{"id":"4d882362-f97d-4efd-a797-4032aaa57be7","arxiv_id":"2607.28506","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":5,"one_line_summary":"A software framework converts PRT topographical maps into heliocentric 3D point clouds; for one 2021 limb CME, Metis, LASCO, and COR2 reconstructions agree with each other and with GCS geometry.","lead":"PRICM3 turns polarization-ratio CME maps into 3D point clouds so multi-spacecraft views can be compared in one frame. It is a practical visualization and validation layer for polarimetric coronagraph data, timed for missions like PUNCH and ASPIICS.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.5","headline":"Interactive ROI plus ad-hoc |z| cutoffs make the reported multi-view consistency and GCS overlaps selection-dependent rather than a pure recovery of PRT geometry.","rationale":"The reader correctly isolated the load-bearing assumption: that filtered PRT |z| points are fair geometric tracers. The manuscript itself supplies the evidence that this step is under-specified (interactive selector in §2.3, ad-hoc 2.1 R⊙ cut in §3/Fig. 4, 95 % CI applied post-hoc). No stronger internal inconsistency exists; the single-event, favorable-limb design and the acknowledged LOS-average limitation are already flagged by the authors and do not overturn the tooling contribution. Because the concern is identical to the reader’s weakest_assumption, the CONDITIONAL verdict (public code + reproducible selection rules or additional events) remains the appropriate gate. The concrete test above directly quantifies how much of the Table 1 headline numbers survive a non-interactive pipeline; that is the minimal check needed before the “remarkably well” language can be taken as robust.","tokens_in":14101,"tokens_out":651,"duration_ms":30863,"concrete_test":"Re-derive the five Table 1 overlap fractions and ΔLon/ΔLat values after replacing the interactive ROI + morphology cut with a single automated brightness-and-|z| mask applied uniformly to all five maps (e.g., tB > 3σ_pre-event and |z| inside the global 90 % CI). If any of the three high-overlap cases (Metis 15:49, COR2 16:08, Metis 16:19) drops below 50 % or the centroid offsets exceed the stated GCS uncertainties, the claimed multi-view/GCS consistency is filter-dependent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (abstract; §3; Table 1) is that independent PRT |z| maps, once lifted by PRICM3 into a common heliocentric frame, produce mutually consistent point clouds that lie inside the GCS shell (overlaps ~74–87 % for the better epochs, centroids within GCS direction errors). That claim rests on treating each pixel’s LOS-weighted average |z| as a faithful 3D sample after (i) interactive ROI masking and (ii) statistical + morphology filters (95 % CI on every map; extra |z| ≥ 2.1 R⊙ streamer cut for LASCO 14:58 UT, Fig. 4). Because PRT already returns only a center-of-mass proxy (explicitly noted via Gibson et al. 2026 and Mierla et al. 2009), these subjective steps can preferentially retain points that happen to fall near the GCS axis while discarding discrepant plasma. The early COR2 15:08 cloud (16 % overlap) already shows how sensitive the metric is to evolutionary stage and filtering; without a fully specified, non-interactive selection rule the high later-epoch overlaps cannot be cleanly attributed to the underlying PRT signal.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript introduces PRICM3, a Python framework that lifts Polarization Ratio Technique (PRT) topographical maps into heliocentric 3D point clouds so that multi-spacecraft PRT reconstructions can be visualized together and compared to Graduated Cylindrical Shell (GCS) geometry. As a proof-of-concept, the authors apply independent PRT analyses of the 28 October 2021 slow limb CME from Solar Orbiter/Metis, SOHO/LASCO C2, and STEREO-A/COR2, transform the maps into a common frame, and report mutually consistent late-epoch clouds whose centroids lie within adopted GCS direction uncertainties and whose maximum GCS-volume overlap fractions reach ~74–87% (Table 1; §3). The paper argues that 3D information already present in PRT maps can be recovered in an intuitive spatial form and positions the tool for Metis, Proba-3/ASPIICS, and PUNCH polarimetry.","tokens_in":14404,"tokens_out":1457,"duration_ms":46862,"significance":"If the demonstration holds under clearer selection