{"id":"fac7bd06-b57d-4607-996e-e48572349d4d","arxiv_id":"2505.18296","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The magnetic ejecta of two interplanetary CMEs remain near-isothermal (Gamma_eff = 0.88 and 0.76) at 1 AU, with Kolmogorov-like turbulence in the fast event and shallower, less developed turbulence in the slow event.","lead":"This study analyzes Wind spacecraft data from two coronal mass ejections that reached Earth in 2011 and 2018. It finds both magnetic ejecta stayed near-isothermal at 1 AU, with different turbulence properties between the fast and slow events.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The near-isothermal Gamma_eff claim depends on 6-point window fits selected by CC>0.8; thousands of overlapping trials make chance high-CC windows likely, and the paper reports no null test, uncertainties, or retained-window counts.","rationale":"After reading in good faith, I agree with the reader's conditional verdict. The paper is a competent two-event study using public Wind data, standard spectral/PVI methods, and a reproducible fitting recipe; the turbulence part is largely descriptive and internally plausible. The central thermal claim, however, is more fragile than the text admits. The Gamma_eff result is dominated by Gamma_e, which is estimated from very short, overlapping windows with a high-correlation selection. With thousands of trials, the CC>0.8 cutoff does not protect against false positives, and the p-value typo (p<0.05 in Section 2.1 vs p<0.5 in Section 3.2) makes the actual selection criterion ambiguous. The absence of uncertainties on Gamma_e, Gamma_p, and Gamma_eff is a serious omission for a claim of near-isothermality. The permutation test I propose would settle whether the selection alone can produce the reported means; the direct Gamma_eff fit would settle whether the weighted-average formula corresponds to the actual combined T-n relation. I do not call for rejection: the data and methods are sufficiently documented that the requested checks can be run, and the authors' near-Sun comparison is a prior modeling result from the same group that should be checked for consistency but is not circular in the present analysis. If the checks pass, the paper can be accepted; if they fail, the near-isothermal conclusion should be softened. The reader identified the window-fitting assumptions as the weakest point; I agree and add the multiple-testing/permutation mechanism, hence partial agreement. Minor typos (labeling electron/proton Gamma in Figure 4, 'faster ICME2' in the abstract) should be fixed but do not change my verdict.","tokens_in":19307,"tokens_out":11607,"duration_ms":103729,"concrete_test":"Carry out a permutation (surrogate) test on the Wind/SWE time series for each ME region: keep the density series fixed, randomly permute the temperature series, and apply the identical 6-point moving-window CC>0.8 (and p<0.05) selection, computing the mean Gamma for each surrogate. Repeat ~1000 times. If the observed mean Gamma_e or Gamma_p falls inside the surrogate distribution, the selection procedure alone can produce the reported values. As a second check, compute Gamma_eff directly by fitting log[(Te+Tp)/2] vs log n over the same accepted windows (or over the full ME) and compare with the temperature-weighted formula; if either event yields a direct Gamma_eff >= 1.2, the near-isothermal conclusion is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline result is Gamma_eff = 0.88 (ICME1) and 0.76 (ICME2), interpreted as near-isothermal ME plasma at 1 AU and consistent with the authors' near-Sun FRIS results. This number is the temperature-weighted average (Section 3.3) of Gamma_e (0.42/0.52) and Gamma_p (1.8/1.47); since Te > Tp, Gamma_eff is dominated by the electron values, so the entire conclusion rests on the reliability of Gamma_e.\n\nGamma_e and Gamma_p come from moving 6-point windows fitted in log T vs log n, retaining windows with CC > 0.8 and p < 0.05 (Section 2.1; Section 3.2 says p < 0.5, an unresolved typo). The windows are moving with a step equal to the cadence, so they overlap by five points; for multi-hour ME intervals at 9-s cadence this is thousands of highly correlated trials. Under the null hypothesis of no T-n correlation, a 6-point sample has roughly 5-6% chance of |CC| > 0.8, so dozens to hundreds of spurious high-CC windows are expected. Because the reported mean Gamma is averaged only over the selected windows, chance correlations can bias it, especially when windows cross boundaries or current sheets and produce strong artificial gradients. The paper does not report how many windows were retained, the Gamma distribution, or any multiple-comparison correction, and it gives no uncertainties on Gamma_e, Gamma_p, or Gamma_eff. A modest bias in Gamma_e (~0.15-0.2) would move Gamma_eff from ~0.8 to ~1.0-1.1, erasing the near-isothermal claim.