{"id":"6eaf32a1-3d8f-41b7-ac5d-3ef5accb9b1a","arxiv_id":"1908.07866","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A spherical-harmonic inversion method with precomputed line-of-sight integrals reconstructs coronal electron density at 5 R⊙ efficiently, with reasonable accuracy on synthetic and real coronagraph data.","lead":"This paper presents a fast method for reconstructing the Sun's coronal electron density from coronagraph images using a spherical harmonic basis. It tests the method on synthetic data and applies it to two spacecraft, aiming to build long-term density maps for solar wind and space weather studies.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The assumed longitude/latitude-independent radial falloff f(r)=(r0/r)^2.2 in Eq. 2 is load-bearing; Sec. 3's synthetic test does not isolate it, and the Sec. 7 LASCO/COR2 38% discrepancy suggests it may bias real reconstructions.","rationale":"Reader identifies the radial-profile assumption; I agree it is the load-bearing point. The computational method is exact conditional on Eq. 3, and the efficiency gain is real and demonstrable. My criticism is not that the paper ignores the assumption, but that its validation does not isolate it: the Section 3 test uses a smooth harmonic target and known L, and the Section 4 test's density is also generated from smooth functions, so LOS averaging can hide a wrong global f(r). The 38% LASCO/COR2 discrepancy in Section 7 is the best available evidence that the assumption may matter, but it is confounded by calibration, binning, regularization, and viewing geometry. A targeted synthetic test with two falloffs would settle this. Until then, the reconstruction should be considered a useful efficient estimator whose bias under non-uniform radial falloff is unquantified; the CONDITIONAL verdict from the reader is appropriate and I do not propose changing it.","tokens_in":10332,"tokens_out":5498,"duration_ms":59572,"concrete_test":"Construct synthetic observations of a model corona with two distinct radial falloff laws, e.g., f_streamer(r)=(r0/r)^2.0 inside the streamer belt and f_hole(r)=(r0/r)^3.0 in coronal holes, with the same r0 surface density. Generate LOS brightness from two viewing geometries, then invert with the paper's method using Eq. 2 with α=2.2 and the exact Section 5 regularization. If the recovered r0 map has a mean absolute deviation above roughly 15%, or if the maps derived from the two viewing geometries differ by a LASCO/COR2-style ~38%, the angle-independent f(r) assumption is a material source of bias in real reconstructions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central efficiency claim is exact only when Eq. 3 holds with a single, angle-independent f(r): ρ(φ,θ,r)=ρ(φ,θ,r0) f(r). In that case the LOS sums Ai in Eq. 5 factor cleanly and Bk=Σ ci Ai. If the falloff actually varies by structure, e.g., streamers fall off more slowly than coronal holes, then ρ(φ,θ,r)=Σ ci Si(φ,θ) f(φ,θ,r), and Eq. 4 cannot be rewritten as Eq. 6 without cross-terms; the precomputed Ai are biased and the coefficients ci absorb the mismatch. The paper acknowledges this is a simplification (Section 2.1), but the validation is weak. In Section 3, the non-uniform radial profiles of Doyle and Gibson are used to make synthetic data, yet the target is a low-order smooth spherical-harmonic field and L is set to the known input order; a 3.8% density error and 0.5% brightness residual are reported. That test does not stress the angle-dependence of f, because a smooth surface density and long LOS averages can absorb a slowly varying radial-profile error. The real-data application (Section 7) is more telling: independent reconstructions from LASCO C2 and COR2 A differ by 38% mean absolute density, with 81% correlation. The paper frames this as a comparison, but what fraction of that discrepancy is caused by the single assumed f(r), regularization, or calibration is not quantified. Without this isolation, the claim that the method yields reliable maps over long time periods is not fully supported, even though the linear algebra and efficiency are sound.