control, PRICM3 is a practically useful contribution: it does not invent a new inversion, but it removes a real barrier to interpreting PRT products by placing independent single-viewpoint depth maps in one heliocentric frame with GCS meshes and simple consistency estimators. The multi-instrument application (Metis, LASCO, COR2) and the explicit quantitative checks (centroid Δlon/Δlat; overlap fractions with stated PRT and GCS uncertainties) are strengths relative to purely visual PRT papers. Timeliness for PUNCH and ASPIICS is genuine, and the discussion of large-elongation Thomson geometry (§4) correctly flags where the framework must be extended. Code-release intent (GitHub upon acceptance) would further raise the work’s value if the selection pipeline is made reproducible.","major_comments":[{"comment":"§2.3–§3 and Fig. 4: The quantitative claims in Table 1 (centroid offsets and max. GCS overlap) rest on interactive ROI selection plus free filtering choices (brightness thresholds before Rm; 95% CI on |z| for every map; an extra morphology cutoff |z|≥2.1 R⊙ for LASCO 14:58 UT). Because PRT |z| is already a LOS-weighted center-of-mass proxy (acknowledged via Mierla et al. 2009 and Gibson et al. 2026), these steps can preferentially retain points near the GCS envelope. The early COR2-A 15:08 UT row (16.2% overlap) already shows strong sensitivity to evolutionary stage and constraints. For the central multi-view/GCS-consistency claim to be attributable to the PRT signal rather than selection, please add a brief sensitivity analysis: report how Δlon/Δlat and overlap change under documented alternate ROIs and under CI cuts (e.g. 90/95/99%) and with/without the morphology cutoff, or replace ad","section":"§2.3–§3, Fig. 4, Table 1"},{"comment":"§3 and Table 1: The overlap estimator is interpreted as geometrical agreement with the GCS shell, yet the text correctly states that PRT cannot recover CME depth extent and only samples an average LOS location. High enclosed fractions are therefore a weaker test than they appear (points can sit inside a thick shell without tracing the flux-rope surface). Please reframe the metric explicitly as a consistency check against an independent large-scale envelope—not volumetric recovery—and, if feasible, add a simple null comparison (e.g. overlap of the same clouds with a moderately mispointed or resized GCS shell) so the reported 74–87% figures can be judged against chance enclosure under your uncertainty model.","section":"§3, Table 1"},{"comment":"Abstract and §4: Language such as “prove that the 3D information… can be successfully recovered” and “aligned remarkably well” overstates a single favorable limb event with mixed epoch quality (LASCO/early-COR2 vs later Metis/COR2). Please temper the abstract and conclusions to “demonstrate for this event” / “support recovery under these conditions,” and state clearly that generalization to halo/complex LOS events remains untested.","section":"Abstract, §4"}],"minor_comments":[{"comment":"§2.2: Front/back ambiguity is resolved from EUV source location; a one-sentence note on residual ambiguity risk for non-limb or multi-source events would help readers applying PRICM3 more broadly.","section":"§2.2"},{"comment":"§3: GCS apex heights are linearly interpolated in time under self-similar expansion between clear-filter/double-exposure times and polarized times. State the time offsets and fitted speed (or apex vs time) so the interpolation uncertainty can be assessed alongside the adopted ±5% height error.","section":"§3"},{"comment":"Fig. 5 caption: Only a random subset of points is shown for visualization while analyses use full clouds—good—but please state the approximate fraction retained in the figure so readers do not misread apparent density.","section":"Fig. 5"},{"comment":"Fig. 6: Axes are intentionally wide; a zoomed inset around the cluster of centroids would make the “within uncertainty” statement easier to verify by eye.","section":"Fig. 6"},{"comment":"§2.3 fn. 16: Repository “upon acceptance” is fine; please commit to releasing the exact ROI masks/filter parameters used for Table 1 with the paper’s data products.","section":"§2.3"},{"comment":"Minor text: “F abi” spacing in the author list; ensure consistent pB/tB/uB italicization; “De Leo et al., 2026b, in preparation” should be cited uniformly when used to justify event selection.","section":"Front matter / §2.1"}],"recommendation":"minor_revision","confidential_remarks":"Solid methods/proof-of-concept paper appropriate for a solar-physics journal. I do not see a load-bearing physical error; the main risk is over-claiming quantitative multi-view agreement without controlling interactive selection. Sensitivity tests and toned-down abstract language should be sufficient—no need for additional events unless the authors want to strengthen generality. Fit and novelty are adequate: the contribution is the integration/visualization/comparison framework, not a new scattering inversion."