verifiability and transparency. More complex models, while commendable, often lack falsifiability and may distort analysis if flawed. We therefore consider the use of basic but rigorous statistical tools appropriate for the aims of this study.  

>The results do not quantify the probability or conditional likelihood of main shock occurrence following observed swarms, nor do they suggest actionable statistical thresholds for prediction.
>The absence of any physical explanation or modeling for the observations or the predictive capability.
>In summary, the manuscript provides a reiteration of a well-known result without significant methodological advancement or practical implications for seismic forecasting. The paper presents nothing new, does not present a critical review of international literature, completely disregards any kind of physical explanation, and adds nothing to the well-established knowledge on the subject.
>Peng, Zhigang, and Xinglin Lei. "Physical mechanisms of earthquake nucleation and foreshocks: Cascade triggering, aseismic slip, or fluid flows?." Earthquake Research Advances 5.2 (2025): 100349.
>Zaccagnino, Davide, et al. "Are foreshocks fore‐shocks?." Journal of Geophysical Research: Solid Earth 129.2 (2024): e2023JB027337.
>Mignan, Arnaud. "The debate on the prognostic value of earthquake foreshocks: A meta-analysis." Scientific reports 4.1 (2014): 4099.
>Lippiello, Eugenio, Cataldo Godano, and Lucilla De Arcangelis. "The relevance of foreshocks in earthquake triggering: A statistical study." Entropy 21.2 (2019): 173.

As indicated in the title, this paper focuses specifically on earthquakes in Japan, where comprehensive data are uniquely available through the Japan Meteorological Agency. The scope of the study is not to provide physical modelling of earthquake mechanisms, but rather to demonstrate predictive methods applicable in this context. We have noted in the introduction that elucidating mechanisms lies beyond the remit of the present work, and this will be addressed in a subsequent paper. We respectfully submit that evaluating the study against aims outside its stated scope is not constructive.

Reviewer 4 Report

Comments and Suggestions for Authors

There is a problem with this paper because many explanations are missing concerning the parameters which are displayed in the Figures.

The fact that some Figures are in a separate file (Figures S) is not justified.

Line 69- 70 “For instance, approximately half of the entries for magnitude 5 and 6 earthquakes pertain to overseas events. Efforts were made to exclude such non-domestic data wherever feasible.” What is the meaning of these sentences ? Do you exclude EQs occurring below the sea ? What is the meaning of “non-domestic data”? In general what is the minimum magnitude of the EQs you consider ? Do you take into account the depth ? It is impossible to understand areas which are considered to compare events related to one EQ and “other” as it is called in the Figures.

In Figure 1B what is the meaning of the parameter “Detection (log)” ? What is the exact signification of the dots and how they are determined ? What is the meaning of the parameter µ  and its value in Figure 1B ? Because there are many definition along the paper (line 15, elevated location; line 98, mean; line 173, location of magnitude). Why only the lowest 14 data points are considered in Figure 1B? What is the exact meaning of σ ? (from line 15 it is a scale). Is it a standard deviation?

Line 97, 2023 ?

In Figure S1 all panels are not labelled. In this Figure what are the meanings of Sorted Detection(log), ideal normal, ideal exponential, Normal Q-Q plot, theoretical quantiles ? Can you explain exactly what is the quantity which is represented in Figure S1B. What are the lines shown in the 4 panels of Figure S1 and how they are calculated ?

Line 111, S1D is related to Kumamoto not to Noto.

Line 114, what is the meaning of P-value ?

In Figure 2 what are the meanings of Proportion Nationwide, Sorted magnitude? Which areas are considered for Noto EQs and other EQs ?

Figure 3 is not correctly described and it is difficult to understand the parameters which are displayed. What is the meaning of λ ?

Line 211, not 2000 for S3D

Figures 6C and 6D are related to December 2021 ?

In Figure S5 problem of position for ideal normal

Line 273 replace magnitudes by energy

Line 358, what is a z-score ?

Finally, it will be interesting to compare your approach with the cumulative number of earthquakes (see for example papers by De Santis et al.)

De Santis, A., Abbattista, C., Alfonsi, L., Amoruso, L., Campuzano, S. A., Carbone, M., ... & Santoro, F. (2019). Geosystemics view of earthquakes. Entropy, 21(4), 412.

