Estimation of date and magnitude of four major earthquakes using integration of precursors obtained from remote sensing data

Khoshgoftar, Mohammad Mahdi; Saradjian, Mohammad Reza

doi:https://doi.org/10.5194/angeo-2021-41

Preprints

https://doi.org/10.5194/angeo-2021-41

Preprints

01 Sep 2021

| 01 Sep 2021

Status: this preprint was under review for the journal ANGEO but the revision was not accepted.

Estimation of date and magnitude of four major earthquakes using integration of precursors obtained from remote sensing data

Mohammad Mahdi Khoshgoftar and Mohammad Reza Saradjian

Abstract. A single precursor is not usually an accurate, precise and adequate measure to predict earthquake parameters. Therefore, it is more appropriate to exploit parameters extracted from several other single precursors, so that their simultaneous combinations may reduce the uncertainty of the prediction. In this study, remote sensing observations in different modalities acquired from several days before impending earthquakes have been investigated to extract earthquake parameters. They are observations in electron and ion density, electron temperature, Total Electron Content (TEC), Land Surface Temperature (LST), Sea Surface Temperature (SST), Aerosol Optical Depth (AOD), Surface Latent Heat Flux (SLHF), and Outgoing Longwave Radiation (OLR) clear sky. Regarding the ionospheric precursors, the geomagnetic indices Dst, Kp, Ap and F10.7 were used to detect pre-earthquake disturbances from frequent anomalies associated with geomagnetic activity. In this study, three methods of median, support vector regression (SVR) and random forest (RF) have been used to detect anomalies. When anomalies associated with impending earthquakes are detected, the number of prior days associated with the earthquake is estimated based on the type of precursor. Then, by estimation of the amount of anomaly deviation from the normal state, the magnitude of the impending earthquake is estimated. The final earthquake parameters (such as date and magnitude) can be obtained by integrating the earthquake parameters extracted from different earthquake precursors using mean square error (MSE) method.

Received: 27 Jul 2021 – Discussion started: 01 Sep 2021

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Mohammad Mahdi Khoshgoftar and Mohammad Reza Saradjian

Status: closed

RC1:
'Comment on angeo-2021-41', Anonymous Referee #1, 29 Sep 2021

This work is a retrospective analysis of four earthquakes (EQs) searching for possible precursory anomalies in some ionospheric and atmospheric parameters measured by satellites, in particular: electron and ion density, electron temperature, TEC, LST, SST, SLHF and AOD. By the way, TEC is an integrated parameter which is characteristic of both atmosphere and ionosphere, but, of course, it is more indicative of ionospheric state, because it is the ionosphere to contribute more to its value. At first glance, the results seem to confirm both magnitude and time of occurrence of EQs. However, after a more detailed attention, there are some potential flaws in the analyses. I am sorry to say that the paper in its present state cannot be published. I leave the Editor to decide whether to provide some possibility to the Authors to try meet my requests or to reject the paper, so that the Authors can have more longer time to revise it deeply and resubmit it again.

Table of equivalence between Dx and Magnitude ranges (Table 1).

It seems to me a strange coincidence that there is an exact correspondence between integer ranges of Dx and magnitudes. Why? Is there any physical reason for this? Or is it just an empirical decision, rounding the limits of each range to the closest integers.

The space-time domain of analysis.

The period of study is about 50 days before each EQ and 10 days after, looking for anomalies in a circular region, around the EQ epicenter, that is defined by Dobrovolsky formula, where the radius scales with magnitude. Therefore, the analysis considers a region that is related with EQ magnitude, so the authors implicitly impose the knowledge of magnitude. Also the period points to about 50 days where searching the anomalies: if then one defines the possible period of EQ occurrence of 15 (or 16) days after the found anomaly, the probability to guess the true time of occurrence is rather high.

In addition, what about to extend back the analysis? For instance, to about three months before EQ: what happens to the search for anomalies? Are there other new ones? This extension would be useful also because there are some evidences that the precursor time could scale with magnitude of the impending earthquake (Rikitake law; Tectonophysics, 1987), while here the Authors impose the same precursory time for all EQ magnitudes.

