The record of the magnetic storm on 15 May 1921 in Stará Ďala (present-day Hurbanovo) and its compliance with the global picture of this extreme event

Koči, Eduard; Valach, Fridrich

doi:https://doi.org/10.5194/angeo-41-355-2023

Articles | Volume 41, issue 2

https://doi.org/10.5194/angeo-41-355-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/angeo-41-355-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 41, issue 2

Regular paper

|

12 Sep 2023

Regular paper |

| 12 Sep 2023

The record of the magnetic storm on 15 May 1921 in Stará Ďala (present-day Hurbanovo) and its compliance with the global picture of this extreme event

Eduard Koči and Fridrich Valach

Download

Final revised paper (published on 12 Sep 2023)
Supplement to the final revised paper
Preprint (discussion started on 12 Apr 2023)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on angeo-2023-12', Anonymous Referee #1, 15 May 2023

The paper introduces newly recovered historical data contributing to the investigation of the 13-15 May 1921 extreme geomagnetic storm. Since both extreme geomagnetic events and their surviving recordings before the 1930s are rare, any new data are important. The unscaled recordings were carefully calibrated. The authors presented the Stara Dala data in the context of other observations of the same storms, as well as closely located aurora observations. The section detailing data processing is too lengthy. Global magnetic data used for this study re taken from WDC. However, it is not clear if this is a complete set of the available data. Figures 4 and 5 includes 10-12 data points, while Hapgood listed 21 observotories (excluding Stara Dala) with available data. The authors seem to consider only Czech aurora observations. Hapgood mentions other sources (see also their references). Relatively little information is given on how other researchers interpreted the available data related to this event. Missing the summary of others findings it is difficult to judge what are the new findings of this paper.

Minor comments:
18: curtailed > limited
29: size > magnitude
38: extent > the extreme equatorward extension
41: an extreme solar storm > the causative solar storm
42: Mountwilson > Mount Wilson
47: will be > could be
52: in low magnetic latitudes > at low magnetic latitudes
63: the descending phases > a declining phase
Table 1: Throughout the paper CGM latitudes are used, geomagnetic coordinates could be removed from the table. They just confuse the reader. See also comment to line 277.
Table 2 and throughout the text: The unit in this form is not correct as it includes a decimal dot. Use „arc minute/mm” or „’/mm” instead.
162: centre > median (?) Please specify clearly!
175: Contemplating > Considering
182: thoroughly selected: based on what selection criteria?
Table 5 and text: Use nT/mm as scale unit!
201: we read the five-minute data (i.e., five-minute means): this needs to be clarified. Means are typically calculated and not just read.
202: not at our disposal > missing OR Where the magnetograms were incomplete
Table 6 caption in … days > on … days
Figure 1: Use wider line, connect sporadic red points with a thin or dotted line to guide the reader eye (line in Fig 2). Now it is difficult the follow the development of the storm in the presented plots. Consider if you could make this figure smaller.
Figure 2: This figure could be made smaller. Note in the caption that the lines connecting the observations is added only to guide the reader’s eyes.
226: limpid > clear
226: sank > set
227: the light was whitish even white: This is not very clear. maybe ’whitish or even white’
243: positive: does it have any relevance? The baseline is instrumental and does not have any physical meaning.
Fig 3: Thicker lines, smaller figure with properly adjusted caption size.
273: reference point > origin or pole
277: Use QDP coordinates everywhere rather than a mixture of coordnates (it is unlikely that you need to redraw any plots because of this technical correction, since as you mentioned QDP and CGM are indistinguishable at high latitudes. Modify Table 1 accordingly.
283: the orientation of the arrows is adjusted as the respective variations would appear in the northern hemisphere if they were the result of currents in auroral ovals or FACs belonging to the northern hemisphere: This is not very clear to me.
288: parallels > circles of latitude
290: increase: of what? Help the reader!
294: outer: ok, but equatorward is more specific.
318: locality > region
320: close to the equatorward boundary of the oval, in this part of which the westward electrojet flowed > close to the section of the equatorward boundary of the oval associated with the westward electrojet [or sg similar]
329: A possibly interesting … : Rephrase this sentence, make it shorter and simpler.
332: each to other > in relation to each other
333 With a great deal of… : Rephrase this sentence [themselves, northernmost]. This is a statement, the reasoning is missing.
Figure 5. There is no need to repeat almost the whole caption of Fig 4. You could say simply: Same as Fig. 4 but for the vertical component.
340: good to mention > worth mentioning
346: To achieve this goal, we dealt with several partial tasks > We achieved this goal through several steps.
355: can be compared > is comparable
Units!
365 and 377: You already mentioned in line 15 the changes of the name of the location throughout its history. It is ok to remind the reader once but it is absolutely unnecessary to do it twice
378: 56.5 nT: Not clear why this arbitrarily chosen temporal variation is relevant for the storm studied.

