The Relationship Between the Seismic Regime and Low-Frequency Variations in Meteorological Parameters Measured at a Network of Stations in Japan

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper investigates how variations in meteorological parameters — humidity, temperature, atmospheric pressure, and wind speed — relate to seismic activity in Japan over the period 1973 to mid-2025.

It uses meteorological data from NOAA’s global hourly dataset, while the seismic data come from the USGS earthquake database.

The authors apply the PCA to compute weighted averages of meteorological data. Wavelet decomposition allows them to extract frequency components (focus on 7th level, representing 16–32 day periods). Hilbert transform generates envelopes of wavelet components. Coupled point process model assesses to find the relationships between local extrema of envelopes and earthquake occurrence.

The authors find two distinct mechanisms for the relationship between meteorological changes and earthquakes: i) Trigger Mechanism (short relaxation time): Associated with cyclones, showing: Minima in humidity, pressure, and temperature and Maxima in wind speed. The authors suggest that cyclones may trigger seismic events. ii) Precursor Mechanism (long relaxation time): they are likely linked to atmospheric precursors caused by radon release and air ionization. They show an opposite pattern of (i) : maxima in humidity, pressure, temperature, and minima in wind speed. The trigger effect is generally stronger than the precursor effect, making precursors harder to detect. A spectral analysis reveals a 12.8-year periodicity, close to the solar activity cycle.

The study contributes to earthquake precursor research by distinguishing between meteorological triggers and atmospheric precursors, offering a potential path toward improved short-term earthquake prediction, however some critical points must be solved before publication. At this stage a major revision is required.

Major points

• The analysis is made over the all dataset. A more convincing method would be to apply it to part of the dataset (e.g. half) and then compare with the remaining dataset (which is independent from that where the method was initially applied). This is a crucial point. Of course the cost would be to explore time scales till to half of the whole available interval, but any result would be more robust and convincing.
• Some of the methodology is complex (e.g. parameter optimization) and would benefit from illustrative diagrams or a conceptual flowchart.
• Discussion is too short and mostly descriptive. It would benefit from clearer statements and practical implications of the research, such as applications in earthquake prediction systems (e.g. De Santis et al. 2025) or early warning.
• English language must be improved.
• Figures lack detailed captions and are not fully auto-explanatory. The color coding should be explained in each figure.

Minor points

• Table 1, which is central to the article, misses a caption.
• The first sentence of the section “7. Discussion” (“This section is not mandatory …”) seems something coming from the journal template, and should be removed.
• A section “Conclusion” is missing.

Reference

De Santis et al. Foundations for an Operational Earthquake Prediction System, Geosciences, 15 (2), 69, 2025. https://doi.org/10.3390/geosciences15020069

 

Author Response

First reviewer

The analysis is made over the all dataset. A more convincing method would be to apply it to part of the dataset (e.g. half) and then compare with the remaining dataset (which is independent from that where the method was initially applied). This is a crucial point. Of course the cost would be to explore time scales till to half of the whole available interval, but any result would be more robust and convincing.

 

Response

Please look point  7, Tables 2 and 3 and Figure 8 in the new version of the paper

 

First reviewer

Some of the methodology is complex (e.g. parameter optimization) and would benefit from illustrative diagrams or a conceptual flowchart.

 

Response

Please look lines 280-290 in new version of the paper.

 

First reviewer

Discussion is too short and mostly descriptive. It would benefit from clearer statements and practical implications of the research, such as applications in earthquake prediction systems (e.g. De Santis et al. 2025) or early warning.

 

Response

Please look lines 441-462 and reference point [44] in new version of the paper

 

First reviewer

English language must be improved.

Response

I suppose that MDPI English language service will improve.

 

First reviewer

Figures lack detailed captions and are not fully auto-explanatory. The color coding should be explained in each figure.

Response

Please look figures captions within new version of the paper.

 

First reviewer

Table 1, which is central to the article, misses a caption

Response

Please look new version of the paper.

 

First reviewer

The first sentence of the section “7. Discussion” (“This section is not mandatory …”) seems something coming from the journal template, and should be removed.

Response

This is occasional mistake, improved.