\n\nThe additional assumption that each retained window samples a single polytropic population is plausible but untested; the windows with strong internal gradients are the most likely to pass the CC threshold. Thus the central claim is not yet supported at the stated confidence.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a two-event case study of a fast ICME (2011 September 26) and a slow ICME (2018 August 25) observed at 1 AU by the Wind spacecraft, analyzing the thermal state and magnetic turbulence properties of the pre-ICME solar wind, the sheath, five equal segments of the magnetic ejecta (ME), and the post-ICME wind. Using 9-second electron and 92-second proton density/temperature data, the authors fit the polytropic relation T n^{1−Γ} = constant in moving 6-point windows, retain fits with CC > 0.8 and a stated p-value threshold (given as p < 0.05 in Section 2.1 and p < 0.5 in Section 3.2), and report region-averaged Γ_e and Γ_p. The headline result is the effective polytropic index of the ME, Γ_eff ≈ (Γ_e T_e + Γ_p T_p)/(T_e + T_p), equal to 0.88 (ICME1) and 0.76 (ICME2), interpreted as near-isothermal, strongly sub-adiabatic states implying sustained heating at 1 AU and consistent with the authors' earlier FRIS-model near-Sun results (Khuntia et al. 2023). Supporting analyses estimate inertial- and dissipation-range magnetic spectral indices α_B from FFT power spectra, magnetic compressibility C_B, and PVI-based intermittency statistics, leading to claims about Kolmogorov-like turbulence in ICME1's ME, less developed turbulence in ICME2's ME, and qualitative correlations between PVI events and local heating or compressibility.","tokens_in":19670,"tokens_out":14622,"duration_ms":115402,"significance":"If the near-isothermal Γ_eff result survives scrutiny, it provides a valuable link between in-situ thermal states of ICME magnetic ejecta at 1 AU and remote-sensing/model-based near-Sun states, strengthening the case for continuous heating of CME plasma throughout heliospheric propagation, which is directly relevant to CME evolution and space-weather modeling. The paper's strengths include the use of publicly available high-resolution Wind data (reproducibility), the systematic dissection of each ejecta into five parts with explicit discussion of boundary contamination, the combination of three complementary turbulence diagnostics (PSD slopes, compressibility, PVI) with the thermal analysis, and an explicit, falsifiable prediction that Γ_eff should remain near unity across heliocentric distances. The main limitation is statistical: the central quantitative claim rests on a windowed-correlation procedure whose uncertainties and selection biases are not characterized, and the manuscript contains an unresolved threshold inconsistency and an abstract-level typo that reverses the comparison object.","major_comments":[{"comment":"The acceptance criterion for 'reliable' polytropic fits is stated as p < 0.05 in Section 2.1 but as p < 0.5 in Section 3.2 (in the caption of Figure 4). These thresholds differ by an order of magnitude, and because all reported mean Γ_e and Γ_p values are computed only from windows passing this criterion, the manuscript must state the operative threshold consistently and demonstrate that the regional means are insensitive to it; as written, a reader cannot determine which filter produced the headline Γ_e values, even though the binding constraint is likely CC > 0.8.","section":"Section 2.1 vs Section 3.2"},{"comment":"The moving-window fits overlap by five of six data points (the step equals the cadence), so multi-hour ME intervals at 9-s cadence produce thousands of highly correlated trials; under the null hypothesis of no T–n correlation, a 6-point sample has roughly 5% probability of |CC| > 0.8, implying that dozens to hundreds of windows are expected to pass the filter by chance. The paper reports neither the number of retained windows per region, nor the full Γ distribution including the rejected windows, nor any null test (e.g., shuffled-density or lagged trials), and it provides