Citation: https://doi.org/10.5194/angeo-2023-12-RC1
- AC1: 'Reply on RC1', Fridrich Valach, 30 May 2023
  
  We would like to thank Referee #1 very much for his/her time and work devoted to carefully reviewing our manuscript and for the helpful comments.
  Our answers to the referee’s general comments are as follows:
  The data calibration for the preserved magnetogram forms an important part of our manuscript because in this part of our study we tried to piece together the sketchy information in order to obtain the most reliable resulting data available. We will try to make the manuscript more attractive by adding a piece of information about the content of the relevant parts of the text. Namely, in Lines 85ff we will clarify that the following sections describe in detail the procedure of the calibration, and the resulting calibrated data are presented later, in Section 2.3. By such an improved navigation in the text, we will allow the reader who is less interested in the calibration procedure to avoid the uninteresting parts.
  In our study, all the one-hour data that are available on the WDC (https://wdc.bgs.ac.uk/catalog/master.html) were used. We will provide information of this data source in the Data availability section in the new manuscript.
  In our work, we wanted to point out to the international scientific community little-known (or nearly unknown) information about the aurora borealis observed during the storm of May 1921 from the territory of the former Czechoslovakia. In the new version of the manuscript, we will also add brief information about the auroras presented in Hapgood's review article, and in the discussion we will indicate how the observations from Czechoslovakia fit into the information published so far.
  Our answers to the referee’s minor comments are as follows:
  Line 162: Yes, we used medians as the centre values. We will add this piece of information to the manuscript.
  Line 182: We selected records in which the light trace on the magnetogram (photo paper) was clear and thin, to allow obtaining values with as little uncertainty as possible.
  Line 201: In determining the 5-minute averages, we imitated the standard method used in determination the hourly averages using a glass scale. The average value was determined using an imaginary horizontal straight line, which was put on the magnetogram so that we were making equal the areas between the trace being scaled and the horizontal line (H. E. McComb, Magnetic Observatory Manual, US Government Printing Office, Washington 1952, pp. 177-178).
  Line 243: We will erase "positive" and only write "eastern".
  Line 283: For data in the Southern Hemisphere, we displayed a positive variation in declination with an arrow pointing west (i.e. in the opposite direction than in the Northern Hemisphere), and an increase in horizontal intensity is shown with an arrow pointing toward the pole (i.e. as in the Northern Hemisphere).
  Line 290: Increase in the vertical intensity.
  Line 333 "With a great deal of caution…" In the new version of the manuscript, we decided to erase the speculative interpretation based on a mere pair of data. It is actually a local variation that we can only explain as a consequence of some local electric current. At De Bilt, after a two-hour data gap (probably caused by an extreme value outside the measurement range of the instrument), the hourly average of declination between 05:00 UT and 06:00 UT was -12.31°, and the next hourly average was -10.61°, the difference in values being up to 1.7°. Based on the studied data, we cannot specify the characteristics of this current in more detail. Nevertheless, it was such a sharp and spatially limited variation, which typically does not occur in mid-latitudes, and resembles variations typically occuring in the vicinity of the auroral oval.
  Line 378: We discard the sentences on line 375 ("For example, ...") and on line 379 ("However, it should also be noted ...") as unnecessary.
  We accept all the comments of the Referee #1, including those not mentioned in our responses above, and we will incorporate them into the manuscript.
  We would like to thank Referee #1 also for his/her comments aimed at improving the language of the text.
  
  Citation: https://doi.org/10.5194/angeo-2023-12-AC1
RC2:
'Comment on angeo-2023-12', Anonymous Referee #2, 22 May 2023

Review comments on the manuscript 10.5194/angeo-2023-12
" The record of the magnetic storm on 15 May 1921 in Stará Dala (present day Hurbanovo) and its compliance with the global picture of this extreme event”
E. Koci and F. Valach

The authors have tried to reconstruct the historical magnetic field recordings of the former Stará Dala observatory for the period of the major magnetic storm in the middle of May 1921. This can be considered a valuable contribution to better characterize space weather conditions during extreme magnetic storms. By applying different approaches for recalibrating the recordings and by comparison with neighboring observatories they obtain reasonably convincing data series for the time period of interest.
Besides these generally positive results the study contains also weaknesses, in particular when it comes to the storm-related magnetic variations. More details of the expected improvements are listed below. Overall, the work is regarded worth being made public, but substantial revisions are expected before it should appear in Annales Geophysicae.

