Pith. sign in

REVIEW 4 major objections 5 minor 25 references

Long-Baseline VLF Observations of Solar Flares from Antarctica

T0 review · 4 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Antarctic VLF observations detect 60.8% of C- and M-class solar flares, rising to 84.2% for M-class events, and the paper argues the lower-frequency path (21.4 kHz) is consistently more sensitive than 24.0 kHz.

desk verdict First Antarctic long-baseline VLF flare data are a real contribution, but the headline detection rates lack a false-alarm baseline and should be read as provisional until null-testing is done. read the letter →

arxiv 2607.13913 v1 pith:T5UTVCMA submitted 2026-07-15 astro-ph.SR astro-ph.HEphysics.space-ph

Kamen Kozarev , Peter Petkov , Ivaylo Nachev , Veselka Radeva , Momchil Dechev , Galin Borisov , Anton Atanasov This is my paper · ORCID
classification astro-ph.SRastro-ph.HEphysics.space-ph
keywords solarflaresVLFradioionosphericD-regionsuddendisturbancesGOESX-rayAntarcticadetectionefficiencysuperposedepochanalysis
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Long-baseline Very Low Frequency (VLF) radio monitoring from a base in Antarctica can detect and time solar flares by watching how the D-region of the ionosphere responds to flare X-rays. Over a 16-day period in early 2025, the paper compares signals from two transmitters more than 11,000 km away — 21.4 kHz from Hawaii and 24.0 kHz from Maine — with GOES soft X-ray records for 250 C- and M-class flares. After removing the daily cycle with superposed epoch analysis, the authors use cross-correlation between VLF amplitude and GOES flux, counting a flare as detected when the maximum correlation within ±180 s reaches 0.7 or higher. They find that 60.8% of all flares, and 84.2% of M-class flares, meet that bar in at least one frequency, with 21.4 kHz consistently outperforming 24.0 kHz. The paper argues these results show Antarctic long-baseline VLF observations are a viable flare proxy, while cautioning that background variability and path-dependent propagation must be quantified.

What carries the argument

The central machinery is a three-stage pipeline: (1) superposed epoch analysis to build and subtract the mean diurnal amplitude curve at each frequency; (2) cross-correlation of the detrended VLF amplitude with GOES XRS-A and XRS-B soft X-ray flux over a symmetric ±180 s lag window, recording the maximum correlation ρ and its lag; (3) an operational detection criterion ρ ≥ 0.7 with |lag| ≤ 180 s, applied per radio–X-ray pairing and also combined across channels. A path-illumination metric, the sunlit fraction of each great-circle path, is used to stratify detection rates by geometry rather than by receiver-local day/night. This pipeline turns raw VLF amplitude records into a labeled set of d

What would settle it

Run the same superposed-epoch detrending and cross-correlation pipeline on VLF data from time windows with no GOES flare (or on GOES light curves shifted by hours), and count how many windows give ρ≥0.7. If this false-alarm rate approaches the reported 60.8% 'any-frequency' detection rate, the claimed detection efficiency cannot be attributed to flares.

Watch

Extended reading notes

Core claim

The central claim is that the Earth-ionosphere waveguide, probed over trans-equatorial paths longer than 11,000 km, responds reliably enough to flare-driven D-region ionization that it can serve as an operational proxy for GOES-class flare occurrence and timing. The evidence is a detection-efficiency analysis of 250 C/M flares: 152 of 250 (60.8%) are detected in at least one VLF–X-ray pairing, M-class flares reach 84.2% detection, and the 21.4 kHz path (NPM, Hawaii) detects 52.4% versus 40.4% for 24.0 kHz (NAA, Maine). The paper also documents a systematic asymmetric distribution of the peak-to-peak delay, with VLF peaks usually lagging the X-ray peak but a non-negligible fraction showing ne

Load-bearing premise

The detection statistics assume that a cross-correlation of at least 0.7 between detrended VLF amplitude and GOES flux, within a ±180 s lag, is a flare signature rather than a chance alignment; no null distribution or false-alarm analysis is provided, and the background variability acknowledged in Section 4.2 could produce spurious detections.

