the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
No statistical link between proton aurora and Pc1 pulsations in the high-latitude dayside using ground-based measurements
Abstract. In order to test the hypothesis that EMIC waves are responsible for the acceleration of auroral protons, we have used spectrograph measurements of proton aurora over Svalbard alongside co-located magnetometer measurements of Pc1 pulsations. No evidence of a link between proton aurora and Pc1 waves was found by three different methods. Firstly, accelerated protons and Pc1 pulsations have no coincident occurrence. Secondly, the proton energy spectrum does not change between Pc1 activity and quiet times. Finally, no imprint of the EMIC wave is found in periodicity of the intensity and blue-shift of the proton H-α line, unlike in flickering electron aurora where intensity fluctuations are caused by EMIC waves. We find no evidence that EMIC waves are the mechanism responsible for accelerating auroral protons in the high-latitude dayside, at least based on the considered ground-based data of proton aurora and magnetic Pc1 pulsations.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-2052', Anonymous Referee #1, 18 Nov 2024
“No statistical link between proton aurora and Pc1 pulsations in the high-latitude dayside using ground-based measurements.” by Rowan Dayton-Oxland et al.
The manuscript presented a statistical relationship between proton auroras and Pc1 wave activities at KHO near the cusp location. There are some ideas for proton precipitation, but the authors focused on the effects of pitch angle scattering by EMIC waves in the magnetosphere through the statistical results. The magnetic field line curvature scattering of protons should be a major factor at higher latitudes (e.g., Ma, L. et al., JGR Space Physics, 127, e2022JA030843, 2022). However, the authors do not discuss and analyze whether the observation location is inside or outside the isotropic boundary (IB). The authors should classify the location as inside or outside the IB in order to discuss the effects of pitch angle scattering by EMIC waves without the effects of the field line curvature scattering. Therefore, this manuscript, which concludes that there is no effect of EMIC waves on proton aurora generation, is a major missing discussion because it does not clearly and quantitatively distinguish whether the events analyzed in this study are outside or inside the IB. Therefore, I recommend that the authors should resubmit the manuscript with significant modifications. Please see detailed comments below.
[Major comments]
1) Without statistically comparing Pc1 and proton aurora at the observation site outside or inside from the IB for all events, it is quantitatively difficult to distinguish between the effects of field line curvature scattering and pitch angle scattering by EMIC waves. Please add the statistical results of a ratio Rc/rho, where Rc is the curvature radius of the field line, rho is the effective particle gyroradius (see, Sergeev et al., JGR, 98(5), pp.7609-7620, 1993), it would be useful for the additional discussion.
2) Pc1 waves can easily propagate thousands of kilometers horizontally through the ionospheric duct. Therefore, taking into account the effects of the ionospheric duct, the authors should classify Pc1 waves as the ionospheric penetration from the magnetosphere along the magnetic field lines at KHO, or as horizontally propagating Pc1 waves.
These two important issues for the relationship between the proton aurora and Pc1 waves, which do not take into account the effects of horizontal propagation and the location of the IB, cannot verify the presence or absence of the influence of pitch angle scattering by the EMIC waves. We hope that these comments will be useful for your further research.
Citation: https://doi.org/10.5194/egusphere-2024-2052-RC1 -
AC1: 'Reply on RC1', Rowan Dayton-Oxland, 21 Feb 2025
Thank you for the useful discussion, to clarify our conclusions please see our response to Reviewer 2.
Thank you for bringing curvature scattering to our attention. This seems to be a likely candidate for precipitation in Svalbard although we can’t rule out other mechanisms. However, we don’t believe this affects our conclusion that EMIC waves cannot be the accelerating/precipitating mechanism for the majority of proton aurora observed over Svalbard. Since we see no change in the energy spectrum of protons when Pc1 are present, it is reasonable to conclude from our observations that they show no evidence for EMIC acceleration.
Since the presence or absence of curvature scattering does not affect our conclusions, we think that calculating whether we are inside or outside of the IB is beyond the scope of this paper. A simpler metric that could be useful would be looking at the open/closed field line boundary based on e.g. SuperDARN or Ovation Prime. Since only a small number of Pc1/proton aurora events are available we aren’t convinced that any additional results based on the IB classification would be significant. We will add further discussion on the IB and curvature scattering in our introduction and discussion sections and thank the reviewer for bringing this alternate mechanism to our attention.