General comments
1) The interpretation of the storm features based on the observed magnetic variations is not convincing. Generally, it is a pity that the northward, H component is missing at Stará Dala. This is most important for the characterization of magnetic activity. Deflections of the vertical, Z component are strongly influenced by the subsurface conductivity distribution. This fact should be clearly stated in the paper. For example, the observatories Wingst and Niemegk exhibit commonly opposite deflections during times of magnetic activity. This is caused by the effect of the so-called northern German anomaly. After having said that, it is worth to continue with the available data from Stará Dala.
In the paper the shown variations recorded at Niemegk give the most complete picture of the storm evolution. However, they are taken about 5° north of Stará Dala. This can make significant differences during a magnetic storm. It would have been very instructive to add complete field recordings from similar latitudes. In my view Munich-Bogenhausen could provide valuable recordings for comparison, complementing very well the Niemegk data. The H variations from Munich could help to better quantify the southward extend of the electrojet.

2) Another topic of concern is the frequent quoting of field-aligned currents (FAC) in connection with observed eastward, D component variations. It is known since Fukushima’s famous publication that field-aligned currents cause virtually not magnetic signature on ground. For that reason, all the parts where FACs are mentions should be revised in this respect. If FACs are quoted, the related Hall currents, that give rise to ground deflections, have to be introduced and made consistent with the observations. In some parts the major storm of October 2003 is taken as reference. For that, detailed observations from ground and satellite are available. For example, Wang et al. (Annales Geophysicae, 24, 311–324, 2006) describes well the relation of FAC to the intensity of solar wind input and ring current activity. This is different for day and nightside. Considering their results may help to support the offered interpretation.

Detailed comments
Line 117: The sentence “We also know that the then device worked in …” is not clear.

Sect. 2.2.3: The variations of the vertical component depend strongly on the subsurface conductivity; opposite deflections are observed between Wingst and Niemegk. The dependence on conductivity has to be mentioned.

Fig. 1 It would have been instructive to mark the times of SCs here, possibly also in Fig. 3.

Fig. 2 The second peak in declination from Clementinum is rather questionable, while the first corresponds reasonably well with the related variations in Niemegk. However, around 18 UT the storm activity has died out. It is thus practically impossible that such a large D deflection could have happened. Here again the Munich data could be decisive. When presenting historical data, it is important to check critically their reliability. This critical assessment of the Clementinum data has to be spelled out clearer, e.g. in the paragraph following line 240.

Line 245ff: “Possibly it might be a manifestation of a field aligned current”, All these statements concerning FACs in relation to declination variations are misleading. They should be removed here and elsewhere.

Lines 320ff: The relations between H und Z variation at Niemegk and possibly Munich, could be used to estimate the latitudinal position of the westward electrojet. Actually, Z goes through zero under the electrojet. A simple model for quantifying the H to Z relation is to assume a line current in the ionosphere.

Fig. 5: I am not sure what to learn from this figure. The sign of Z deflection varies from place to place, and not conclusions are drawn in this work from it. It well could be dropped or put into the Supplements.

Lines 341ff: In comparison with the 29 Oct. 2003, storm the Wang et al. paper should be used to help quantifying the expansion of the auroral oval. Useful information can be obtained in this regard from a reconstruction of the ring current intensity from low-latitude stations.

Lines 388ff: The conclusions listed here are presently pure speculation. After revision of the manuscript, they have to be improved.

Citation: https://doi.org/10.5194/angeo-2023-12-RC2
- AC2: 'Reply on RC2', Fridrich Valach, 30 May 2023
  
  We would like to thank Referee #2 very much for his/her time and work devoted to carefully reviewing our manuscript and for the helpful comments.
  
  Our answers to the referee’s general comments are as follows:
  
  We will remove disputed and incorrect interpretations from Section 3.2.2 and from Discussion. Instead, we focus on the declination variations that were observed in the mid-latitudes in the morning sector between 03:00 UT and 07:00 UT.
  Because the geometry of the geomagnetic field in mid-latitudes does not strictly meet the assumptions of Fukushima's Theorem (field lines in mid-latitudes cannot be considered perpendicular to the earth's surface; moreover, in disturbed geomagnetic conditions, the assumption of uniform conductivity might not be well fulfilled), even ground-based geomagnetic observations can capture at least some manifestations of FACs. In our opinion, this may also be the case for the mentioned mid-latitude declination variations in the morning sector between 03:00 UT and 07:00 UT. This is why we might consider the pronounced easterly variation to be a likely consequence of an upward-moving electric current; in this time sector, the R2 FACs are possible candidates for such a current.
  