Editorial extensions

If this is right

  • A single-frequency VLF receiver at 21.4 kHz can serve as a low-cost flare monitor, catching roughly half of C-class flares and 84% of M-class flares without needing direct X-ray data.
  • Adding a second frequency (24.0 kHz) provides internal confirmation: only 32% of all flares, but 63% of M-class flares, are detected at both frequencies, giving a high-confidence subset.
  • Detection efficiency is strongly modulated by how much of the propagation path is sunlit; interpreting a VLF detection therefore requires knowledge of the path's illumination geometry, not just local time at the receiver.
  • Measured VLF–X-ray lags, including negative lags, should be treated as empirical timing indicators with tolerance, not as physical ionospheric delays, until spectral X-ray information is added.
  • The detection framework can be refined by adjusting the ρ threshold, and the main trends (21.4 kHz superiority, M-class advantage, dual-frequency confirmation) survive threshold choices from 0.5 to 0.8.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Inference: Because no false-alarm rate is given for the ρ≥0.7 threshold, the reported 60.8% detection efficiency likely includes some spurious correlations from residual background VLF variability; applying the same pipeline to flare-free time windows would bound the false-positive rate.
  • Inference: The negative-lag population (VLF peak before soft-X-ray peak) is consistent with hard X-ray (≳40 keV) contributing to early D-region ionization; combining the VLF timing with HXR observations would test whether negative lags correspond to harder flare spectra.
  • Inference: The path-illumination dependence suggests that a global network of long-baseline VLF paths, rather than a single receiver, could yield nearly continuous flare detection coverage by always keeping some sunlit paths in view.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents 16 days of continuous VLF amplitude observations at 21.4 kHz (NPM) and 24.0 kHz (NAA) received at the Bulgarian Antarctic base, together with GOES soft X-ray data for 250 C- and M-class solar flares. After removing the diurnal trend by superposed epoch analysis, the authors define a flare as 'detected in VLF' when the maximum positive cross-correlation with the GOES X-ray profile within a ±180 s lag window exceeds ρ_min = 0.7. They report overall detection efficiency of 60.8% (152/250), higher efficiency for M-class flares (84.2%) than C-class flares (56.6%), a consistent advantage of 21.4 kHz over 24.0 kHz, and systematically negative peak-to-peak delays Δt_peak. They also examine sensitivity to the correlation threshold and the dependence of detection on local day/night and path-integrated illumination.

Significance. The dataset is valuable: it is, to my knowledge, the first Antarctic long-baseline VLF flare study, and the paper publicly provides a complete flare list, per-event correlation results (Table 7), and a sensitivity analysis over correlation thresholds (Table 5). The authors are careful to label their detection measure as an operational, benchmark-based definition and to acknowledge limitations in Sect. 4.2. If the reported detection efficiencies and frequency ordering survive a proper false-alarm analysis, this would be a useful contribution to space-weather monitoring and to understanding D-region responses at trans-hemispheric paths. However, the central quantitative claims currently depend on an untested null hypothesis for the ρ≥0.7 criterion.