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"Punchline: this is a methods/tooling paper, not a new Thomson-scattering inversion. PRICM3 takes standard PRT |z| maps, turns them into heliocentric point clouds, overlays multi-spacecraft views, and scores them against GCS with simple centroid and overlap metrics. That is genuinely useful for Metis/PUNCH/ASPIICS work and was missing as an operational layer.\n\nWhat is new is the common-frame point-cloud workflow plus the quantitative consistency checks (Table 1: late-epoch overlaps ~74–87%, centroids inside the stated GCS direction errors). The 28 Oct 2021 limb event is well chosen—near-POS, three instruments, independent PRT runs that mostly land in the same volume. They are honest that PRT only returns a LOS-weighted average (Mierla, Gibson et al.), that early COR2 is weak because the CME is barely above the occulter, and that density is lost in the maps. Citations and the GCS comparison look standard and fair; circularity is low because PRT and GCS are independent techniques.\n\nSoft spots, in proportion: single-event proof-of-concept; interactive ROI plus 95% CI and the ad-hoc |z|≥2.1 R⊙ streamer cut (Fig. 4) mean the high overlaps are partly selection-dependent, not a pure automatic recovery of geometry. Code is promised on acceptance, not shipped yet. None of that breaks the central claim that the 3D information already in PRT maps can be visualized and cross-checked this way; it does mean referees should ask for fully specified selection rules and more events.\n\nWho it is for: people doing polarimetric coronagraph CME geometry, especially multi-mission campaigns. Not for readers hunting a new physical model of CMEs. Math is elementary geometry plus established PRT; data handling is careful. I would send it to peer review. Engage if you care about multi-view PRT practice; skim if you only want inversion physics.","headline":"Useful multi-mission packaging of existing PRT maps into heliocentric point clouds; solid single-event demo, not a new inversion, with transparent but selection-dependent filtering.","tokens_in":15137,"tokens_out":505,"would_cite":true,"duration_ms":19946,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Polarization-ratio maps of CMEs can be turned into explicit 3D point clouds that match across spacecraft and line up with standard flux-rope geometry.","keywords":["solar physics","coronal mass ejections","polarimetry","Thomson scattering","3D reconstruction","coronagraphs","polarization ratio technique"],"falsifier":"Apply the same PRICM3 pipeline without morphology-driven cutoffs to another multi-viewpoint limb CME with dense polarized cadences; if independent point-cloud centroids systematically leave the GCS direction uncertainty box or overlap fractions stay low even at later, well-resolved epochs, the claim that PRT maps recover a consistent CME volume fails.","tokens_in":14955,"feed_emoji":"☀️","tokens_out":949,"duration_ms":26112,"temperature":0.7,"pith_summary":"White-light coronagraphs only see two-dimensional projections of coronal mass ejections, even though the eruptions are three-dimensional. The Polarization Ratio Technique already encodes a proxy for how far scattering plasma sits from the plane of the sky, but that information is usually left as color-coded 2D topographical maps that are hard to compare across viewpoints or against geometric models. This paper introduces PRICM3, a framework that converts those maps into filtered 3D point clouds in a shared heliocentric frame so independent reconstructions can be visualized together and checked against a Graduated Cylindrical Shell model. On a slow limb CME from 28 October 2021, reconstructions from three coronagraphs converge on a consistent volume whose centroids stay inside the model’s direction