{"id":"05ba1477-da02-4ff1-90c2-45c8c90c5851","arxiv_id":"2506.12854","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"high","formal_verification":"none","parameter_count":5,"one_line_summary":"Two unresolved far-infrared sources, detected twice within 24 hours and absent after six months in AKARI single-scan data, are presented as Planet Nine candidates awaiting follow-up.","lead":"A far-infrared search using AKARI single-scan detections has found two point sources that stay fixed within a day, vanish after six months, and fall inside the sky region where Planet Nine was predicted. If one is real, it would be the first thermal detection of a distant solar-system planet; if not, the search method still shows how single-scan far-infrared catalogs can be mined for moving objects.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The motion claim rests on six-month non-detections whose sensitivity is never quantified; absent candidates could be missed stationary sources.","rationale":"I read the paper as a transparent search that explicitly presents two candidates warranting follow-up rather than a confirmed detection. The reader's CONDITIONAL verdict is appropriate. The single most load-bearing weakness is the unquantified completeness of the six-month non-detections, exactly as the reader identified. This is load-bearing because the search strategy's core discriminator is 'stationary within 24 hours, absent after six months.' If a stationary source of the candidate flux would not have been detected in the later scans, then the absence criterion is empty and the candidates reduce to two-epoch detections of unknown nature. The paper gives no scan-count, sensitivity, or injection-recovery information for the later epochs at the two candidate positions. The transient and noise estimates in Section 5 do not cover this branch. A concrete injection-recovery test on the actual six-month-separated detection maps would settle whether the non-detections are meaningful. Because the paper's conclusion is already hedged as 'possible candidates' and explicitly calls for follow-up, this concern does not change the reader's CONDITIONAL verdict; it does mean the claim should not be strengthened without the sensitivity check and, ideally, a positive detection at the parallax-shifted position.","tokens_in":17396,"tokens_out":5045,"duration_ms":58724,"concrete_test":"For each final candidate, take the six-month-separated AKARI detection-probability maps at that position, inject a point source with the measured per-scan flux (0.61–1.62 Jy) at the candidate coordinates, run the same FIS source extraction with the >21 threshold, and record whether it is recovered. If either injected source is not recovered, the six-month non-detection cannot distinguish a moving Planet Nine from a stationary source missed because of limited coverage or sensitivity, and the two candidates should be reclassified as requiring a positive detection at the parallax-shifted position.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The inference that the two final candidates are moving objects, and hence possible Planet Nine, depends on the claim in Sections 3.6 and 3.7 that they are detected twice within 24 hours and are 'confirmed to have no appearances before six months or disappear after six months.' The paper never quantifies the sensitivity of the six-month-separated scans at the candidate positions. In particular, it does not state the number of later-epoch scans, the local noise or confusion level, or the detection-probability-map threshold that would have been required to recover a 0.5–1.6 Jy point source at those coordinates. Without this, a non-detection is equally consistent with (a) a Planet Nine that has moved out of the 32-arcsecond search radius, and (b) an unresolved stationary far-infrared source that was not covered, or was below the local detection threshold, in the later scans. The transient-rate calculation in Section 5 estimates contamination by astrophysical transients and noise fluctuations, but it does not include the missed-stationary-source branch. The candidates therefore currently lack the positive evidence of displacement that the search strategy is designed to provide.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a search for Planet Nine in the AKARI all-sky survey using the FIS Single Scan Detection List, which, unlike the standard FIS Bright Source Catalogue, does not require confirmation across multiple scans. The authors select sources in the Millholland and Laughlin (2017) sky region (30°<R.A.<50°, –20°<Dec.<20°), cross-match against nine optical and infrared catalogues, apply flux and background cuts (FLUX90/FERR90>5, BG90<0.2), require two detections within 24 hours and no detection after six months, and visually inspect detection probability maps to reject cosmic rays and edge artifacts. After this pipeline they identify two candidate sources with per-scan 90 µm fluxes between 0.5 and 1.6 Jy, at positions within the selected region, and argue that their fluxes are consistent with a 6–12 Earth-mass planet at 400–900 au with an effective temperature of 28–53 K. The paper estimates contamination from astrophysical transients and noise fluctuations, reports expected counts of about 0.014 transients and 0.029 noise overlaps, and proposes Subaru HSC follow-up observations.","tokens_in":17519,"tokens_out":10440,"duration_ms":96758,"significance":"If the two candidates are genuine moving objects, this would be the first far-infrared detection of a plausible Planet Nine and would provide strong motivation for targeted