major comments (4)
  1. [Sect. 2.4, Table 7] The detection criterion requires the maximum cross-correlation ρ within a ±180 s lag window to exceed 0.7, so the timing condition |Δt|≤180 s is redundant by construction. The whole discriminator is ρ≥0.7. No null distribution or false-alarm analysis is presented. Table 7 shows many best lags pinned at +180 s or -180 s (e.g., IDs 3, 14, 17, 26, 30, 45), which is the hallmark of correlation maxima driven by partial overlap of smooth, unrelated trends rather than a well-centered flare response. Section 4.2 explicitly acknowledges residual background variability and 'apparent flare-like excursions,' but does not quantify how often such fluctuations alone would produce ρ≥0.7. Without such a null baseline, the headline efficiencies (60.8% overall, 84.2% M-class, and the 21.4 vs 24.0 kHz ordering) are unproven and could be substantially inflated by chance alignments, especially for C-class eve
  2. [Sect. 4.3, Table 5] The sensitivity analysis in Table 5 shows only that detection efficiencies decrease monotonically as ρ_min increases from 0.5 to 0.8. This monotonicity is exactly what would be expected under the null hypothesis of no flare-VLF association, because raising the threshold rejects more of any distribution, signal or noise. The table therefore does not establish that the observed detection count exceeds the expected false-alarm count at any threshold. The authors should compute the null efficiency (or false-alarm rate) from randomized flare times or from non-flare VLF windows at the same thresholds and report a significance measure (e.g., binomial p-value or false-alarm probability) for the observed efficiencies.
  3. [Abstract, Sect. 5] The Abstract and Conclusions state that VLF observations 'detect' 60.8% of flares and can 'serve as a viable proxy for solar flare occurrence and timing.' Given the operational definition in Sect. 2.4, the quantity measured is the fraction of GOES flares whose VLF signal correlates with the GOES X-ray profile, not an independent detection rate. The authors make this caveat clear in the body, but the abstract and conclusions overstate the result. If the null test requested in the first major comment shows a statistically significant excess over chance, the claims can be retained with explicit wording that the rates are correlation-based; without such a test, the current wording is not supportable.
  4. [Sect. 3.4, Table 4] The Δt_peak means in Table 4 are reported without standard errors. For example, the C-class 21A row has N=79 and σ=340 s, yielding SEM≈38 s, so the mean of -156 s would be about 4σ from zero; by contrast, the M-class 24A row has N=25 and σ=490 s, giving SEM≈98 s, making its mean of -125 s statistically indistinguishable from zero. Without SEMs or confidence intervals, the reader cannot assess the systematic negative-delay claim or compare rows. Please add the standard error of the mean or a confidence interval to Table 4.
minor comments (5)
  1. [Sect. 2.5] The 'frequency consistency check' is said to require a positive zero-lag correlation C0 'above a certain threshold,' but no threshold value is specified and this criterion is not used in the reported efficiencies. Either specify the threshold and report its results, or remove this paragraph.
  2. [Table 7 caption] The column header ρ_min is not defined in Sect. 2.2, where BL, BC, and C0 are defined. Presumably it denotes the minimum correlation over the lag window, but this should be stated explicitly.
  3. [Tables 2 and 3] Several illumination bins contain only one or two events (e.g., 1/1, 2/2, 4/5). Percentages based on such tiny samples are fragile; the counts are shown, which is good, but for n<5 it would be better to suppress the percentage or show a binomial confidence interval.
  4. [References and figures] There are small presentation errors: the Clilverd reference has 'SeppäLä' (should be 'Seppälä'); the Grubor reference misspells 'Šulić'; Fig. 4 uses the acronym BGPAO without defining it; and the caption of Fig. 2 has 'The trend is quite clearer in the lower frequencies.'
  5. [Sect. 3.4] The paper computes both the correlation lag Δt and the peak-to-peak delay Δt_peak, but does not compare them. A brief statement of whether the two timing measures agree in sign and magnitude for the detected subset would help the reader interpret the 'sluggishness' discussion.

Circularity Check

2 steps flagged · score 2.0 of 10

Mild definitional circularity: detection is defined as GOES-correlation rate and the timing criterion is redundant; no fitted parameters or self-citations, so score stays low.

  1. self definitional [Sect. 2.4 (Operational definition), Eq. (2); Table 1]
    "we adopt an operational definition of flare detectability based on the temporal correspondence between the de-trended VLF signal and the GOES soft X-ray flux. ... A flare is defined as “detected in VLF” if it satisfies two operational criteria ... (i) a shape-matching criterion, ρ≥ρmin ... (ii) a timing plausibility criterion, |Δt|≤Δt0 ... we compute the VLF flare detection efficiency as η=N_det/N_total"

    N_det counts flares whose VLF segment reaches ρ≥0.7 in cross-correlation with the GOES X-ray profile. Thus η is, by construction, the rate at which VLF resembles the GOES X-ray shape within GOES-defined flare windows. The headline “60.8% of flares are detected in VLF” is a restatement of this correlation rate, not an independent detection measurement. The paper is transparent about the operational character, and ρmin is a priori, so the circularity is mild, but the proxy claim adds no information beyond the definition used to construct the efficiency.