Reviewer 1 has summarized our conclusions - ‘which concludes that there is no effect of EMIC waves on proton aurora generation’. We don’t make this conclusion in general and will rephrase our conclusions section to avoid misunderstanding. We are happy that in general EMIC waves have a role in generation of proton aurora, particularly in the night-side proton aurora, and have highlighted relevant studies in our introduction. We are concluding that our ground-based observations are not consistent with EMIC waves as the principal acceleration/precipitation mechanism in the Svalbard region dayside. We will address this comment by rewriting our conclusions section to be more clear on this.
To address the comment on EMIC waves propagating horizontally through the ionospheric duct, we understand that this would lead to an increase in EMIC observations on the ground, due to them travelling over a larger horizontal area (Noh+2022, 10.1029/2022JA030262) - which also states that 75% of EMIC waves are observed simultaneously on the ground. This supports our conclusion that since for the majority of proton aurora there is no Pc1 activity, it is not likely that EMIC are the principal generation mechanism, regardless of the direction of propagation.
Citation: https://doi.org/10.5194/egusphere-2024-2052-AC1
-
AC1: 'Reply on RC1', Rowan Dayton-Oxland, 21 Feb 2025
-
RC2: 'Comment on egusphere-2024-2052', Anonymous Referee #2, 22 Nov 2024
This paper statistically examines the correlation between proton auroras and Pc1 waves observed at the same ground locations. The authors found no correlations between (1) proton flux, (2) proton energy spectrum, (3) intensity of H-alpha lines, and Pc1 activity. Although all the results in the paper indicate that there is no clear evidence for a relationship between Pc1 waves and proton auroras, there are still some methodological points that need to be clarified and considered to draw a definitive conclusion.
Major comments
1) Regarding the contingency table (Table 1):
The occurrence rate of proton auroras is much higher than that of Pc1 wave events. Thus, it is somewhat reasonable to conclude that the majority of proton auroras are not driven by (or correlated with) Pc1 waves. However, it is still possible that Pc1 waves are highly likely to drive proton auroras when they occur. Considering the low occurrence rate of Pc1 waves, the difference in numbers between p11 and p01 seems large. Therefore, it should not be definitively stated that there is no link between Pc1 waves and proton auroras.
Additionally, the authors should consider a more lenient contingency condition for identifying events. Counting events strictly based on simultaneous observations seems overly restrictive. It is plausible that proton precipitation and wave propagation occur at slightly different times due to various factors, including differing propagation times. Even wave activities coincidentally observed in space and on the ground at conjugate locations can have different start times and durations.
2) Potential statistical bias:
In L100, the authors mention excluding Pc1 events contaminated by auroral noise. How much time (or how many Pc1 events) was excluded from the study? If this represents a large portion of the dataset, the statistics may be biased. Moreover, if the exclusion of wave events is related to auroral activities, the authors should handle the dataset carefully to avoid introducing bias into the analysis.
Minor comments
L10: Provide the energy range of these energetic protons.
L18: Provide the frequency range of Pc1 waves.
L24: Instabilities causing EMIC waves are not a mixture of hot and cold plasmas. They are driven by either anisotropy or ion beam instabilities.
L58: Approximating the L-shell coverage of the KHO observatory would be helpful for mapping this process to magnetospheric activities.
L78: Why is this event considered “unstructured”?
L88: The identification method for Pc1 waves is not described in Appendix A. Please explain how the authors processed the search coil data and identified wave activities.
L100: The authors mention that data recorded during daylight were eliminated, but later state that data from 6–18h MLT were considered. This paragraph should be rewritten for clarity. Additionally, the authors should specify how much data were excluded due to daylight or bright cloud, as this could significantly affect the statistics, as noted in the second major comment.
L103: Clarify the meaning of “correlation.” Does this refer to the coincidence of the two phenomena?
L108: Elaborate on why an odd ratio of 1.856 is not significant. From my understanding, 1.923 is the maximum odd ratio from 1,000 subsample trials. If true, this value may lack significance, as it could vary depending on the number of subsamples or sampling runs. If this is incorrect, please provide a clear explanation of how the threshold was determined. The odd ratio’s statistical significance should also be considered, as it would approach 1 if the two phenomena were entirely independent with a sufficiently large sample size.
L114: Explain how the 200 minutes of data were determined. Using only 200 minutes of proton aurora data without Pc1 waves represents significant under-sampling and could introduce statistical bias.
L128: Change "proton events" to "proton aurora events."
L130: Why is a 1–10 s periodic signal required? Is this related to the Pc1 frequency?