uncertainties and whose better-constrained epochs put most points inside the model shell. The result is meant to show that the depth already latent in polarimetric data can be recovered as usable 3D geometry for current and upcoming polarimetric missions.","feed_headline":"CME polarization maps become matching 3D point clouds","feed_subtitle":"Three coronagraphs recover one volume that sits inside a standard flux-rope shell","key_machinery":"PRICM3 (Polarization Ratio Integrated CME Mapper in 3D): a workflow that takes PRT topographical |z| maps, interactively selects the CME region, attaches each pixel’s plane-of-sky position to its line-of-sight distance, transforms the resulting points into a shared heliocentric frame, and overlays them with observing geometry and a GCS mesh for multi-viewpoint comparison.","core_discovery":"Independent Polarization Ratio Technique reconstructions of the 28 October 2021 limb CME, once converted by PRICM3 into a common heliocentric point-cloud frame, yield mutually consistent three-dimensional localizations that align with the corresponding Graduated Cylindrical Shell geometry, with cloud centroids inside the adopted direction uncertainties and volumetric overlap fractions reaching roughly 74–87 percent for the better-constrained epochs.","pith_inferences":["If the filtering step remains partly interactive, multi-event catalogs will need automated, reproducible ROI and outlier rules before PRICM3-style clouds can feed space-weather ensembles without human tuning.","Centroid and overlap scores could become a standard cross-check whenever GCS fits and single-view PRT disagree on propagation direction.","Extending the method to halo or backside events will stress the front/back ambiguity resolution that this limb case settled with EUV source location."],"forward_implications":["PRT topographical maps from different coronagraphs can be placed in one heliocentric frame and judged by shared centroid-offset and volume-overlap metrics.","Polarimetric sequences from Metis and similar instruments become routinely usable as 3D geometric products, not only as 2D depth-coded images.","Inner-corona joint views (for example high-resolution polarimetric pairs at similar heights) can be checked for fine CME substructure locations against a common shell.","Wide-field polarimetry out to large elongations can reuse the same mapper once the Thomson-scattering geometry is reformulated for non-parallel lines of sight and a point-like Sun."],"fun_headline_variants":["PRT maps convert to matching 3D CME point clouds","Three coronagraphs rebuild one CME volume inside GCS shell","PRICM3 turns polarization ratios into aligned 3D clouds","Limb CME localizations converge in shared heliocentric frame","Polarimetric CME maps recover consistent 3D geometry"],"cache_read_input_tokens":128,"weakest_assumption_plain":"Each pixel’s polarization-ratio depth can be treated as one fair three-dimensional sample of the CME after interactive region selection and ad hoc filtering, rather than a selection-dependent average that mixes the eruption with other structures along the line of sight.","fun_headline_variants_meta":{"raw":{"variants":["PRT maps convert to matching 3D CME point clouds","Three coronagraphs rebuild one CME volume inside GCS shell","PRICM3 turns polarization ratios into aligned 3D clouds","Limb CME localizations converge in shared heliocentric frame","Polarimetric CME maps recover consistent 3D geometry"]},"model":"grok-4.5","effort":"low","cost_usd":0.003736,"raw_usage":{"total_tokens":1199,"prompt_tokens":810,"num_sources_used":0,"completion_tokens":83,"cost_in_usd_ticks":37364000,"prompt_tokens_details":{"text_tokens":810,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":306,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":810,"tokens_out":83,"duration_ms":5859,"temperature":1.0,"reasoning_tokens":306,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T05:11:16.537925+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Apply the same PRICM3 pipeline without morphology-driven cutoffs to another multi-viewpoint limb CME with dense polarized cadences; if independent point-cloud centroids systematically leave the GCS direction uncertainty box or overlap fractions stay low even at later, well-resolved epochs, the claim that PRT maps recover a consistent CME volume fails.","supporting_citations":[],"review_version":1}