Wrong EQ occurrence Predictions.

Some predictions are wrong. In Table 6 and Table 8, some predictions of Samoa EQ occurrence are clearly wrong: the ionospheric anomalies found on 5 September produce a prediction period 6-20 September and NOT 6 Sept.-20 Oct., so missing the true EQ occurrence of Samoa EQ.

MSE of the different precursors.

This work considers 8 types of precursors. The authors use MSE to provide a good estimator for each EQ parameter prediction. However, I suspect that the prediction is highly biased by the fact that some estimators are each other dependent: for instance the ionospheric parameters, or most of the atmospheric parameters (temperature in its different aspects). To be more correct, I would suggest to group all atmospheric or all ionospheric parameters, define V and M for each group and then finally estimate a mean and the absolute deviation between the best result from each group. Of course, this operation would not be important for Gaussian independent variables, but here the analysed precursors are not independent and neither Gaussian.

Minor points

Line 34. Please change “facilitate” with “facilitated”

Lines 36-39. Some recent important references about satellite precursors are missing, as likely the Authors did not read them before writing this paper: e.g. De Santis et al (Scientific Reports, 2019; https://doi.org/10.1038/s41598-019-56599-1) analysed Swarm satellite magnetic field and electron density data. They also found confirmation of the Rikitake law, for which the larger the impending EQ magnitude, the longer the anticipation time of the precursory anomaly; Picozza et al. (Frontiers in Earth Sc., 2021; https://doi.org/10.3389/feart.2021.676775) made a critical review of satellite EQ precursors.

Line 57. The “temperature (K)” is mentioned twice.

Line 84. Is “SLHF” here actually “SST”?

Line 126. Is here “intense” actually “intensity” or (probably better) “strength”?

Eq. (1). Probably for “m” you mean “M” to be in agreement with following equations (2) and (4). You should also change it in the text (line 124)

Line 194. Please correct “December” with “September”.

Citation: https://doi.org/10.5194/angeo-2021-41-RC1
- CC2:
  'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
  Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
  Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
  
  Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
  
  Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
  
  Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
  
  Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-CC2
  - AC1: 'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
    
    Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
    
    Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
    
    Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
    
    Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
    
    Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
    
    Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
    
    Citation: https://doi.org/10.5194/angeo-2021-41-AC1
- AC1:
  'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
  Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
  
  Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
  
  Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
  
  Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
  
  Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
  
  Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC1
CC1:
'Revision', Iman Khosravi, 15 Oct 2021

It seems that one of the appropriate methods for identifying the precursors and examining the anvals and their dependence on the seismic activity of the study on earthquakes has occurred, in which many studies have been done and are being done. . Similarly, regarding the estimation of earthquake parameters, the best way to achieve certainty is to study the earthquakes that have occurred. Due to the fact that in most articles only pre-indicators are studied, in this work, an attempt has been made to estimate the parameters of the earthquake by combining different pre-indicators and reducing uncertainties, which has received less attention from researchers. Is. Due to the totality of the work, it is suggested that this article be accepted with some minor changes.

1. Have the precursors been studied only at the epicenter, or have they been studied in a spatial range?

2. Is this MSE relation correct? And that variance and median are calculated for all predictors. That is, all together and in one model? How the MSE method can be used to estimate earthquake parameters.

3- Minor point: It is suggested that more up-to-date articles be used as research proposals.

Citation: https://doi.org/10.5194/angeo-2021-41-CC1
- AC2:
  'Reply on CC1', Mohammad Mahdi Khoshgoftar, 31 Oct 2021
  Thank you for supporting the paper and the positive points you made in this regard. I will state the following in response to your points:
  None of the precursors are focused solely on the epicenter, and the time series consisting of all of them is collected in a spatial interval relative to the epicenter.
  
  Yes, this equation is correct for MSE. This equation is applied when MSE is used as an estimator. The variance and median for the predicted upper and lower limits for the date and magnitude of the earthquake are calculated separately for all precursors. Finally, any parameter that has a minimum value of MSE is considered as the final parameter of the earthquake.
  