  2. self definitional [Sect. 2.4]
    "we compute the maximum positive cross-correlation coefficient, ρ, within a symmetric lag window of ±180s, together with the corresponding lag Δt at which this maximum occurs. ... we adopt ρmin=0.7 ... and Δt0=180s, consistent with the lag search window."

    Because Δt is defined as the lag at which the maximum ρ occurs inside the ±180s search window, the condition |Δt|≤Δt0=180s is satisfied for every event by construction. Criterion (ii) therefore adds no independent timing-plausibility constraint; only the ρ threshold discriminates detections from non-detections. This is a definitional tautology rather than a load-bearing element of the efficiency numbers, but it makes the two-criterion detection definition partly redundant.

full rationale

The paper is self-contained and benchmarked against an external GOES catalog; it contains no fitted physical parameters and no load-bearing self-citations. The central efficiency values are empirical counts, not predictions from a model, and the VLF data are independent observations. The only circular aspects are definitional. First, “detection” is defined as high cross-correlation with the GOES X-ray profile, so η=N_det/N_total restates the correlation rate rather than providing an independent detection measurement; the paper explicitly calls this an operational definition. Second, the timing criterion |Δt|≤180s is redundant because the maximizing lag is selected from within that same ±180s window. These issues are mild and partly acknowledged by the paper’s own caveats about background variability and the operational nature of the thresholds. The absence of a null/false-alarm distribution is a real statistical validation gap, but it is a correctness risk rather than a circularity, so it does not raise the score beyond 2.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new entities or physical parameters. Its central numbers are descriptive statistics of an observational dataset; the main 'free' choices are detection thresholds, which are a priori and sensitivity-tested. The key unexamined assumption is the significance of correlations against background VLF variability.

free parameters (2)
  • rho_min detection threshold
    Hand-chosen correlation threshold ρmin=0.7 for defining VLF detection (Sect. 2.4); affects all efficiency numbers, though sensitivity analysis (Table 5) shows trends robust.
  • dt0 timing window = 180 s
    Hand-chosen timing plausibility criterion |Δt|≤180s; redundant since the max correlation is searched within ±180s, so it imposes no additional constraint.
assumptions (4)
  • domain assumption VLF amplitude changes in the Earth-ionosphere waveguide are a reliable proxy for D-region ionization changes caused by solar flares.
    Standard theory in VLF remote sensing (Wait & Spies 1964; Thomson et al. 2005), invoked throughout §1.
  • domain assumption The GOES soft X-ray flare catalog is complete and accurate for the studied period.
    Flares are selected from NOAA/SWPC GOES catalog (Table 6); any misclassification directly affects detection efficiency.
  • ad hoc to paper Superposed epoch analysis removes the diurnal trend sufficiently that residual VLF variability is small compared to flare signatures.
    Used in §2.1 to detrend; residual variability is acknowledged in §4.2 as a limitation, but no quantitative false-alarm analysis is provided.
  • ad hoc to paper Cross-correlation of detrended VLF with GOES X-ray flux within ±180s lag identifies flare-related responses.
    Operational definition in §2.4; it presupposes that any high correlation is physically due to the flare rather than noise or other ionospheric disturbances.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Long-Baseline VLF Observations of Solar Flares from Antarctica." pith.science (2026). https://pith.science/paper/T5UTVCMA

@misc{pith2026260713913,
  author       = {Pith},
  title        = {Pith review of: Long-Baseline VLF Observations of Solar Flares from Antarctica},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T5UTVCMA}},
  note         = {Machine review of arXiv:2607.13913}
}
read the original abstract