no uncertainties on the fitted slopes or on the mean Γ_e and Γ_p. The skeptical concern about selection bias is therefore well-founded, and the manuscript should address it with retained-window counts, bootstrap or fitting uncertainties on each regional mean, and a null test.","section":"Section 2.1"},{"comment":"Γ_eff = (Γ_e T_e + Γ_p T_p)/(T_e + T_p) is evaluated from single mean values with no uncertainty propagation even though it is the paper's headline result, and because T_e > T_p, Γ_eff is dominated by Γ_e. For ICME1, an upward bias of about 0.17 in Γ_e — well within the range made plausible by the selection issue raised above — shifts Γ_eff from 0.88 to roughly 1.0, making it indistinguishable from exactly isothermal and destroying the quantitative 'substantial heating' claim; a similar but smaller shift affects ICME2. In addition, the approximation T_e + T_p ∝ n^{Γ_eff−1} treats the ratio T_p/T_e as locally constant, which is inconsistent with the very different reported values Γ_e ≈ 0.4–0.5 and Γ_p ≈ 1.5–1.8 (T_p/T_e ∝ n^{Γ_p−Γ_e}); the authors should quantify the error of this local-logarithmic-derivative approximation rather than assuming it.","section":"Section 3.3"},{"comment":"The choice of six data points and the CC > 0.8 threshold is justified only by the statement that it yielded the maximum number of reliable Γ estimates, with no sensitivity analysis shown. Since the regional mean Γ values and therefore Γ_eff depend on these choices, the paper should present how the means and their scatter vary with window length (3–10 points) and with the correlation threshold (e.g., CC from 0.7 to 0.9) before the reported Γ_e and Γ_p can be treated as robust.","section":"Section 2.1"}],"minor_comments":[{"comment":"In the abstract and in Conclusions item 4, 'the ME of slower ICME2 is less affected by the ambient medium than the faster ICME2' should read 'faster ICME1'; the same error appears in the abstract's dissipation-scale sentence.","section":"Abstract and Section 4"},{"comment":"The caption states 'The derived PVI values for ICME1 (c) and ICME2 (d)', which conflicts with the text in Section 3.5 referring to 'Figure 5(b) and (d)' for the PVI panels; panel (b), not (c), is presumably the ICME1 PVI panel.","section":"Figure 5 caption"},{"comment":"The sentence 'Figure 4(a) and (b) show the polytropic index for electron Γp and proton Γe' has the species labels reversed relative to the figure caption, which correctly assigns panels (a,c) to Γ_e and (b,d) to Γ_p.","section":"Section 3.2"},{"comment":"The sentence 'The transition from Gaussian to non-Gaussian magnetic field fluctuations is observed around PVI = 3' has no citation attached; the Osman et al. (2011a,b) references appear only in the next sentence and should be cited directly at the transition statement.","section":"Section 3.5"},{"comment":"The phrase 'to derived proton polytropic index' should read 'to derive the proton polytropic index'; also re-read the sentence beginning 'The higher resolution of the electron data captures finer variations' for clarity, since the claim that the resolution difference does not affect the regional means is asserted without justification.","section":"Section 2.1"},{"comment":"The mean in-situ parameters are presented without uncertainties or spread measures; adding standard errors or interquartile ranges would help the reader evaluate the significance of the differences between regions discussed in Sections 3.1 and 3.2.","section":"Table 1"},{"comment":"Spectral slopes α_B are quoted to two significant figures (e.g., −1.6 versus −1.7) without fitting uncertainties; given that differences of ~0.1 are interpreted physically, confidence intervals on each fitted slope should be reported alongside the R² values shown in the appendix.","section":"Section 3.4 and Appendix A"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the journal's scope as a two-event case study, and the framework is publishable in principle. My main concern is that the headline claim is stated with a precision (Γ_eff = 0.88 and 0.76) that the underlying statistics do not yet support; I would want the sensitivity analysis and uncertainty quantification described in the major comments before endorsing the near-isothermal interpretation. The comparison with the authors' own FRIS model is legitimate and not circular, but the in-situ derivation must stand on its own, and the emphasis on self-citations should not compensate for the missing internal statistical validation. The unresolved p-value discrepancy (0.05 versus 0.5) and the reversed comparison-object typo in the abstract suggest a hasty final pass, and both should be corrected at the next revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: if you need a quick reference for ICME thermal/turbulence structure at 1 AU, this is a decent two-event study. But the central near-isothermal claim is built on window fits that aren't error-bounded or corrected for multiple comparisons, so I'd treat the specific Gamma_eff values as provisional.