{"id":"f2d7b8b8-5ae9-4b46-a792-1d165fa958e2","arxiv_id":"2411.17108","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A boulder on comet 67P moved ~140 m within 14 hours on 2015 October 3, and thermal modeling shows the boulder's southern side was much hotter, possibly enabling self-propulsion.","lead":"Using Rosetta OSIRIS images, this paper narrows the time of a ~30-meter boulder's 140-meter shift on comet 67P to a 14-hour window on October 3, 2015, and finds night-time mini-outbursts nearby. Thermal modeling shows a strong temperature split between the boulder's sunlit and shadowed sides, suggesting its own sublimation could have helped move it.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 14-hour timing window rests on identifying a ~30-m boulder in ~20 m/pix images via synthetic rendering from a shape model that itself contains the boulder at both locations; an unquantified registration or shape-model error at the 09:26 'before' image would erase the constraint.","rationale":"The reader's weakest assumption is exactly the shape-model fidelity and synthetic-image matching for the 14-hour window, and I agree that this is the most load-bearing concern for the paper's central claim. The timing constraint is the paper's principal advance over the previous months-long window, and it rests on a 1.5-pixel boulder identification in ~20 m/pix images. The SHAP7 model's dual representation of the boulder makes the agreement between real and synthetic images partly self-fulfilling, and the paper provides no quantitative registration metric or pointing-error analysis. This is not an internal inconsistency or an ad hominem issue; it is a specific, addressable methodological gap. The proposed self-propulsion mechanism (Scenario C) is also quantitatively weak, since the paper's own estimate gives an acceleration an order of magnitude below surface gravity, but that weakness affects the interpretation rather than the timing claim itself. The paper is honest about the shape-model limitation in Sec. 3.3, and the underlying data are public, so a focused re-analysis could either confirm or refute the 14-hour window. For these reasons the appropriate verdict remains CONDITIONAL, and my read does not change the reader's verdict.","tokens_in":15,"tokens_out":4648,"duration_ms":108757,"concrete_test":"Reproject the SHAP7 model into the 2015-10-03 09:26 NAC image using archived SPICE pointing, and render two versions with identical illumination and camera geometry: (i) boulder present at its original location, and (ii) boulder removed (masked) with all other topography unchanged. Compute a normalized cross-correlation or chi-squared residual over the region containing the observed shadow for both renders; also perturb the camera boresight within the tabulated OSIRIS pointing uncertainty and measure the shift of the predicted shadow. If the boulder-absent render matches the observation as well as the boulder-present render, or if the predicted shadow shifts by more than ~20 m (about one pixel) under pointing perturbations, the 09:26 'before' identification is not secure.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central new claim is the ~14-hour migration window on 2015 October 3 (Sec. 3.1). This claim depends entirely on two detections: the boulder 'at its original location' in the 09:26 NAC image and its 'faint shadow' at the new position in the 22:49 NAC image. Both images have a surface resolution of ~19–20 m/pix (Table 1), so a 30-m boulder subtends roughly 1.5 pixels. The identifications are made by blinking real images against synthetic renderings of the SHAP7 shape model, whose ~4-m horizontal resolution (Preusker et al. 2017) is far better than the images. However, SHAP7 was derived from images acquired between August 2014 and February 2016, so it explicitly contains the boulder at both its original and final locations (Sec. 3, Fig. 1a). Rendering that model at 09:26 necessarily places a boulder-shaped feature at the original location, making agreement between synthetic and real images partly circular. The 'lengthened shadow' used for the 09:26 detection is the only anchor for the pre-migration state; if that shadow is actually cast by adjacent topography, or if the boulder's true size/shape differs from the model (as the authors themselves note for the new location in Sec. 3.3: 'the identification of the boulder at its new location is subject to the accuracy of the shape model'), then the boulder may already have been gone by 09:26 and the true window would degrade to the Sep 26 – Oct 3 interval or worse. No quantitative match metric, residual map, or pointing-error propagation is provided for either detection. The 22:49 'faint shadow' is even lower signal-to-noise. Thus the 14-hour constraint is load-bearing on