 

First reviewer

A section “Conclusion” is missing.

Response

Please look new version of the paper.

Reviewer 2 Report

Comments and Suggestions for Authors

A large database for atmospheric parameters and high intensity earthquakes in Japan is analyzed, both statistically and phenomenologically and the results are similar to those from other authors, extensively presented in the introductory part.

Comments on the Quality of English Language

The use of English is satisfactory along the manuscript.

Author Response

Second reviewer

 

A large database for atmospheric parameters and high intensity earthquakes in Japan is analyzed, both statistically and phenomenologically and the results are similar to those from other authors, extensively presented in the introductory part.

 

Response

Thank you. I suppose that the new version of the paper is more adequate.

 

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript investigates the relationship between low-frequency variations in meteorological parameters (such as humidity, pressure, temperature, and wind speed) and seismic activity in the Japanese archipelago. Based on multi station observation data from 1973 to 2025, a method is proposed to analyze a large number of long-term meteorological time series measured on the station network. Principal component analysis, wavelet decomposition, envelope extraction, and point process coupling models are used to compare their anomalies with earthquake conditions; Propose two identification mechanisms: one related to earthquakes triggered by cyclones, and the other related to atmospheric precursor phenomena. The research methods and results have certain scientific value. But there are still some issues that need further clarification and improvement:

1. Although the manuscript mentions two leading mechanisms (trigger and precursor), it does not specifically discuss their application prospects, feasibility, or limitations in actual earthquake prediction. The author needs to supplement;
2. The introduction section of the manuscript is a bit lengthy, and citing references when reviewing the research background appears somewhat rigid. It is recommended to polish the language expression, sentence statement, and other aspects;
3. In the manuscript, the explanation of the figures appears messy, as shown in Figure 4, which reads "The graphs are presented: (a) - moments of time of 213 earthquakes with a magnitude of at least 6.5; (b) - (e) - amplitudes of the envelopes of the...". It is necessary to first summarize and then describe them according to the subgraphs. In addition, the icon description overlaps too much with the following description in Figure 4, and this part needs to be adjusted;
4. It is suggested to delete the phrase "This section is not mandatory but may be added if there are patents resulting from the work reported in this manual;
5. The title of the last section is suggested to be changed to “Conclusion and Discussion” for better structure.
6. There are too many keywords.

Author Response

Although the manuscript mentions two leading mechanisms (trigger and precursor), it does not specifically discuss their application prospects, feasibility, or limitations in actual earthquake prediction. The author needs to supplement;

Response

Please look sections Discussion and Conclusion in the new version of the paper

 

Third reviewer

The introduction section of the manuscript is a bit lengthy, and citing references when reviewing the research background appears somewhat rigid. It is recommended to polish the language expression, sentence statement, and other aspects;

Response

I suppose that MDPI English language service will improve.

 

Third reviewer

In the manuscript, the explanation of the figures appears messy, as shown in Figure 4, which reads "The graphs are presented: (a) - moments of time of 213 earthquakes with a magnitude of at least 6.5; (b) - (e) - amplitudes of the envelopes of the...". It is necessary to first summarize and then describe them according to the subgraphs. In addition, the icon description overlaps too much with the following description in Figure 4, and this part needs to be adjusted;

Response

I tried to make the captions to the figures as detailed as possible so that each figure could be understood separately from the main text.

 

Third reviewer

It is suggested to delete the phrase "This section is not mandatory but may be added if there are patents resulting from the work reported in this manual;

Response

This is occasional mistake, improved.

 

Third reviewer

The title of the last section is suggested to be changed to “Conclusion and Discussion” for better structure.

Response

Please look new version of the paper. I introduce section Conclusion and section Discussion separately.

 

Third reviewer

There are too many keywords.

Please look new version of the paper. I decreased the number of keywords

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, thanks for your revision. I am almost satisfied. However, Discussion and Conclusions do not take completely into account the results from the new section 7: the great differences in some parameters must give more caution in the interpretation of the results over the whole dataset. Some more discussion must be included. Some mentions should be also given in the Conclusions.

Author Response

Please look lines 447-458 within section Discussion in the new version of the paper