We present long-baseline Very Low Frequency (VLF) observations of solar flare-induced ionospheric disturbances obtained at the Bulgarian Polar Astronomical Observatory (St. Kliment Ohridski Base) on Livingston island, Antarctica. Using continuous VLF transmissions at 21.4 kHz (NPM, Hawaii) and 24.0 kHz (NAA, Maine), propagating over trans-hemispheric paths exceeding 11000 km, we analyze observations of solar flares during the period 24 January--8 February 2025. After removing the strong diurnal signal via superposed epoch analysis, we analyse the flare-related perturbations in VLF amplitude and their correlation with GOES soft X-ray flux for 250 flares of C and M class. The long propagation paths provide enhanced sensitivity to flare-driven changes in D-region ionization. The observations reveal clear, frequency-dependent responses and measurable time delays between X-ray and VLF peaks. These delays, including cases of near-zero or negative lag for stronger events, highlight the role of flare spectral characteristics and D-region recombination processes. Our results demonstrate the scientific value of long-baseline Antarctic VLF observations for detecting and timing GOES-class flares, while also highlighting key limitations, such as background variability and path-dependent propagation effects, which must be quantified for reliable VLF-based flare monitoring.

Figures

Figures reproduced from arXiv: 2607.13913 by the authors.

Figure 1
Figure 1. The VLF observations obtained from the St. Kliment Ohridski Base during the period 01/24/2025 – 02/08/2025 at 21.4 kHz (green) and 24.0 kHz (red). K. Kozarev et al.: Preprint submitted to Elsevier Page 10 of 27 [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Daily superposed time series of 21.4 kHz (top) and 24.0 kHz (bottom) VLF observations. Different colors are used for each day. The black lines show the mean for each time step, and the gray shadings denote the standard deviations about the mean. K. Kozarev et al.: Preprint submitted to Elsevier Page 11 of 27 [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Detrended VLF signal in 21.4 kHz (green) and 24.0 kHz (red) during the studied period. The local daytime periods are shaded in yellow. Vertical dashed lines denote the starting times of >C7.0 flares [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison between the GOES X-ray and BGPAO VLF observations during a flare on 31.01.2025. K. Kozarev et al.: Preprint submitted to Elsevier Page 12 of 27 [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Stacked histograms of the distributions of the Δ𝑡 𝑝𝑒𝑎𝑘 parameter for C-class and M-class flares for the overall sample of 250 observed flares. K. Kozarev et al.: Preprint submitted to Elsevier Page 13 of 27 [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 2 canonical work pages

  1. [1]

    Wait, J. R. and Spies, K. P. , title =. 1964 , journal =

    1964

  2. [2]

    1965 , publisher =

    Physics of the lower ionosphere , author=. 1965 , publisher =

    1965

  3. [3]

    Thomson, N. R. and Rodger, C. J. and Clilverd, M. A. , title =. Journal of Geophysical Research , volume =. 2005 , pages =

    2005

  4. [4]

    Geophysical Research Letters , keywords =

    Ionosphere gives size of greatest solar flare. Geophysical Research Letters , keywords =

  5. [5]

    Journal of Geophysical Research (Space Physics) , keywords =

    Lower Ionosphere Sensitivity to Solar X-ray Flares Over a Complete Solar Cycle Evaluated From VLF Signal Measurements. Journal of Geophysical Research (Space Physics) , keywords =

  6. [6]

    Geophysical Research Letters , year = 2007, month = feb, volume =

    Solar flare induced D region perturbation in the ionosphere, as revealed from a short-distance VLF propagation path. Geophysical Research Letters , year = 2007, month = feb, volume =

    2007

  7. [7]

    Annales Geophysicae , year = 2008, month = jun, volume =

    Classification of X-ray solar flares regarding their effects on the lower ionosphere electron density profile. Annales Geophysicae , year = 2008, month = jun, volume =

    2008

  8. [8]