an unquantified image-to-model registration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reanalyzes Rosetta/OSIRIS imaging to constrain the timing of a ~30 m boulder's ~140 m displacement in the Khonsu region of comet 67P. Using the SHAP7 stereo-photogrammetric shape model and synthetic-image matching, it claims the boulder was still at its original position at 09:26 UTC on 2015 October 3 and had arrived at its new position by 22:49 UTC, yielding a migration window of within 14 hours. The paper also reports night-time mini-outbursts near the new location, documents morphological changes in the surrounding area, and applies a 1-D thermophysical model (Hu & Shi 2021) to local facets. The model shows a pronounced insolation and temperature dichotomy between the boulder's southern and northern sides, and the authors propose that the boulder's own sublimation-driven activity may have contributed to or triggered the migration.","tokens_in":15091,"tokens_out":4658,"duration_ms":43115,"significance":"If the 14-hour timing claim withstands scrutiny, it would be the best-constrained large-boulder migration event observed on a cometary nucleus, demonstrating a sudden, burst-like displacement rather than gradual creep, and enabling a direct search for coincident activity. The paper exploits unique Rosetta data, provides a clear observational catalog in Table 1, and uses a published thermophysical model without tuning its parameters to reproduce the migration, which is a methodological strength. However, the central timing constraint rests on identifying a ~2-pixel boulder in ~20 m/pix images, and the synthetic comparisons are partly circular because the shape model contains the boulder at both its pre- and post-migration locations. The proposed self-propulsion mechanism is also quantitatively weak by the authors' own estimate. These issues need to be addressed before the central claims can be considered robust.","major_comments":[{"comment":"The 14-hour migration window is based on visually matching ~20 m/pix NAC images against synthetic renderings of SHAP7, a shape model that contains the boulder at both its pre- and post-migration positions because it was derived from images spanning the migration (Sec. 3, Fig. 1a). Consequently, the synthetic image used for the 09:26 detection contains a boulder at the original location by construction, so agreement between the real and synthetic images does not independently establish the boulder's presence there at 09:26. The paper provides no quantitative match metric, residual map, or pointing/registration uncertainty for this key detection. If the 'lengthened shadow' at 09:26 is not uniquely attributable to the boulder, the constraint degrades to the Sep 26 - Oct 3 interval. Please add a quantitative registration analysis, for example by measuring residuals against fixed landmarks and by testing whether masking the boulder in the shape model removes the observed signal at the original location.","section":"Sec. 3.1, Fig. 4; Sec. 3, Fig. 1a"},{"comment":"The paper states that the source of the 19:30 mini-outburst is 'unclear whether the outburst is from the unilluminated boulder itself or its surroundings,' and that 'the identification of the boulder at its new location is subject to the accuracy of the shape model.' These admissions are important: the coincident dust activity on October 3 is not established to be directly associated with the migration, and the boulder's arrival at the new location by 22:49 is not as secure as the text in Sec. 3.1 implies. The timeline in Fig. 7 should visibly distinguish confirmed boulder locations from activity whose source is ambiguous, and the 'coincident activity' interpretation should be softened accordingly.","section":"Sec. 3.3, Fig. 6a"},{"comment":"The proposed self-propulsion scenario (Scenario C) is not quantitatively supported by the paper's own numbers: a mean water-vapor flux of ~3.2 x 10^-5 kg m^-2 s^-1 at facet S yields an acceleration of ~2 x 10^-5 m s^-2, one order of magnitude smaller than the local gravitational acceleration of ~2 x 10^-4 m s^-2. The paper acknowledges this shortfall but nonetheless presents the mechanism as viable. To make Scenario C convincing, the authors need to show how the net force could exceed gravity, for example by invoking supervolatile (CO2) sublimation at depth or by computing the reduced force needed to initiate rolling on a ~25-degree slope. As written, Scenario C is a qualitative suggestion rather than a demonstrated mechanism.","section":"Sec. 4; Sec. 5.3"}],"minor_comments":[{"comment":"The symbol f0 is used for both the dust volume-filling fraction (0.2) and the initial ice volume fraction (0.1); please use distinct symbols to avoid ambiguity in the parameter definitions.","section":"Table 