follow-up. The paper also makes a methodological contribution by exploiting the AKARI FISSSDL, which is not subject to the moving-source rejection of the standard catalogue. The claimed result is, however, not yet fully supported: the key evidence of motion—the six-month non-detection—is not quantified in terms of sensitivity, and the selection pipeline relies on several post hoc thresholds and subjective image inspection. The transient and noise contamination estimates are useful but do not cover all possible false-positive branches, so the significance of the two candidates remains uncertain pending a quantitative demonstration that the non-detections are meaningful.","major_comments":[{"comment":"The central claim that the two candidates are moving objects rests on the statement that they were 'confirmed to have no appearances before six months or disappear after six months' (Section 3.7). However, the paper never quantifies the sensitivity of the six-month-separated scans at the candidate positions. It does not report the number of later-epoch scans, the local detection-probability-map values, or the background and noise level at the candidate coordinates. Without this information, a non-detection could equally be explained by a stationary unresolved source that was not covered by, or fell below the detection threshold of, the later scans. The authors should provide, for each of the two candidates, the detection-probability-map value or an equivalent flux upper limit at the six-month epoch, and demonstrate that a source with the observed first-epoch flux would have been detected had it remained stationary.","section":"Sections 3.6–3.7, Table 2"},{"comment":"The contamination estimate in Section 5 accounts for astrophysical transients (CCSNe, SNe Ia, novae, TDEs) and for overlapping noise fluctuations, but it does not include the branch in which a stationary far-infrared source is detected in the first epoch and missed in the later epoch because of coverage gaps or sensitivity variations. Since the six-month non-detection is the only evidence that the candidates moved, the authors should either quantify how many stationary sources would be expected to pass the full selection (for example, by injecting artificial stationary sources into the FISSSDL and running the selection pipeline) or explicitly add this branch to the false-positive budget.","section":"Section 5"},{"comment":"The description of the candidate selection after the one-year check is ambiguous. The text states that only 8 out of 10 candidates have detection maps in the subsequent or preceding year, and then all 8 of those are removed, leaving two candidates. This implies that the final two candidates are among those without one-year detection maps. The authors should clarify, for each final candidate, which epochs were actually observed, whether the six-month-separated scans covered the candidate position, and how the 'no appearances before six months or disappear after six months' confirmation was performed if the one-year maps are absent. A table listing the scan epochs and the detection-probability-map values at the candidate positions for both the initial and the follow-up epochs would remove this ambiguity.","section":"Section 3.7"},{"comment":"Several selection thresholds (BG90 < 0.2, detection-probability-map value > 21, and the two-detections-within-24-hours requirement) are chosen after inspecting the data and are not accompanied by a robustness test. Because the final candidate list shrinks from 393 to 2 through these cuts, the authors should demonstrate that the two surviving candidates are not an artifact of the specific threshold choices—for example, by varying each threshold over a reasonable range and reporting how the candidate list changes, or by estimating the effective number of trials. This is particularly important given that the visual rejection of cosmic rays and edge detections is subjective.","section":"Sections 3.5 and 3.7"}],"minor_comments":[{"comment":"The header contains the placeholder-like text 'Cambridge Large T wo(2024)' and the running title should be checked for typographical errors before final submission.","section":"Title page"},{"comment":"The caption states 'The X-axis shows the number of sources in each bin' for the right panel; the number of sources is plotted on the y-axis, not the x-axis. In addition, the sentence 'The dark blue histogram shows ... with the same y-axis as the left panel' is unclear because the two panels have different y-axis quantities.","section":"Figure 2 caption"},{"comment":"The text refers to an 'arbitral unit' for the detection probability map; this should be 'arbitrary unit'. Also, the sentence describing the selection of 165 candidates could be reworded to clarify the difference between 'clearly detected in all scans within 24 hours' and 'confirmed to have no appearances before six months or disappear after six months'.","section":"Section 3.7"},{"comment":"The transient time-window factor is written as '(T – 1)/365'; if the requirement is two detections within one day, the probability of catching a transient of duration T in a one-year survey is not simply (T–1)/365. Please explain the derivation of this factor.","section":"Section 5"},{"comment":"The table lists per-scan fluxes for the two candidates but does not provide flux uncertainties. Given that the fluxes vary by factors of 2–3 between the two scans of each candidate, the authors should report the photometric errors to allow the reader to judge whether the variability is significant.","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents an interesting search and a useful dataset, but the central claim of moving-object detection is not yet quantitatively supported. The most important revision is to provide sensitivity limits for the six-month non-detections and a completeness/false-positive analysis that includes stationary sources missed in later epochs. I also note that the acknowledgment thanks an anonymous referee, which is unusual in a submitted manuscript and may need to be removed or revised before publication; this does not affect my substantive assessment."