    2014 , issn =

    Sensing the Earth’s low ionosphere during solar flares using VLF signals and goes solar X-ray data , journal =. 2014 , issn =. doi:https://doi.org/10.1016/j.asr.2014.02.022 , author =

    work page doi:10.1016/j.asr.2014.02.022 2014

Show all 25 references
  1. [9]

    , title =

    Cannon, Paul S. , title =. Space Weather , volume =. doi:https://doi.org/10.1002/swe.20032 , year =

    doi

  2. [10]

    Space Weather , keywords =

    The May 1967 great storm and radio disruption event: Extreme space weather and extraordinary responses. Space Weather , keywords =

    1967

  3. [11]

    Belcher, Samuel R. G. and Clilverd, Mark A. and Rodger, Craig J. and Cook, Sophie and Thomson, Neil R. and Brundell, James B. and Raita, Tero , title =. Space Weather , volume =. doi:https://doi.org/10.1029/2021SW002820 , year =

    doi

  4. [12]

    2025 , issn =

    Low-latitude sub-ionospheric VLF radio signal disturbances due to solar flares: Effects on the attenuation and phase velocities of the waveguide modes , journal =. 2025 , issn =. doi:https://doi.org/10.1016/j.jastp.2025.106433 , author =

    2025

  5. [13]

    Journal of Atmospheric and Terrestrial Physics , keywords =

    Experimental daytime VLF ionospheric parameters. Journal of Atmospheric and Terrestrial Physics , keywords =

  6. [14]

    Ferguson, J. A. , title =. 1998 , month =

    1998

  7. [15]

    1961 , publisher=

    The Wave-guide Mode Theory of Wave Propagation , author=. 1961 , publisher=

    1961

  8. [16]

    Journal of Atmospheric and Solar-Terrestrial Physics , keywords =

    Solar flare induced ionospheric D-region enhancements from VLF phase and amplitude observations. Journal of Atmospheric and Solar-Terrestrial Physics , keywords =. doi:https://doi.org/10.1016/j.jastp.2003.09.009 , adsnote =

    2003 doi

  9. [17]

    Space Weather , keywords =

    Remote sensing space weather events: Antarctic-Arctic Radiation-belt (Dynamic) Deposition-VLF Atmospheric Research Konsortium network. Space Weather , keywords =. doi:https://doi.org/10.1029/2008SW000412 , adsnote =

    doi

  10. [18]

    and Gallagher, Peter T

    Hayes, Laura A. and Gallagher, Peter T. and McCauley, Joseph and Dennis, Brian R. and Ireland, Jack and Inglis, Andrew , journal =. Pulsations detected in ionospheric D-region are synchronized with flare X-ray pulsa-tions , year =

  11. [19]

    Mitra, A. P. , title =. 1974 , doi=

    1974

  12. [20]

    and O'Hara, Oscar S

    Hayes, Laura A. and O'Hara, Oscar S. D. and Murray, Sophie A. and Gallagher, Peter T. , journal =. Solar Flare Effects on the Earth's Lower Ionosphere , year =

  13. [21]

    Appleton, E. V. , journal =. A note on the ‘sluggishness’ of the ionosphere , year =

  14. [22]

    and Grubor, D

    Žigman, V. and Grubor, D. and Šulić, D. , journal =. D-region electron density evaluated from VLF amplitude time delay during X-ray solar flares , year =

  15. [23]

    Role of hard X-ray emission in ionospheric D-layer disturbances during solar flares , year =

    Briand, Carine and Clilverd, Mark and Inturi, Srivani and Cecconi, Baptiste , journal =. Role of hard X-ray emission in ionospheric D-layer disturbances during solar flares , year =

  16. [24]

    and Briand, C

    Teysseyre, P. and Briand, C. and Marshall, R. and Cohen, M. , journal =. Effect of Ground Conductivity on VLF Wave Propagation , year =

  17. [25]

    Ferguson, J. A. , journal =. Computer programs for assessment of long-wavelength radio communications, version 2.0: user’s guide and source files , year =

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.