De Santis, A., Cianchini, G., Favali, P., Beranzoli, L., & Boschi, E. (2011). The Gutenberg–Richter law and entropy of earthquakes: Two case studies in Central Italy. Bulletin of the Seismological Society of America, 101(3), 1386-1395.

Author Response

>There is a problem with this paper because many explanations are missing concerning the parameters which are displayed in the Figures.

I thank the reviewer for these comments and acknowledge that additional clarification was needed.  
Regarding the parameters displayed in the Figures, I have now added explicit explanations in the text. For example, readers unfamiliar with normal distributions may not immediately recognise the meaning of “σ”; this has now been clarified.  

>The fact that some Figures are in a separate file (Figures S) is not justified.

Concerning the placement of Figures, we originally separated supplementary figures to avoid excessive length in the main text. These figures were intended to provide statistical background or illustrative examples. However, Figure S1 has now been moved into the main text, as it concerns fundamental statistical properties.  

>Line 69- 70 “For instance, approximately half of the entries for magnitude 5 and 6 earthquakes pertain to overseas events. Efforts were made to exclude such non-domestic data wherever feasible.” What is the meaning of these sentences ? Do you exclude EQs occurring below the sea ? What is the meaning of “non-domestic data”? In general what is the minimum magnitude of the EQs you consider ? Do you take into account the depth ? It is impossible to understand areas which are considered to compare events related to one EQ and “other” as it is called in the Figures.

With respect to Line 69–70, we have revised the wording to clarify that “non-domestic data” refers to earthquakes occurring outside Japan. Offshore events within Japanese territory are included, but overseas events were excluded wherever feasible. I have also specified that no data was ommited. The areas compared in the Figures have been explained in greater detail to avoid ambiguity.

>In Figure 1B what is the meaning of the parameter “Detection (log)” ? What is the exact signification of the dots and how they are determined ? What is the meaning of the parameter µ  and its value in Figure 1B ? Because there are many definition along the paper (line 15, elevated location; line 98, mean; line 173, location of magnitude). Why only the lowest 14 data points are considered in Figure 1B? What is the exact meaning of σ ? (from line 15 it is a scale). Is it a standard deviation?

I apologise for the confusion caused and thank the reviewer for pointing this out, as it has helped improve the clarity of the paper. Additional explanations have now been included. Specifically, σ is one of the parameters of the normal distribution and is obtained as the standard deviation; in English it is often referred to as the “scale” parameter. µ denotes the location parameter, which in this context corresponds to the mean of the log-transformed values. These clarifications are now provided in the revised text and illustrated in the new Figure S1.  

apologise for the confusion caused and thank the reviewer for pointing this out, as it has helped improve the clarity of the paper. Additional explanations have now been included. Specifically, σ is one of the parameters of the normal distribution and is obtained as the standard deviation; in English it is often referred to as the “scale” parameter. µ denotes the location parameter, which in this context corresponds to the mean of the log-transformed values. These clarifications are now provided in the revised text and illustrated in the new Figure S1.  

Finally, we note that parameters are mathematical concepts used within statistical models, whereas axis labels in figures indicate the data being displayed. We have revised the text to ensure this distinction is clear.

>Line 97, 2023 ?

I have added an explanation.

>In Figure S1 all panels are not labelled. In this Figure what are the meanings of Sorted Detection(log), ideal normal, ideal exponential, Normal Q-Q plot, theoretical quantiles ? Can you explain exactly what is the quantity which is represented in Figure S1B. What are the lines shown in the 4 panels of Figure S1 and how they are calculated ?

I have added an explanation.

>Line 111, S1D is related to Kumamoto not to Noto.

To demonstrate that this phenomenon occurs everywhere, I showed multiple regions.

>Line 114, what is the meaning of P-value ?

I thank the reviewer for this comment. I have added an explanation.

>In Figure 2 what are the meanings of Proportion Nationwide, Sorted magnitude?
> Which areas are considered for Noto EQs and other EQs ?

I thank the reviewer for this comment. I have added an explanation.

>Figure 3 is not correctly described and it is difficult to understand the parameters which are displayed. What is the meaning of λ ?