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a tomographic inversion method for white-light coronal observations at heights around 5 Rsun. The key idea is to represent the density on a spherical shell by spherical harmonics and, under the assumption of a radially structured corona with a uniform height profile f(r) = (r0/r)^alpha, to precompute line-of-sight integrations once per harmonic basis. The density reconstruction then reduces to a regularized linear least-squares fit, claimed to be very efficient. The method is tested on synthetic data from a simple spherical-harmonic density and from a more intricate, narrowly peaked distribution, with added noise and data gaps, and is applied to LASCO C2 and STEREO COR2 observations for March 2009. The paper concludes that the method is efficient, stable, and suitable for routine long-term coronal mapping.","tokens_in":10670,"tokens_out":3269,"duration_ms":36608,"significance":"If the efficiency and stability claims hold, the method could enable routine tomographic maps of the extended corona over long periods, which is valuable for solar-wind studies, space-weather applications, and linking coronal structure to in situ measurements. The manuscript has clear strengths: the linear algebra is transparent, the synthetic tests include a non-trivial target and explicitly test robustness to noise and large data gaps, and the real-data comparison is honest about the LASCO/COR2 discrepancy. However, the central assumption of a single, angle-independent radial falloff is only weakly validated, and the regularization parameter selection is heuristic. These issues do not invalidate the basic approach, but they need to be addressed before the long-term mapping claim is fully supported.","major_comments":[{"comment":"The factorization that permits line-of-sight sums to be computed once per harmonic is exact only when f(r) in Eq. (3) is independent of longitude and latitude. If the radial falloff varies with structure, for example in streamers versus coronal holes, Eq. (4) contains cross terms and Eq. (6) with precomputed A_i is only approximate; the reconstructed coefficients then absorb the mismatch. The paper acknowledges this simplification, but the validation is not isolating. In Section 3, the synthetic data are generated with the non-uniform Doyle and Gibson radial profiles while the reconstruction assumes Eq. (2), yet the target is a smooth L=11 spherical harmonic with known order, so a slowly varying radial-profile error can be absorbed by the surface-density coefficients and long line-of-sight averages. I request a quantitative test that specifically varies the radial falloff with angular position, e.g., a streamer-like slower falloff and a coronal-hole-like faster falloff, and reconstructs with the uniform Eq. (2), reporting the resulting bias in density and in the coefficient estimates.","section":"Section 2.1, Eqs. (4)-(6)"},{"comment":"The real-data comparison reports a 38% mean absolute difference between the LASCO C2 and COR2 A reconstructions with an 81% correlation. This is a large discrepancy, and the manuscript does not quantify how much of it is caused by the assumed uniform f(r), by regularization choices, by calibration differences between instruments, or by temporal changes over the two-week interval. Since the stated goal is reliable maps over long time periods, this decomposition is important. A concrete improvement would be to apply the same reconstruction pipeline to synthetic data viewed from two different spacecraft with known instrumental offsets, or to cross-calibrate the two reconstructions against a common standard and state the residual attributable to the radial-profile assumption.","section":"Section 7, Figs. 12-13"},{"comment":"The selection of the regularization parameter lambda and the minimum density threshold rho' is defined by a heuristic: the 'optimal point' is halfway between the centroid of the low-chi region and the farthest boundary point, with the region itself set by the 15% percentile of chi. This is said to be based on 'tests using several different density distributions,' but no sensitivity analysis is shown. Because the final density maps depend on these two parameters, the manuscript should either justify the heuristic with a reproducible criterion (for example, an L-curve or generalized cross-validation) or demonstrate that reconstructions are insensitive to the exact choice within the flat region of Fig. 8. Without this, the regularization step is not fully specified.","section":"Section 5, Eq. (14) and Fig. 8"}],"minor_comments":[{"comment":"The appendix restarts equation numbering at (1), which creates ambiguity when the text refers to 'equation 5 of section 2.1' (main-text Eq. 5) while the appendix also has its own Eq. (5). Renumbering the appendix as (A1), (A2), etc. would remove this confusion.","section":"Appendix, equations"},{"comment":"The description of the white contour as 'the 15% percentile minimum value of chi' is ambiguous: it is not clear whether the shaded region contains values below the 15th percentile of chi or above it. Please clarify the definition in the text or caption.","section":"Section 