\n\nWhat's actually new: they apply the interval-based polytropic fitting (Dayeh & Livadiotis style) to a fast and slow ICME and try to connect the 1 AU thermal state back to their own FRIS-model near-Sun results. The turbulence diagnostics (spectral slopes, magnetic compressibility, PVI) are standard, but they lay them out carefully for each region. The paper is transparent about data and methods, re-implementable, and the qualitative story—sheaths hotter and more intermittent, ME near-isothermal, slower ICME with less developed turbulence—is coherent.\n\nThe soft spot is the statistics of Gamma_e and Gamma_p. They use moving 6-point windows at 9-s cadence, which overlap by five points, so there are thousands of correlated trials. Filtering on CC > 0.8 and p < 0.05 will capture many spurious correlations, and the paper gives no number for retained windows, no slope uncertainties, no null test, no multiple-comparison correction. Electron gamma values around 0.4–0.5 are driving the near-isothermal Gamma_eff because Te >> Tp; a modest bias in Gamma_e could push Gamma_eff from ~0.8 up to ~1.0–1.1. That would erase the headline result. So the central claim is plausible but not yet supported at the stated confidence. Also note the p-value threshold inconsistency (p<0.05 in Section 2.1, p<0.5 in Section 3.2) and a couple of label typos (e.g., 'faster ICME2' in conclusion point 4; swapped Gamma_p/Gamma_e labels in the Figure 4 text). Those are easy fixes.\n\nThe comparison to the near-Sun states leans on their own FRIS model, but the in-situ measurements are independent, so I don't see circularity in the main inference. The sample is two events, so treat this as a case study, not a statistical result.\n\nBottom line: it deserves referee time, but it needs a revision that puts error bars on the polytropic indices, reports how many windows survived, and ideally includes a null/synthetic test showing the high-CC windows aren't just multiple-comparison noise. If that comes back clean, the near-isothermal claim would be a useful addition. I'd send it to review with the expectation that the statistics get strengthened.","headline":"Two-event case study with a plausible near-isothermal result; the statistics behind the headline need real work before it can be trusted.","tokens_in":20301,"tokens_out":2444,"would_cite":false,"duration_ms":18872,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The central claim is that the magnetic ejecta of coronal mass ejections are still being heated at 1 AU, with effective polytropic indices near isothermal (0.88 and 0.76) that match near-Sun states.","keywords":["coronal mass ejections","ICME","polytropic index","solar wind turbulence","magnetic ejecta","plasma heating","magnetic compressibility","intermittency"],"falsifier":"Recompute $\\Gamma_e$ and $\\Gamma_p$ without the correlation filter, or with windows aligned to plasma parcels identified by composition and velocity, and check whether $\\Gamma_{\\rm eff}$ stays near 1; if unfiltered or parcel-aligned values drift toward $\\Gamma=5/3$ or outside roughly 0.7--1.3, the sustained-heating claim fails. Resolving the stated $p<0.05$ versus $p<0.5$ threshold inconsistency and repeating the fit over all windows would settle the point directly.","tokens_in":19055,"feed_emoji":"☀️","tokens_out":10303,"duration_ms":100372,"temperature":0.7,"pith_summary":"This paper uses in-situ measurements at 1 AU to reconstruct the thermal and turbulent state of two interplanetary coronal mass ejections (ICMEs), one fast and one slow. Its central claim is that the magnetic ejecta — the CME's magnetized plasma body — is near-isothermal at Earth: the effective polytropic index, a thermodynamic measure where 1 means constant temperature and 5/3 means adiabatic expansion, is 0.88 for the fast 2011 event and 0.76 for the slow 2018 event. That implies expansion cooling is offset by ongoing heating from near the Sun to 