2"},{"comment":"The statement that the boulder 'extends over only two pixels' is slightly inconsistent with a ~30 m boulder at ~19-20 m/pix, which corresponds to roughly 1.5 pixels; please phrase this as 'one to two pixels.'","section":"Sec. 3.1"},{"comment":"The model equations do not explicitly state the surface boundary condition that couples the sublimation flux to the surface energy balance; a pointer to Hu & Shi (2021) is given, but a short statement of the surface boundary condition would make the reported flux values in Fig. 10 reproducible by the reader.","section":"Eqs. (1)-(2) and Sec. 4"}],"recommendation":"major_revision","confidential_remarks":"The 14-hour constraint is a valuable observational result, but the identification evidence as presented is not sufficient to exclude a registration or shape-model artifact at the 09:26 image. I would require the quantitative masking/registration test described in Major Comment 1 before acceptance. The self-propulsion scenario may be better framed as speculation unless its acceleration shortfall is addressed; if it cannot be strengthened, the authors should present it only as a qualitative possibility. There is also a question of fit with the journal's scope, but the Rosetta data analysis and the new timing constraint make this a suitable contribution if the robustness issues are resolved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a useful piece of work that genuinely narrows the timing of the largest boulder displacement seen on 67P, but the marquee 14-hour window is softer than the abstract implies. It deserves peer review, but a referee should push on the identification methodology.\n\nWhat's actually new: the authors whittle the migration window from months to a single rotation of the comet (09:26 to 22:49 UTC on 2015 October 3). That is a real observational constraint, and it is not a fitted model output. They also turn up two night-time mini-outbursts near the boulder's new position—one during the migration window, one a day later—and they lay out the thermal state of the boulder's two sides with a published thermophysical code. The north-south insolation dichotomy is a credible, quantitative description of the local environment.\n\nThe soft spot is the load-bearing 14-hour window. The two bracketing images have roughly 19–20 m/pix resolution; a 30-m boulder occupies maybe 1.5 pixels. The authors identify the boulder by blinking real images against synthetic renderings of the SHAP7 shape model—the same model that contains the boulder at both its old and new locations. If the 09:26 'before' detection is actually a shadow from adjacent topography, or if the boulder's true size differs from the model, the boulder could already have been gone by 09:26, and the real window degrades toward the September 26 – October 3 interval. The paper gives no quantitative match metric, residual map, or pointing-error propagation to rule that out. The authors acknowledge the shape-model limitation for the post-migration identification (Sec. 3.3), but the same caveat applies to the pre-migration side.\n\nThat said, this is not fatal. The 14-hour window is consistent with a sudden event, and the Oct 4 image independently confirms the boulder at its new location. The self-propulsion scenario is explicitly underpowered—the acceleration estimate is an order of magnitude below surface gravity—and the paper says so, which is honest. The mechanism discussion is labeled as speculation.\n\nFor a cometary-surface specialist, this is worth reading and citing. For a general audience, the headline number should be taken with a grain of salt until the identification is quantitatively validated. I would send it to peer review with a request for a more rigorous image-to-model comparison.","headline":"Useful and honest paper that narrows a 67P boulder migration to a 14-hour window, but the marquee claim sits on a 1.5-pixel visual identification against a shape model that contains the boulder at both locations—so the number is plausible but not yet airtight.","tokens_in":15665,"tokens_out":3052,"would_cite":true,"duration_ms":26972,"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":"A ~30-meter boulder on comet 67P shifted ~140 meters within a 14-hour window on 2015 October 3, and the authors propose that uneven heating and outgassing of the boulder itself helped trigger the displacement.","keywords":["comet 67P","boulder migration","Khonsu region","Rosetta mission","OSIRIS images","thermophysical modeling","comet activity"],"falsifier":"If a reprojection of the October 3 09:26 or 22:49 images with a refined shape model or slightly different camera pointing placed the boulder's shadow on the opposite side of the migration, or if a higher-resolution image from the same day under better illumination