I thank the reviewer for this comment. I have now added a clear explanation of the parameter λ in the text. Specifically, λ is the sole parameter of the exponential distribution and represents the rate at which events occur. This has been clarified in the revised manuscript and is further illustrated in Figure S1, which provides a textbook-style explanation for readers less familiar with statistical distributions.

>Line 211, not 2000 for S3D

I thank the reviewer for this comment. This was my mistake. Ireplaced it.

>Figures 6C and 6D are related to December 2021 ?

As stated. The earthquake occurred in 2022, so I am investigating the period slightly prior to that.

>In Figure S5 problem of position for ideal normal

I thank the reviewer for this comment. This was my mistake. Ireplaced it.

>Line 273 replace magnitudes by energy

I thank the reviewer for this comment. This was my mistake. Ireplaced it.

>Line 358, what is a z-score ?

I thank the reviewer for this comment. I explained this in the new Figure S1 (becouse this was too long to be included in the main text).

>Finally, it will be interesting to compare your approach with the cumulative number of earthquakes (see for example papers by De Santis et al.)
>De Santis, A., Abbattista, C., Alfonsi, L., Amoruso, L., Campuzano, S. A., Carbone, M., ... & Santoro, F. (2019). Geosystemics view of earthquakes. Entropy, 21(4), 412.

>De Santis, A., Cianchini, G., Favali, P., Beranzoli, L., & Boschi, E. (2011). The Gutenberg–Richter law and entropy of earthquakes: Two case studies in Central Italy. Bulletin of the Seismological Society of America, 101(3), 1386-1395.

I thank the reviewer for the suggestion. The cited works by De Santis et al. adopt approaches based on the Gutenberg–Richter law. As demonstrated in our previous paper [2], I have shown that this law does not hold under rigorous statistical analysis. For this reason, I do not consider revisiting it to be productive in the present study.  

Our approach deliberately avoids unnecessary assumptions, employs models with as few parameters as possible, and ensures falsifiability. The purpose of this paper is not to criticise other methodologies, but rather to introduce a statistically sound method that may contribute to earthquake prediction in Japan, where comprehensive data are uniquely available.

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Please check the attached file.

Comments for author File: Comments.docx

Author Response

>Minor mistakes:
>In Figure 1, there are three panels unlabeled.
Thank you！I had forgotten, so I have corrected it.

>Line 156, it is should be Figures 3A and 3B? 
My apologies. I've corrected it.

>And lines 171-173 the right format?
Thank you. I changed the fiels code.

 

Reviewer 2 Report

Comments and Suggestions for Authors

I thank the authors for carefully responding to my comments. I do not have further questions. I am now glad to accept it.

Author Response

Thank you for your thoughtful review. I cannot thank you enough.

Reviewer 4 Report

Comments and Suggestions for Authors

I thank the author for the corrections of his paper regarding my comments.

In supplementary:

Line 19 EQ intervals ? I think you mean time intervals between EQs ?

Line 60 to be completed

Line 62 lm ?

Line 63 different instead of differing

Figure S7B ideal normal still missing

Main test:

Line 125 2000-2023 also please specify that one point corresponds to one year.

 

Comments on the Quality of English Language

Last point: it will be good to use AI to check the English (and to carefully examine the results).

Author Response

>Comments and Suggestions for Authors
>I thank the author for the corrections of his paper regarding my comments.

I should be the one thanking you; I'm very grateful for your careful review. I believe it has become considerably easier to read.

>In supplementary:

>Line 19 EQ intervals ? I think you mean time intervals between EQs ?
Quite right. I've amended it.

>Line 60 to be completed

Thank you！I had forgotten, so I have corrected it.

Line 62 lm ?

This is the name of an R function. It is thought to be an abbreviation for Fitting Linear Models. Brackets have been added to make this clear.

Line 63 different instead of differing
I've amended it.

Figure S7B ideal normal still missing
My apologies. I've corrected it.

Main test:

>Line 125 2000-2023 also please specify that one point corresponds to one year.
 I've corrected it.

>Comments on the Quality of English Language
>Last point: it will be good to use AI to check the English (and to carefully examine the results).

Thank you. Just to be sure, I asked Copilot, but it seems there weren't many corrections needed. I've amended a few terms.