1 AU, matching states the authors previously derived close to the Sun. The paper also reports that the fast ejecta shows Kolmogorov-like magnetic turbulence while the slow ejecta has less developed turbulence, and that intermittent current-sheet-like structures cluster in the sheath and trailing regions and correlate with local heating. If right, heliospheric CME models must include sustained internal heating rather than assuming adiabatic expansion.","feed_headline":"Two CMEs arrive at Earth still being heated, not cooling","feed_subtitle":"Their magnetic ejecta are near-isothermal at 1 AU, matching their state near the Sun.","key_machinery":"The load-bearing object is the effective polytropic index, a temperature-weighted blend of the electron and proton polytropic indices, $\\Gamma_{\\rm eff}=(\\Gamma_e T_e+\\Gamma_p T_p)/(T_e+T_p)$, derived from the polytropic relation $T n^{1-\\Gamma}=\\text{constant}$. Each component index is obtained by linear fits of $\\log T$ versus $\\log n$ over moving six-point windows, filtered to retain only high-correlation fits, which lets the analysis map how heating varies across the sheath, five subdivisions of the magnetic ejecta, and the surrounding solar wind. The turbulence argument runs on the trace magnetic power spectrum $P_{\\rm tr}\\sim f^{\\alpha_B}$, the magnetic compressibility $C_B=P_t/P_{\\rm tr}$, and the partial variance of increments (PVI), which together connect spectral shape, Alfvénic character, and intermittent dissipation sites to the thermal state.","core_discovery":"On the paper's own terms, the discovery is that the magnetic ejecta of an ICME does not cool adiabatically as it travels. By fitting the polytropic relation $T n^{1-\\Gamma}=\\text{constant}$ in short moving windows of electron and proton data, the authors obtain localized polytropic indices for each substructure, then combine them through a temperature-weighted average, $\\Gamma_{\\rm eff}=(\\Gamma_e T_e+\\Gamma_p T_p)/(T_e+T_p)$. This yields $\\Gamma_{\\rm eff}=0.88$ (ICME1) and $0.76$ (ICME2), close to the isothermal value $\\Gamma=1$ and far from the adiabatic $\\Gamma=5/3$, implying substantial heating inside the ejecta at 1 AU despite its expansion. The same near-isothermal state was found near the Sun at roughly 15--20 $R_\\odot$ in the companion modeling, so the authors conclude that heating persists across heliospheric distances. In addition, inertial-range magnetic spectra are near Kolmogorov ($\\alpha_B\\approx -5/3$) in the fast ejecta but shallower in the central parts of the slow ejecta, and both ejecta have magnetic compressibility $C_B\\ll 1$, indicating mostly Alfvénic fluctuations.","pith_inferences":["The paper does not separate event speed from ambient environment; a larger sample of ICMEs with decoupled speed and wind-type combinations would test whether the fast/slow turbulence contrast is general.","If the near-isothermal state truly persists from roughly 15--20 $R_\\odot$ to 1 AU, the implied corollary is that heating approximately tracks expansion over the whole radial range; multi-spacecraft radial crossings of the same ejecta would provide a direct test.","Because most fitted windows are discarded by the correlation filter, the sensitivity of $\\Gamma_{\\rm eff}$ to the correlation-coefficient and p-value thresholds is not quantified; recomputing with all windows or with relaxed thresholds is the natural robustness check.","The PVI--$C_B$ association is qualitative; conditioning PVI statistics on $C_B$, plasma beta, or local temperature would test whether intermittent structures causally drive compressive fluctuations and dissipation."],"forward_implications":["Heliospheric CME transport models should not treat ejecta expansion as adiabatic; the near-isothermal indices imply a heating term that roughly balances expansion cooling out to 1 AU.","Fast and slow ejecta can arrive with different turbulence development states: the fast ejecta's magnetic ejecta is Kolmogorov-like, while the slow ejecta's central magnetic ejecta has a shallower inertial spectral index, indicating a less fully developed cascade.","Magnetic fluctuations inside both magnetic ejecta are predominantly Alfvénic ($C_B\\ll 1$), so models that treat ejecta fluctuations as compressive will mischaracterize their energy cascade.","Intermittent structures identified by PVI, likely current sheets and reconnection sites, cluster in sheaths and post-ICME regions and coincide with elevated temperatures and