  Thank you for bringing Wang's work to our attention. We will use his determination of the equatorward extent of the auroral oval in Discussion to support our conclusion that the group of observatories around De Bilt might indeed have been in proximity to the auroral oval for some time.
  
  We will discard Figure 5, which shows the variations in the vertical intensity, from the new manuscript. The series of the images will only be kept in the supplement; however, we will correct an error in them, because we mistakenly displayed as small variations also variations that we did not want to display (variations less than 100 nT). In the text of the manuscript, we will only draw attention to the significant variation of Z observed in Niemegk between 03:00 UT and 07:00 UT. In addition to the amplification of the vertical intensity due to the proximity of the equatorward boundary of the west-flowing currents in the electrojet, the magnetic fields generated by the induced currents in the conductive ground play an essential role in the variations of the vertical intensity. The Z variations are thus strongly dependent on the subsurface conductivity distribution (as Referee #2 alerted us to). These facts greatly complicate the interpretation of vertical intensity variations, and we cannot meaningfully interpret them in our work.
  
  Hourly data were not available in Munich for the studied period [H. C. Soffel, History of the Munich–Maisach–Fürstenfeldbruck Geomagnetic Observatory, Hist Geo Space Sci, 6, 65–86, 2015]. In Discussion we will explain the absence of Munich Observatory data in this sense. We are aware that, in the geomagnetic community, the Munich observatory is a respected observatory, and the reader will naturally wonder why the Munich data are not included here.
  
  Detailed comments:
  
  Line 117: The recording in the variation device was made on photographic paper, which was moved at a speed of 1 cm per hour using a clockwork machine.
  
  Fig. 1: We will add the SCs to Figs. 1 and 3.
  
  Fig 2: We will spell out the critical assessment of the Clementinum data clearer. In Figure 2, we will also use some graphical method to emphasize the doubtfulness about the second peak (e.g. by changing the colour of symbols and connecting lines to grey).
  
  Line 245, Lines 320ff, Lines 341ff, Lines 388ff: We will correct the manuscript as indicated in the responses to the general comments.
  
  Fig. 5: This figure will be dropped from the text of the manuscript. The figures with the vertical variations will only be kept in the supplement.
  
  We would like to thank Referee #2 also for his/her comments aimed at improving the language of the text.
  
  Citation: https://doi.org/10.5194/angeo-2023-12-AC2
RC3:
'Comment on angeo-2023-12', Hans-Joachim Linthe, 31 May 2023
Referee on Eduard Koci and Fridrich Valach: The record of the magnetic storm on 15 May 1921 in Stará Dala (present day Hurbanovo) and its compliance with the global picture of this extreme event
The authors contribute with this article a data set of the strong magnetic storm of 13-15 May 1921, which was up to now not yet considered. The publication extends with this important data set the possibilities of researching an interesting geomagnetic event. The authors found a suitable method to recalibrate the data. For my opinion the authors succeeded in this comprehensive challenge. I recommend this article for the publication in Anales Geophysicae. Only minor changes are necessary.
I have the following detailed comments:
Caption and last line of table 1, line 238 and further lines: At the time of the magnetic storm the observatories Potsdam and Seddin existed, but not Niemegk. Niemegk was started to be established only in 1929 and opened in 1932. Therefore, Niemegk should be skipped or only mentioned as the successor station of Potsdam and Seddin.

Line 107: Supplement ( S1) – There is no supplement and no Fig. S1

Line 281: Supplement ( S2… – There is no supplement and no Fig. S2

Line 205 and further lines: supplement – There is no supplement

Line 207 and further lines: The word “registration” should be changed into “recording”

Caption of Fig. 3: A better word for “one-hour means” is “hourly means” or “hourly averages”. The data of Potsdam and Seddin distinguish by several nT, especially during geomagnetic storms. Therefore, it is not common to present a plot of “Potsdam/Seddin”. You need to plot the data of Potsdam or Or, if you averaged the data of both observatories to mention about this.