5, Fig. 8 caption and text"},{"comment":"The phrase 'the A^T A covariance matrix is pre-computed' refers to the regularized normal equations; the word 'covariance' is nonstandard here and could be replaced by 'normal-equations matrix' for clarity.","section":"Section 7, line after Eq. (12)"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and the efficiency idea is attractive. My main concern is that the load-bearing uniformity assumption of f(r) is not sufficiently stress-tested: the synthetic test in Section 3 uses a smooth target with known harmonic order, and the real-data comparison in Section 7 leaves a 38% discrepancy largely unexplained. These are fixable with additional experiments and sensitivity analysis, not fundamental flaws. I would encourage a revision that adds a targeted synthetic test with angle-dependent radial profiles and a more quantitative decomposition of the LASCO/COR2 difference."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, here's my read of Morgan's spherical harmonic tomography paper. The genuinely new thing is the precomputation trick: assuming a radial corona above the height of interest, each spherical harmonic's line-of-sight integral is computed once, so the inversion becomes a small linear least squares problem. That is real and useful. The synthetic tests are thoughtful—they create data using a non-uniform radial falloff (Doyle/Gibson) while reconstructing with a uniform falloff, and the method copes. The robustness to datagaps is worth emphasizing; losing a third of the observing period is a serious stress test, and the reconstruction degrades gracefully.\n\nThe soft spots are not fatal, but they are real. The regularization scheme has two free parameters (λ and a density floor), and the 'optimal' choice is a rule of thumb validated on a handful of synthetic cases. That is acceptable for a method paper, but it means the real-data maps should be taken as demonstrations, not calibrated products. The larger caveat is the assumed radial falloff f(r)=(r0/r)^2.2, uniform over all longitudes and latitudes. If streamers fall off more slowly than coronal holes, the clean separation in Eq. 6 breaks down and the coefficients will absorb the mismatch. The stress-test note worries that the synthetic tests don't isolate this, and it is right that they don't. The 38% mean difference between the LASCO and COR2 reconstructions sits right on top of that worry, though calibration and instrument differences probably contribute too. The paper doesn't quantify how much of that discrepancy comes from the assumed f(r), from regularization, or from calibration.\n\nOne thing the test does support clearly is the efficiency and stability of the linear algebra. Solving in minutes on a desktop, with stable results under noise and datagaps, is a genuine advance for routine long-term mapping. Citation practice looks fine: the prior tomography methods (Frazin, Morgan) are cited, and the novelty claim about the precomputation is consistent with that prior art.\n\nWho should read it: anyone building coronal density maps as inner boundary conditions for solar wind models, or wanting a fast tomographic method for large datasets. It deserves a serious referee. My recommendation is to send it to review, and to ask the author to (1) quantify the sensitivity to the radial falloff assumption, ideally with a synthetic test where the falloff varies by latitude, and (2) be explicit in the real-data section that the 38% cross-spacecraft difference is an upper bound on the method's current accuracy until calibration and f(r) effects are separated. With those changes it would be a solid contribution.","headline":"A genuinely new and efficient spherical-harmonic inversion trick for coronal tomography, with a heuristic regularization and an unresolved radial-falloff assumption that the real-data cross-spacecraft discrepancy (38%) keeps alive; worth a serious referee.","tokens_in":11183,"tokens_out":1687,"would_cite":true,"duration_ms":17177,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that coronal electron density can be reconstructed from coronagraph images by expanding the density in spherical harmonics on a shell at about 5 solar radii, assuming a radial corona above that height, and solving a…","keywords":["coronal electron density","spherical harmonics","coronal tomography","coronagraph inversion","LASCO C2","STEREO COR2","regularized least squares","radial density profile"],"falsifier":"Generate synthetic observations from a model in which the radial falloff differs between streamers and coronal holes, then reconstruct using the paper's fixed profile