L132: The authors state that they used FFT to analyze signal periodicity but do not present any FFT results in the paper.
L161: Could the authors provide references or background information for this process? To my knowledge, this complicated process is unlikely to result solely from linear or quasi-linear wave-particle interactions.
L175: If the alternative explanations also suffer from the same issue (transience), they cannot be considered valid alternatives.
L187: The results do not indicate that EMIC waves are not an acceleration mechanism. EMIC waves can still accelerate ions, but they are not the governing mechanism responsible for the majority of proton auroras.
Citation: https://doi.org/10.5194/egusphere-2024-2052-RC2 -
AC2: 'Reply on RC2', Rowan Dayton-Oxland, 21 Feb 2025
Thank you for the detailed and useful comments.
- Firstly I feel like our conclusions were not well conveyed in the original manuscript which I will change - this should solve several of the comments. Our conclusions should be phrased as - through the ground-based evidence of proton aurora and Pc1 measurements, we cannot see a statistically significant link between Pc1 and proton aurora, contrary to our expectations. It seems that EMIC waves are not responsible for the majority of proton acceleration in this high-latitude dayside region, and another mechanism is likely needed. It is certainly still possible that Pc1 drives proton aurora when it occurs, but from our observations it is not true in the majority of cases. We state that there is no significant link in our observations, not that there is significantly no link in the physical phenomena. We will make sure to rephrase our conclusions so that this distinction is clear. Our contingency condition has an overlap of 10-minutes, so we aren’t only considering exactly simultaneous observations. We are confident that 10-minutes is a long enough overlap to account for propagation time to the ground.
- Only a few hours over the entire season of observations were removed due to auroral noise contamination, and those times were removed from both datasets. Futhermore, auroral noise contamination typically occurs during energetic substorm electron precipitation on the nightside. Since we are investigating dayside phenomena this only represents a small fraction of the dataset. We will add some sentences in the method section, detailing how much auroral noise was removed and how it was identified in a ‘field-guide’ similar to the one given for proton aurora.
Minor comments
Again, thank you for the detailed comments. We are happy to add these suggestions to the manuscript.
L10-58: These will be added, the L-shell at Svalbard is 12
L78: This Pc1 event is considered unstructured because it lacks the bead-like temporal structure usually seen in EMIC Pc1 further south.
L88: More detail will be given in the method section - search coil data was pre-processed by the instrument owners, see our data availability statement. Pc1 waves were identified as high-amplitude and narrow-bandwith features in the Pc1 frequency range. We can provide a field-guide for Pc1 pulsation identification.
L100: Since our observatory is located at 78N, we see no daylight as our observing season is polar night. Later in the season there is occasionally daylight around local noon. The bigger problem for observation is moonlit cloud which we can provide statistics on, and was removed from both data sets.
L103: We refer here to the results of the Odd’s ratio, there is no significant temporal correlation between the phenomena. We will clarify in the manuscript.
L108: We will add a paragraph here describing the statistics of the Odd’s ratio outputs for randomised shuffles, accounting for the number of subsample runs.
L114: 200 minutes of data were chosen at random from the available dataset to give a comparable integration time to the data where Pc1 are present.
L128: Will update
L130: Yes, this is related to the Pc1 frequency. We did look for a broader range of frequencies up to 60s, but we expect the frequency to be comparable to the Pc1 frequency see (Ozaki+2018, 10.1002/2017GL076486) (Whiter+2010, 10.1029/2010JA015805).
L132: The FFTs didn’t turn up any signal, similar to the autocorrelation plot. We didn’t think another plot showing no signal was necessary.
L161: This sentence isn’t really necessary and we will remove it, as on reflection this was speculation on our part. We will be adding more discussion on simultaneous observation of EMIC and Pc1 as response to Reviewer 1.
L175: We agree - we will rewrite this sentence.
L187: We agree with this as well - the sentence will be rephrased for clarity.
Citation: https://doi.org/10.5194/egusphere-2024-2052-AC2
-
AC2: 'Reply on RC2', Rowan Dayton-Oxland, 21 Feb 2025
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
125 | 0 | 0 | 125 | 0 | 0 |
- HTML: 125
- PDF: 0
- XML: 0
- Total: 125
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
Country | # | Views | % |
---|---|---|---|
United States of America | 1 | 60 | 42 |
United Kingdom | 2 | 20 | 14 |
Japan | 3 | 13 | 9 |
China | 4 | 9 | 6 |
Germany | 5 | 6 | 4 |
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
- 60