  Regarding minor points, all were corrected.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC2
RC2:
'Comment on angeo-2021-41', Anonymous Referee #2, 23 Oct 2021

The manuscript titled 'Estimation of date and magnitude of four major earthquakes using integration of precursors obtained from remote sensing data' by Khoshgoftar et al., is an interesting study dealing with the detection of earthquake precursors. This is a well-written study with the use of a large number of various data sets. However, I have serious reservations on the results and I list my objections below in the form of major comments:

1. As this study deals with anomalies and the authors have made major use of TEC anomalies in their study, I do not find mention of TEC uncertainties anywhere in this manuscript. The authors mention that they have used TEC GIMs for earthquake precursor detection but the uncertainty levels have not been mentioned. As the TEC uncertainties in GIMs except over North America and Europe generally exceed 3 TECu and the authors are dealing with this level of DTEC, it is difficult to ascertain whether the TEC anomalies are genuine or just random noise.

2. The same goes for other datasets and their uncertainty levels. Unless this is clarified, it is difficult to make valid judgement about the results. I'd request the authors to make appropriate changes in the manuscript based on these comments.

I recommend a major revision for this manuscript purely because it is difficult to take a correct call on the shown results and the efficacy of the methods that are mentioned.

Citation: https://doi.org/10.5194/angeo-2021-41-RC2
- AC3:
  'Reply on RC2', Mohammad Mahdi Khoshgoftar, 02 Nov 2021
  We would like to thank the reviewer’s fruitful comments on our manuscript. Each point has been explained as follows:
  In response to the reviewer’s comment, it should be noted that regarding the case of ionospheric data, factors such as geomagnetic and solar activities that can cause anomalies in the data have been already removed from the time series in this study, and the remaining anomalies are due to seismic action. Therefore, regarding the time series, the anomalies detected near the time of earthquakes occurrence are due to the seismic activity solely.
  
  By using the combination of precursors in this study, the uncertainties in estimating earthquake parameters have been removed implicitly. However, one method to explicitly and directly deal with the uncertainty of precursors is to study the precursors in different earthquakes that have occurred in the past.Another method is to study similar and comparable years’ data in which no earthquake occurred.The idea suggested by the referee is interesting but regarding the aim of the study, it falls outside of the scope of the study.However, we will work on assessing uncertainties directly and provide additional experimental information on our next manuscript.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC3

Status: closed

RC1:
'Comment on angeo-2021-41', Anonymous Referee #1, 29 Sep 2021

This work is a retrospective analysis of four earthquakes (EQs) searching for possible precursory anomalies in some ionospheric and atmospheric parameters measured by satellites, in particular: electron and ion density, electron temperature, TEC, LST, SST, SLHF and AOD. By the way, TEC is an integrated parameter which is characteristic of both atmosphere and ionosphere, but, of course, it is more indicative of ionospheric state, because it is the ionosphere to contribute more to its value. At first glance, the results seem to confirm both magnitude and time of occurrence of EQs. However, after a more detailed attention, there are some potential flaws in the analyses. I am sorry to say that the paper in its present state cannot be published. I leave the Editor to decide whether to provide some possibility to the Authors to try meet my requests or to reject the paper, so that the Authors can have more longer time to revise it deeply and resubmit it again.

Table of equivalence between Dx and Magnitude ranges (Table 1).

It seems to me a strange coincidence that there is an exact correspondence between integer ranges of Dx and magnitudes. Why? Is there any physical reason for this? Or is it just an empirical decision, rounding the limits of each range to the closest integers.

The space-time domain of analysis.

The period of study is about 50 days before each EQ and 10 days after, looking for anomalies in a circular region, around the EQ epicenter, that is defined by Dobrovolsky formula, where the radius scales with magnitude. Therefore, the analysis considers a region that is related with EQ magnitude, so the authors implicitly impose the knowledge of magnitude. Also the period points to about 50 days where searching the anomalies: if then one defines the possible period of EQ occurrence of 15 (or 16) days after the found anomaly, the probability to guess the true time of occurrence is rather high.