showed the boulder absent from both locations, the 14-hour window would be falsified.","tokens_in":14526,"feed_emoji":"☄️","tokens_out":8664,"duration_ms":68869,"temperature":0.7,"pith_summary":"The paper revisits the largest boulder movement seen on comet 67P: a ~30-meter boulder in the Khonsu region that shifted ~140 meters between August and October 2015. By matching OSIRIS images to synthetic views generated from the SHAP7 shape model, the authors constrain the migration to a single 14-hour interval on 2015 October 3, showing the boulder moved suddenly rather than creeping. They also detect night-time mini-outbursts near the boulder's new location on the migration day and the next day. Thermophysical modeling reveals a sharp thermal dichotomy between the boulder's southern and northern sides, with the southern side receiving twice the insolation and reaching peak subsurface temperature at the time of movement. On this basis they propose a third triggering mechanism: the boulder's own uneven outgassing could have produced a net 'rocket force,' possibly acting alongside an outburst or seismic shaking.","feed_headline":"Comet boulder's 140-meter leap pinned to a 14-hour window","feed_subtitle":"Rosetta data show a 30-meter rock on comet 67P moved 140 meters in a day.","key_machinery":"The central tool is synthetic-image matching against the stereo-photogrammetric SHAP7 shape model of comet 67P (~4 m horizontal resolution, 1 million vertices). Because the shape model was built from images spanning the whole mission, it contains the migrating boulder at both its original and final locations, so synthetic views with realistic ray-traced illumination can be blinked against actual low-resolution OSIRIS images (~20 m/pixel) to tell which side of the migration the boulder is on. The second component is a 1-D facet-wise thermophysical model that conserves energy and volatile mass, used to reconstruct the diurnal and orbital temperature and sublimation-flux history of the boulder's southern and northern sides. Together these yield the 14-hour timing constraint and the thermal asymmetry that underpins the proposed 'rocket force' mechanism.","core_discovery":"The paper establishes that the ~30-m boulder in Khonsu migrated ~140 m on 2015 October 3 between 09:26 and 22:49 UTC, a window of under 14 hours, by blinking OSIRIS NAC images against synthetic renders of the SHAP7 shape model. This makes the displacement a sudden event rather than a gradual process. The same analysis finds a mini-outburst at 19:30 UTC on the migration day emanating from near the boulder's new location, and another mini-outburst two rotations later, plus a jet on October 9. Facet-wise thermophysical modeling shows the boulder's southern side had been illuminated almost continuously for over 150 rotations, received roughly twice the cumulative insolation of the northern side, and reached orbital-maximum temperatures near 225 K at 10 cm depth at the time of migration, while the northern side was in a cold 'winter' state. The authors therefore argue that asymmetric sublimation from the boulder's own interior could have exerted a net propulsion toward the north, complementing the earlier hypotheses of direct outburst, erosion, or seismic destabilization; the actual trigger may have been a combination.","pith_inferences":["If the migration truly happened in the afternoon of October 3, the peak in surface temperature on the boulder's southern side (~15:00 UTC) points to a thermal-stress or sublimation trigger; one could test this by checking whether other large boulder displacements on 67P cluster in local afternoon.","The proposed self-propulsion mechanism would be most convincing if the boulder's actual shape produces a net force direction roughly downslope; a numerical simulation using the SHAP7 facet geometry and the modeled outgassing could show whether the estimated acceleration is sufficient to overcome static friction on a ~25° slope.","The synthetic-image matching technique, treating a moving object as present at two locations in a single shape model, could be applied to other candidate surface changes on 67P (e.g., the other Khonsu boulder) to shrink their timing windows.","A direct search for fresh albedo or exposure changes at the original site using the highest-resolution post-migration images, or a re-analysis with a shape model that excludes the boulder's dual presence, would test whether the identification at the new location is robust."],"forward_implications":["The 14-hour constraint shows the boulder moved in a sudden, event-like displacement rather than through gradual creep, since the window is barely longer than one 12.4-hour rotation.","Night-time mini-outbursts occurred on the migration day and the following rotation near the boulder's