with $\\alpha = 2.2$; if the streamer-belt density error exceeds the roughly 12% seen in the paper's own synthetic tests, the uniform-profile assumption fails. An independent check is to compare the reconstructed density at 5 solar radii with in situ density measurements from a spacecraft crossing that height.","tokens_in":10119,"feed_emoji":"☀️","tokens_out":9662,"duration_ms":80273,"temperature":0.7,"pith_summary":"This paper claims that coronal electron density can be reconstructed from coronagraph images by expanding the density in spherical harmonics on a shell at about 5 solar radii, assuming a radial corona above that height, and solving a precomputed regularized least-squares system. Under that assumption, the costly line-of-sight integral of each harmonic is computed once at initialization, so the inversion reduces to a fast linear fit for the harmonic coefficients. Synthetic tests show the reconstruction reproduces a smooth target to within 3.8% mean absolute fractional deviation and a sharply ridged target to within 12.3% after regularization, while tolerating 5% noise and data gaps totaling four days. Applied to two coronagraphs observing the same period in 2009, the method yields density maps in minutes that are smoother and about half as dense as an earlier reconstruction, with a 38% difference between the two instruments.","feed_headline":"Precomputed line-of-sight sums make coronal density mapping fast","feed_subtitle":"A single radial-falloff assumption lets a two-week reconstruction run in minutes, robust to noise and data gaps.","key_machinery":"The central object is the precomputed line-of-sight integral of each spherical harmonic, $A_i = \\sum_j g_j f(r_j) S_{ij}$, evaluated once for a fixed radial falloff $f(r) = (r_0/r)^\\alpha$ with $\\alpha = 2.2$. Because the corona is assumed radial above the height of interest, $A_i$ does not depend on the unknown density, so the brightness becomes a linear system $b = A c$. The coefficients $c$ are then found by noise-weighted regularized least squares, with a diagonal penalty $w_i = (l_i + |m_i|)/\\sum_i(l_i + |m_i|)$ that increasingly damps high-order harmonics, and a two-parameter scan over the smoothing factor $\\lambda$ and a minimum density threshold $\\rho'$ sets the final smoothness and non-negativity.","core_discovery":"The central claim is that the assumption of a radially structured corona above the height of interest decouples the tomography problem: the line-of-sight summation for each spherical harmonic basis function, $A_i = \\sum_j g_j f(r_j) S_{ij}$, depends only on geometry and the fixed radial profile $f(r) = (r_0/r)^\\alpha$ with $\\alpha = 2.2$, not on the unknown density. Once these $A_i$ are precomputed, the observed brightness is a linear combination $b = A c$, and the density is recovered by solving a noise-weighted, regularized least-squares problem whose diagonal penalty damps high-order harmonics and whose minimum-density threshold is selected by a two-parameter goodness-of-fit scan. In synthetic tests the reconstruction reaches 3.8% mean absolute fractional deviation for a harmonic-generated target and 12.3% for a narrowly ridged target after regularization, with no negative densities; with added noise and multi-day gaps the deviation stays near 14%. On real data from 2009/03/20 the method yields a streamer belt at 5.5 solar radii that is narrower and about half as dense as a previous reconstruction, with a 38% mean absolute fractional difference between the LASCO C2 and STEREO COR2 results.","pith_inferences":["If the uniform radial falloff is genuinely wrong in a spatially varying way, the 38% difference between the two spacecraft may partly reflect that assumption rather than instrument calibration; an iterative scheme that updates $f(r)$ per region or per harmonic order would be a direct test.","The same precomputation trick transfers to any basis that factorizes the line-of-sight integral, such as wavelet bases on the sphere, which the paper already suggests for the non-radial low corona.","In situ density measurements from a spacecraft crossing the 5-solar-radii shell would give an independent, quantitative check of the radial-profile assumption and the reconstructed values.","The reported downward revision of streamer density by roughly a factor of two relative to earlier work is an empirical claim that could be checked against simultaneous white-light eclipse measurements at the same height."],"forward_implications":["Routine long-term mapping becomes feasible: with integrations done once, a two-week reconstruction completes in minutes, enabling multi-year atlases of coronal