In addition, what about to extend back the analysis? For instance, to about three months before EQ: what happens to the search for anomalies? Are there other new ones? This extension would be useful also because there are some evidences that the precursor time could scale with magnitude of the impending earthquake (Rikitake law; Tectonophysics, 1987), while here the Authors impose the same precursory time for all EQ magnitudes.

Wrong EQ occurrence Predictions.

Some predictions are wrong. In Table 6 and Table 8, some predictions of Samoa EQ occurrence are clearly wrong: the ionospheric anomalies found on 5 September produce a prediction period 6-20 September and NOT 6 Sept.-20 Oct., so missing the true EQ occurrence of Samoa EQ.

MSE of the different precursors.

This work considers 8 types of precursors. The authors use MSE to provide a good estimator for each EQ parameter prediction. However, I suspect that the prediction is highly biased by the fact that some estimators are each other dependent: for instance the ionospheric parameters, or most of the atmospheric parameters (temperature in its different aspects). To be more correct, I would suggest to group all atmospheric or all ionospheric parameters, define V and M for each group and then finally estimate a mean and the absolute deviation between the best result from each group. Of course, this operation would not be important for Gaussian independent variables, but here the analysed precursors are not independent and neither Gaussian.

Minor points

Line 34. Please change “facilitate” with “facilitated”

Lines 36-39. Some recent important references about satellite precursors are missing, as likely the Authors did not read them before writing this paper: e.g. De Santis et al (Scientific Reports, 2019; https://doi.org/10.1038/s41598-019-56599-1) analysed Swarm satellite magnetic field and electron density data. They also found confirmation of the Rikitake law, for which the larger the impending EQ magnitude, the longer the anticipation time of the precursory anomaly; Picozza et al. (Frontiers in Earth Sc., 2021; https://doi.org/10.3389/feart.2021.676775) made a critical review of satellite EQ precursors.

Line 57. The “temperature (K)” is mentioned twice.

Line 84. Is “SLHF” here actually “SST”?

Line 126. Is here “intense” actually “intensity” or (probably better) “strength”?

Eq. (1). Probably for “m” you mean “M” to be in agreement with following equations (2) and (4). You should also change it in the text (line 124)

Line 194. Please correct “December” with “September”.

Citation: https://doi.org/10.5194/angeo-2021-41-RC1
- CC2:
  'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
  Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
  Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
  
  Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
  
  Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
  
  Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
  
  Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-CC2
  - AC1: 'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
    
    Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
    
    Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
    
    Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
    
    Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
    
    Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
    
    Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
    
    Citation: https://doi.org/10.5194/angeo-2021-41-AC1
- AC1:
  'Reply on RC1', Mohammad Mahdi Khoshgoftar, 23 Oct 2021
  Thanks to the reviewer’s fruitful comments. Due to the rejection, however, I would like to describe some points on the objections that may help change the decision.
  
  Regarding the Table 1 on correspondence between Dx and Magnitude ranges, I would like to point out that they are from the studies in the literature on earthquakes and the severity of anomalies.
  
  Regarding the space-time domain of analysis, I would like to point out that an extension to the study is underway for a period of 3 months and more, and more promising result have obtained. Studies on the radius of the areas affected by earthquake were carried out according to the Dobrovolsky formula. Periods of 15 or 16 days have also been selected based on other researchers' studies on each precursor.
  
  Regarding “Wrong EQ occurrence Predictions”, I would like to point out that any single precursor may give wrong prediction therefore multiple precursors with multiple predictions have been exploited and combined so that the uncertainties and false predictions are decreased and even avoided.
  
  Regarding MSE of the different precursors, it should be noted that V and M‌ are obtained from the dates and intensities obtained from the precursors. Therefore, the difference in the type of precursors has no effect on this issue, because its input is not a physical quantity.
  