compressibility, making them plausible local dissipation sites.","The two events' ambient wind has similar thermal states but different turbulence properties, implying the surrounding medium shapes ejecta turbulence more than ejecta temperature."],"supporting_citations":[{"why":"Supplies the near-Sun polytropic states (near-isothermal around 15-20 R_sun) for the same two CMEs that the present 1 AU effective-index values are compared against.","marker":"Khuntia et al. (2023)"},{"why":"Provides the established range of ICME polytropic indices (1.15-1.33) at large heliocentric distances that the new values are contrasted with.","marker":"Liu et al. (2005, 2006a)"},{"why":"Justifies the interval-based approach by showing that the polytropic index varies across ICME sheath, magnetic ejecta, and ambient wind.","marker":"Dayeh & Livadiotis (2022)"},{"why":"Source of the interpretation that a negative electron temperature-density correlation in magnetic ejecta measures an electron polytropic index.","marker":"Osherovich et al. (1993); Sittler & Burlaga (1998)"},{"why":"Present the competing interpretation that the negative correlation arises from differing expansion histories and pressure balance, which the paper's localized-window analysis is designed to address.","marker":"Hammond et al. (1996); Gosling (1999)"},{"why":"Provides the sheath turbulence and PVI methodology and context that the present intermittency and spectral-slope results are compared with.","marker":"Kilpua et al. (2020)"},{"why":"Gives the recent result that ICME magnetic-flux-rope spectral indices remain close to -5/3 across heliospheric distances, the baseline for the observed alpha_B values.","marker":"Good et al. (2023)"}],"fun_headline_variants":["Both fast and slow CMEs stay heated at 1 AU","CMEs don't cool: near-isothermal ejecta at 1 AU","ICMEs maintain heat from Sun to Earth","CME ejecta near-isothermal at 1 AU, not cooling","Near-isothermal CME ejecta: no adiabatic cooling by 1 AU"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each six-point density-temperature window samples one coherent plasma parcel obeying a single polytropic relation, so that keeping only high-correlation fits (CC>0.8 and, as stated variously, p<0.05 vs p<0.5) does not bias the remaining $\\Gamma$ values; if windows blend parcels with different expansion histories, the near-isothermal effective indices could be an artifact.","fun_headline_variants_meta":{"raw":{"variants":["Both fast and slow CMEs stay heated at 1 AU","CMEs don't cool: near-isothermal ejecta at 1 AU","ICMEs maintain heat from Sun to Earth","CME ejecta near-isothermal at 1 AU, not cooling","Near-isothermal CME ejecta: no adiabatic cooling by 1 AU"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001081,"raw_usage":{"total_tokens":4626,"prompt_tokens":1152,"completion_tokens":3474,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":768,"completion_tokens_details":{"reasoning_tokens":3381}},"tokens_in":768,"tokens_out":3474,"duration_ms":21271,"temperature":1.0,"reasoning_tokens":3381,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:33:31.812669+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $\\Gamma_e$ and $\\Gamma_p$ without the correlation filter, or with windows aligned to plasma parcels identified by composition and velocity, and check whether $\\Gamma_{\\rm eff}$ stays near 1; if unfiltered or parcel-aligned values drift toward $\\Gamma=5/3$ or outside roughly 0.7--1.3, the sustained-heating claim fails. Resolving the stated $p<0.05$ versus $p<0.5$ threshold inconsistency and repeating the fit over all windows would settle the point directly.","supporting_citations":[{"cited_title":"A., Livadiotis G., 2022, @doi [ ] 10.3847/2041-8213/aca673 , https://ui.adsabs.harvard.edu/abs/2022ApJ...941L..26D 941, L26","cited_arxiv_id":null,"evidence_quote":"Justifies the interval-based approach by showing that the polytropic index varies across ICME sheath, magnetic ejecta, and ambient wind."},{"cited_title":"A., Farrugia C","cited_arxiv_id":null,"evidence_quote":"Source of the interpretation that a negative electron temperature-density correlation in magnetic ejecta measures an electron polytropic index."},{"cited_title":"M., Phillips J","cited_arxiv_id":null,"evidence_quote":"Present the competing interpretation that the negative correlation arises from differing expansion histories and pressure balance, which the paper's localized-window analysis is designed to address."}],"review_version":1}