3: A plot of hourly mean values is not very valuable due to the ramps every hour. I could offer you analogue copies of the original magnetograms of both observatories Potsdam and Seddin, which are available in the Niemegk data archive. It is up to you to decide, what is better to be used. You may get an image of such copies in:

Linthe, H.-J.: History of the Potsdam, Seddin and Niemegk Geomagnetic Observatories – First Part: Potsdam, History of Geo- and Space Sciences, submitted 2022

Hans-Joachim Linthe
Volunteer of the Niemegk Adolf Schmidt Geomagnetic Observatory
Helmholtz Centre Potsdam, GFZ German Recearch Centre for Geosciences
Citation: https://doi.org/10.5194/angeo-2023-12-RC3
- AC3: 'Reply on RC3', Fridrich Valach, 31 May 2023
  
  Dear Dr. Linthe,
  
  Thank you very much for your time and work devoted to our manuscript and for the comments. We are especially grateful to you for the offered magnetograms from Potsdam and Seddin.
  Here are our answers to the detailed comments:
  
  - We agree that only Potsdam and Seddin observatories existed when the studied magnetic event occurred. We will correct the manuscript accordingly. We will explain that the hourly means used in the study were the data of Potsdam or Seddin reduced to Niemegk. (They are available in that form in the WDC database <https://wdc.bgs.ac.uk/catalog/master.html>. The headers of the data files that we used list the coordinates of Niemegk.)
  
  - The supplement is available on the address <https://angeo.copernicus.org/preprints/angeo-2023-12/angeo-2023-12-supplement.zip>.
  
  - We will use the proposed correct terminology: ‘recording’ and ‘hourly means’.
  
  - The analogue copies of the original magnetograms of observatories Potsdam and/or Seddin would certainly be great information to include in the manuscript and might be used instead of Figure 3. We would be very grateful if you could provide such magnetograms for us.
  Best regards,
  
  Fridrich Valach and Eduard Koči
  
  Citation: https://doi.org/10.5194/angeo-2023-12-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (01 Jun 2023) by Christos Katsavrias

AR by Fridrich Valach on behalf of the Authors (20 Jun 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (21 Jun 2023) by Christos Katsavrias

RR by Anonymous Referee #2 (23 Jun 2023)

RR by Hans-Joachim Linthe (28 Jun 2023)

ED: Publish subject to revisions (further review by editor and referees) (04 Jul 2023) by Christos Katsavrias

AR by Fridrich Valach on behalf of the Authors (18 Jul 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (19 Jul 2023) by Christos Katsavrias

RR by Anonymous Referee #2 (24 Jul 2023)

Suggestions for revision or reasons for rejection

Comments on Re-reply to RC2', of Fridrich Valach, 19 July 2023

I am much satisfied with the authors’ responses to my main concerns.
However, in this new version there appear a number of weak points and flaws that need to be corrected before publication.

Line 205: The diurnal Bz variations at your station in summer are probably caused by the Sq currents. However, the amplitudes of those variations depend strongly on the solar flux. Are the solar activity conditions the same in 1910 and 1921?.
Owolabi et al. (2022) https://doi.org/10.1029/2021JA029903, showed in their Figs. 2 and 3 the strong dependence of Sq intensity on F10.7. For your period no F10.7 is available, but the sun spot number can also do a good job in this respect. This kind of dependence should be taken into account for the Bz scale factor or at least be discussed.

Fig. 1: I am confused by your lines marked, SCs. I the Introduction you mention three SCs: 13 May, 13:06 and 19:24 UT, 14 May, 14:05 UT. The first line may well mark the 13:06 UT event, but for the second, on 14 May, I find no correspondence in your list.

Your dealing with SCs, in general, is quite confusing for me. When comparing the listed times with the recordings at Sedin I can confirm the times listed for 13 May but not for the following day. Rather, a clear SC signature appeared at 16:0x UT, not at 14:05 UT, and there is definitely no SC at 22:10 UT, where the mark appears. The time of an SC should not be confused with the start of magnetic activity, which typically comes a few hours later. Please check again and correct the SC occurrences. I would love to see correct lines for all three SCs in both figures.

Line 267: In the present version no hourly means of the Sedin observatory appear. It would be highly desirable to add that former Fig. 3 to the final version as Fig. 4 because it well outlines the gross features of that storm. Without such an overview plot the global variations shown in the figure below are difficult to understand.

After removing these weaknesses, the manuscript is regarded as suitable for publication in Ann. Geophys.

Hide

ED: Publish subject to minor revisions (review by editor) (25 Jul 2023) by Christos Katsavrias

AR by Fridrich Valach on behalf of the Authors (10 Aug 2023) Author's response Author's tracked changes Manuscript

ED: Publish as is (12 Aug 2023) by Christos Katsavrias

AR by Fridrich Valach on behalf of the Authors (14 Aug 2023)

Short summary

We dealt with the most intense magnetic storm of the 20th century, which took place on 13–15 May 1921. It was also observed at Stará Ďala (present-day Hurbanovo). However, the record was not complete. We estimated the missing sensitivity scale values and presented the resulting digitized Stará Ďala’s data for 13–15 May 1921. The data were compared with the records from other observatories. The analysis suggests that the auroral oval got close to Stará Ďala in the morning hours on 15 May 1921.