  Regarding minor points, all cases except for the one mentioned for line 57 were corrected. In the case of line 57, it should be noted that the first temperature is related to the ion and the second temperature is related to the electron.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC1
CC1:
'Revision', Iman Khosravi, 15 Oct 2021

It seems that one of the appropriate methods for identifying the precursors and examining the anvals and their dependence on the seismic activity of the study on earthquakes has occurred, in which many studies have been done and are being done. . Similarly, regarding the estimation of earthquake parameters, the best way to achieve certainty is to study the earthquakes that have occurred. Due to the fact that in most articles only pre-indicators are studied, in this work, an attempt has been made to estimate the parameters of the earthquake by combining different pre-indicators and reducing uncertainties, which has received less attention from researchers. Is. Due to the totality of the work, it is suggested that this article be accepted with some minor changes.

1. Have the precursors been studied only at the epicenter, or have they been studied in a spatial range?

2. Is this MSE relation correct? And that variance and median are calculated for all predictors. That is, all together and in one model? How the MSE method can be used to estimate earthquake parameters.

3- Minor point: It is suggested that more up-to-date articles be used as research proposals.

Citation: https://doi.org/10.5194/angeo-2021-41-CC1
- AC2:
  'Reply on CC1', Mohammad Mahdi Khoshgoftar, 31 Oct 2021
  Thank you for supporting the paper and the positive points you made in this regard. I will state the following in response to your points:
  None of the precursors are focused solely on the epicenter, and the time series consisting of all of them is collected in a spatial interval relative to the epicenter.
  
  Yes, this equation is correct for MSE. This equation is applied when MSE is used as an estimator. The variance and median for the predicted upper and lower limits for the date and magnitude of the earthquake are calculated separately for all precursors. Finally, any parameter that has a minimum value of MSE is considered as the final parameter of the earthquake.
  
  Regarding minor points, all were corrected.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC2
RC2:
'Comment on angeo-2021-41', Anonymous Referee #2, 23 Oct 2021

The manuscript titled 'Estimation of date and magnitude of four major earthquakes using integration of precursors obtained from remote sensing data' by Khoshgoftar et al., is an interesting study dealing with the detection of earthquake precursors. This is a well-written study with the use of a large number of various data sets. However, I have serious reservations on the results and I list my objections below in the form of major comments:

1. As this study deals with anomalies and the authors have made major use of TEC anomalies in their study, I do not find mention of TEC uncertainties anywhere in this manuscript. The authors mention that they have used TEC GIMs for earthquake precursor detection but the uncertainty levels have not been mentioned. As the TEC uncertainties in GIMs except over North America and Europe generally exceed 3 TECu and the authors are dealing with this level of DTEC, it is difficult to ascertain whether the TEC anomalies are genuine or just random noise.

2. The same goes for other datasets and their uncertainty levels. Unless this is clarified, it is difficult to make valid judgement about the results. I'd request the authors to make appropriate changes in the manuscript based on these comments.

I recommend a major revision for this manuscript purely because it is difficult to take a correct call on the shown results and the efficacy of the methods that are mentioned.

Citation: https://doi.org/10.5194/angeo-2021-41-RC2
- AC3:
  'Reply on RC2', Mohammad Mahdi Khoshgoftar, 02 Nov 2021
  We would like to thank the reviewer’s fruitful comments on our manuscript. Each point has been explained as follows:
  In response to the reviewer’s comment, it should be noted that regarding the case of ionospheric data, factors such as geomagnetic and solar activities that can cause anomalies in the data have been already removed from the time series in this study, and the remaining anomalies are due to seismic action. Therefore, regarding the time series, the anomalies detected near the time of earthquakes occurrence are due to the seismic activity solely.
  
  By using the combination of precursors in this study, the uncertainties in estimating earthquake parameters have been removed implicitly. However, one method to explicitly and directly deal with the uncertainty of precursors is to study the precursors in different earthquakes that have occurred in the past.Another method is to study similar and comparable years’ data in which no earthquake occurred.The idea suggested by the referee is interesting but regarding the aim of the study, it falls outside of the scope of the study.However, we will work on assessing uncertainties directly and provide additional experimental information on our next manuscript.
  