density.","The method's tolerance of several-day data gaps and 5% noise means degraded images can be discarded without destroying the reconstruction, simplifying processing pipelines.","Regularization eliminates negative densities, a known problem in coronal tomography, at the cost of smoothness near the equator where gradients are sharpest.","A time-dependent extension with harmonic coefficients varying in time is identified as the next step, but is found challenging, especially for step changes associated with coronal mass ejections."],"supporting_citations":[{"why":"Supplies the data processing, calibration, and dynamic separation technique used to prepare the LASCO and COR2 observations.","marker":"Morgan (2015)"},{"why":"Provides the Thomson-scattering line-of-sight integration formulation used for the $g_j$ factors.","marker":"Quemerais & Lamy (2002)"},{"why":"Introduces the regularization approach for coronal tomography that the paper adapts with a harmonic-order-dependent penalty.","marker":"Frazin 2000"},{"why":"Gives the previous reconstruction at 5.5 solar radii used as the comparison baseline for the new density maps.","marker":"Frazin et al. (2010)"},{"why":"Sets the coronal-hole minimum density radial profile used in the simple synthetic test.","marker":"Doyle et al. (1999)"},{"why":"Sets the streamer maximum density radial profile used in the synthetic tests.","marker":"Gibson et al. (2003)"},{"why":"Demonstrates time-dependent regularized tomography, the extension identified as the next step.","marker":"Vibert et al. (2016)"}],"fun_headline_variants":["Spherical harmonics accelerate coronal electron density mapping","Precomputed basis sums make coronal density reconstruction fast","Spherical harmonic model cuts coronal density mapping to minutes","Decoupled spherical harmonics enable fast, noise-robust coronal density maps"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that above the height of interest the corona is radially structured everywhere and that a single, uniform radial density falloff $f(r) = (r_0/r)^\\alpha$ with $\\alpha = 2.2$ describes the decrease at every longitude and latitude; if the true falloff varies in space, the separation of harmonics in the precomputed integrals is only approximate and the recovered coefficients are biased.","fun_headline_variants_meta":{"raw":{"variants":["Spherical harmonics accelerate coronal electron density mapping","Precomputed basis sums make coronal density reconstruction fast","Spherical harmonic model cuts coronal density mapping to minutes","Decoupled spherical harmonics enable fast, noise-robust coronal density maps"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000609,"raw_usage":{"total_tokens":2868,"prompt_tokens":1012,"completion_tokens":1856,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":1787}},"tokens_in":628,"tokens_out":1856,"duration_ms":12785,"temperature":1.0,"reasoning_tokens":1787,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:53:54.993327+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Generate synthetic observations from a model in which the radial falloff differs between streamers and coronal holes, then reconstruct using the paper's fixed profile with $\\alpha = 2.2$; if the streamer-belt density error exceeds the roughly 12% seen in the paper's own synthetic tests, the uniform-profile assumption fails. An independent check is to compare the reconstructed density at 5 solar radii with in situ density measurements from a spacecraft crossing that height.","supporting_citations":[{"cited_title":"2015, ApJS, 219, 23","cited_arxiv_id":null,"evidence_quote":"Supplies the data processing, calibration, and dynamic separation technique used to prepare the LASCO and COR2 observations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the regularization approach for coronal tomography that the paper adapts with a harmonic-order-dependent penalty."},{"cited_title":"A., Lamy, P., Llebaria, A., & V´ asquez, A","cited_arxiv_id":null,"evidence_quote":"Gives the previous reconstruction at 5.5 solar radii used as the comparison baseline for the new density maps."},{"cited_title":"G., Teriaca, L., & Banerjee, D","cited_arxiv_id":null,"evidence_quote":"Sets the coronal-hole minimum density radial profile used in the simple synthetic test."},{"cited_title":"E., Foster, D","cited_arxiv_id":null,"evidence_quote":"Sets the streamer maximum density radial profile used in the synthetic tests."},{"cited_title":"A., & Wojak, J","cited_arxiv_id":null,"evidence_quote":"Demonstrates time-dependent regularized tomography, the extension identified as the next step."}],"review_version":1}