new position, indicating a spatial and temporal link between local dust activity and the displacement.","Thermal modeling predicts that at migration time the boulder's southern side, 10 cm below the surface, was at its orbital maximum temperature (~225 K), a state favorable for sublimation of volatiles at depth.","If uneven outgassing contributed, large boulders on comets can act as self-propelled objects, a mechanism that has not been considered for tens-of-meters-scale blocks before.","The two Khonsu 'jumping boulders' may share a common triggering scenario, either ejection by an outburst or self-propulsion, though the other boulder's original position is unknown."],"supporting_citations":[{"why":"first reported the boulder's ~140 m displacement between August and October 2015 and proposed the outburst-lifting scenario.","marker":"El-Maarry et al. (2017)"},{"why":"provided the SHAP7 stereo-photogrammetric shape model that contains the boulder at both locations, enabling synthetic-image matching for the timing.","marker":"Preusker et al. (2017)"},{"why":"provides the camera pointing and positioning used in geometric calculations and synthetic views.","marker":"Acton et al. (2016)"},{"why":"narrowed activity to within two months after perihelion and documented the other Khonsu jumping boulder and outburst context.","marker":"Hasselmann et al. (2019)"},{"why":"supplies the 1-D thermophysical code and insolation calculation used for the thermal history modeling.","marker":"Hu et al. (2017a,b)"},{"why":"supplies the coupled energy-and-mass-conservation model that gives temperature and sublimation flux profiles.","marker":"Hu & Shi (2021)"},{"why":"provides the 'rocket force' equation used to estimate the acceleration from uneven outgassing on the boulder.","marker":"Agarwal et al. (2016)"}],"fun_headline_variants":["Sudden 140-m boulder jump on comet 67P linked to its own heat","How a comet boulder moved 140 m in a 14-hour burst","Comet 67P boulder's 14-hour 140-m sprint shown by Rosetta","Asymmetric warming likely drove comet boulder's 140-m move","Rosetta camera pins boulder's 140-m leap to 14 hours"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the SHAP7 shape model, with its ~4-meter horizontal resolution, faithfully represents the 30-meter boulder and its surroundings in both the pre- and post-migration states, so that the boulder's shadow in ~20 m/pixel images can be trusted to place it at one location or the other.","fun_headline_variants_meta":{"raw":{"variants":["Sudden 140-m boulder jump on comet 67P linked to its own heat","How a comet boulder moved 140 m in a 14-hour burst","Comet 67P boulder's 14-hour 140-m sprint shown by Rosetta","Asymmetric warming likely drove comet boulder's 140-m move","Rosetta camera pins boulder's 140-m leap to 14 hours"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00164,"raw_usage":{"total_tokens":6541,"prompt_tokens":991,"completion_tokens":5550,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":607,"completion_tokens_details":{"reasoning_tokens":5444}},"tokens_in":607,"tokens_out":5550,"duration_ms":35179,"temperature":1.0,"reasoning_tokens":5444,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:30:03.174834+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If a reprojection of the October 3 09:26 or 22:49 images with a refined shape model or slightly different camera pointing placed the boulder's shadow on the opposite side of the migration, or if a higher-resolution image from the same day under better illumination showed the boulder absent from both locations, the 14-hour window would be falsified.","supporting_citations":[{"cited_title":"2017, Astronomy & Astrophysics, 607, L1, 10.1051/0004-6361/201731798","cited_arxiv_id":null,"evidence_quote":"provided the SHAP7 stereo-photogrammetric shape model that contains the boulder at both locations, enabling synthetic-image matching for the timing."},{"cited_title":"2016, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 41, 357","cited_arxiv_id":null,"evidence_quote":"provides the camera pointing and positioning used in geometric calculations and synthetic views."},{"cited_title":"H., Barucci, M","cited_arxiv_id":null,"evidence_quote":"narrowed activity to within two months after perihelion and documented the other Khonsu jumping boulder and outburst context."},{"cited_title":"2021, Journal of Geophysical Research (Planets), 126, e06594, 10.1029/2020JE006594","cited_arxiv_id":null,"evidence_quote":"supplies the coupled energy-and-mass-conservation model that gives temperature and sublimation flux profiles."},{"cited_title":"F., Vincent , J","cited_arxiv_id":null,"evidence_quote":"provides the 'rocket force' equation used to estimate the acceleration from uneven outgassing on the boulder."}],"review_version":1}