  Citation: https://doi.org/10.5194/angeo-2021-41-AC3

Mohammad Mahdi Khoshgoftar and Mohammad Reza Saradjian

Viewed

Total article views: 1,298 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
997	247	54	1,298	44	48

HTML: 997
PDF: 247
XML: 54
Total: 1,298
BibTeX: 44
EndNote: 48

Views and downloads (calculated since 01 Sep 2021)

Month	HTML	PDF	XML	Total
Sep 2021	245	18	4	267
Oct 2021	232	21	10	263
Nov 2021	89	23	4	116
Dec 2021	15	6	1	22
Jan 2022	32	8	0	40
Feb 2022	27	3	0	30
Mar 2022	16	5	1	22
Apr 2022	12	4	0	16
May 2022	9	5	1	15
Jun 2022	4	3	1	8
Jul 2022	8	4	0	12
Aug 2022	8	6	0	14
Sep 2022	4	3	0	7
Oct 2022	5	3	1	9
Nov 2022	12	1	0	13
Dec 2022	11	3	0	14
Jan 2023	6	18	0	24
Feb 2023	10	3	1	14
Mar 2023	14	1	0	15
Apr 2023	5	4	1	10
May 2023	11	2	1	14
Jun 2023	10	8	1	19
Jul 2023	15	12	1	28
Aug 2023	11	7	2	20
Sep 2023	14	2	7	23
Oct 2023	31	2	1	34
Nov 2023	14	4	0	18
Dec 2023	18	6	3	27
Jan 2024	14	10	0	24
Feb 2024	19	12	0	31
Mar 2024	15	14	0	29
Apr 2024	13	1	4	18
May 2024	10	3	3	16
Jun 2024	6	6	1	13
Jul 2024	8	2	4	14
Aug 2024	4	5	0	9
Sep 2024	3	3	0	6
Oct 2024	6	4	0	10
Nov 2024	8	2	1	11
Dec 2024	3	0	3

Cumulative views and downloads (calculated since 01 Sep 2021)

Month	HTML	PDF	XML	Total
Sep 2021	245	18	4	267
Oct 2021	232	21	10	263
Nov 2021	89	23	4	116
Dec 2021	15	6	1	22
Jan 2022	32	8	0	40
Feb 2022	27	3	0	30
Mar 2022	16	5	1	22
Apr 2022	12	4	0	16
May 2022	9	5	1	15
Jun 2022	4	3	1	8
Jul 2022	8	4	0	12
Aug 2022	8	6	0	14
Sep 2022	4	3	0	7
Oct 2022	5	3	1	9
Nov 2022	12	1	0	13
Dec 2022	11	3	0	14
Jan 2023	6	18	0	24
Feb 2023	10	3	1	14
Mar 2023	14	1	0	15
Apr 2023	5	4	1	10
May 2023	11	2	1	14
Jun 2023	10	8	1	19
Jul 2023	15	12	1	28
Aug 2023	11	7	2	20
Sep 2023	14	2	7	23
Oct 2023	31	2	1	34
Nov 2023	14	4	0	18
Dec 2023	18	6	3	27
Jan 2024	14	10	0	24
Feb 2024	19	12	0	31
Mar 2024	15	14	0	29
Apr 2024	13	1	4	18
May 2024	10	3	3	16
Jun 2024	6	6	1	13
Jul 2024	8	2	4	14
Aug 2024	4	5	0	9
Sep 2024	3	3	0	6
Oct 2024	6	4	0	10
Nov 2024	8	2	1	11
Dec 2024	3	0	3

Viewed (geographical distribution)

Total article views: 1,266 (including HTML, PDF, and XML) Thereof 1,266 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Dec 2024

Short summary

This article is about using remote sensing data to estimate the date and magnitude of the impending earthquakes. An earthquake can also affect the environment. Damage to the environment can destroy and lose lives and be more dangerous than earthquakes. The damage sustained by earthquakes is due to the secondary effects of man-made materials on the environment. Water pipes are broken and the ground is flooded, and gas, electricity and scattered fuel lines cause fires.


Total:	0
HTML:	0
PDF:	0
XML:	0