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe useful summary: this is a genuine new search — the first use of AKARI single-scan data to look for Planet Nine's thermal emission — and it produces two candidates that were not in prior catalogs. The two sources sit in the Millholland-Laughlin sky region, with 90 µm fluxes between 0.5 and 1.6 Jy, and the authors are appropriately cautious: they call them 'possible' candidates and push for follow-up. That is the honest frame, and it should be respected.\n\nWhat is actually new: the FISSSDL catalog, built from single-scan AKARI detections, relaxing the confirmation requirement that would exclude moving objects. That is a real, reusable product. The idea of searching for six-month parallax displacement in far-infrared single-scan data is also new. The cross-match with nine optical/IR catalogs is careful, and the transient-rate calculation for supernovae, novae, and TDEs is a reasonable first-order contamination estimate.\n\nThe soft spots are where the reader's conditional verdict lands. The biggest is the motion claim. The paper requires the candidates to have no detection six months before or after, and it excludes monthly-confirmed sources as non-moving. But it never quantifies whether a stationary source at the candidates' flux would actually have been recovered in the later scans. The detection probability maps are inspected, but no local noise, coverage, or completeness threshold at the candidate positions is reported. A non-detection could equally mean a source that was simply below the local detection limit in that epoch, which would make the 'moved away' inference unsupported. This is load-bearing: it is the only thing separating a moving Planet Nine candidate from a stationary far-infrared source the search failed to re-detect.\n\nThe secondary issue is the number of post hoc thresholds — BG90 < 0.2, FLUX90/FERR90 > 5, detection probability > 21 — chosen after inspecting the data, and the subjective rejection of cosmic rays and edge detections. The two final candidates also have mutually inconsistent fluxes (0.61 vs 1.62 Jy; 1.27 vs 0.51 Jy) with no uncertainties. None of this is fatal on its own; the paper does not claim a confirmed detection. But it does mean the evidence supports 'targets worth follow-up,' not 'probable detections.'\n\nWho gets value: anyone working on Planet Nine searches, far-infrared source extraction, or moving-object identification in single-scan survey data. The paper deserves a serious referee. The novel catalog and search strategy justify the referee time, and the motion claim is fixable if the authors can show that the later epochs were sensitive enough at those coordinates. I would send it to review with that demand attached.","headline":"A genuinely new far-infrared search with two follow-up-worthy candidates, but the motion inference rests on an unquantified six-month non-detection.","tokens_in":18193,"tokens_out":3282,"would_cite":false,"duration_ms":33555,"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 paper claims that two far-infrared sources in the AKARI all-sky survey, each seen twice within a day and missing from scans six months later, are viable Planet Nine candidates with fluxes matching a 6-12 Earth-mass planet at 300-900 au.","keywords":["Planet Nine","far-infrared astronomy","AKARI survey","moving object search","trans-Neptunian objects","Kuiper belt","parallax"],"falsifier":"Re-analyze the AKARI detection probability maps at the positions of FISSSDL J0250422-150114 and FISSSDL J0301112-164240 in the six-month-separated scans, injecting a fake 0.5 Jy point source to measure the local detection limit; if such a stationary source would be confidently detected, the absence supports motion, and if not, the candidates could be stationary objects.","tokens_in":17115,"feed_emoji":"🪐","tokens_out":10013,"duration_ms":88345,"temperature":0.7,"pith_summary":"This paper claims that a dedicated search of the AKARI far-infrared all-sky survey has found two Planet Nine candidates: point-like sources detected twice within 24 hours and absent from scans six months later, as expected for a distant object whose apparent position shifts by parallax. The search works in the thermal far-infrared rather than reflected optical light, where a planet's signal falls off as distance squared instead of distance to the fourth power, allowing a wider volume to be probed. After excluding known sources, cirrus contamination, cosmic-ray artifacts, and fast-moving objects, the remaining two sources have 90-micron fluxes of 0.5-1.6 Jy, consistent with a 6-12 Earth-mass planet at 300-900 au from the Sun with a temperature of 28-53 K. The paper argues these are the strongest far-infrared Planet Nine candidates to date and that they warrant follow-up observations, though it stops short of claiming a confirmed detection.","feed_headline":"Far-infrared search yields two Planet Nine candidates","feed_subtitle":"Each was detected twice in a day and absent after six months, the signature of a distant planet.","key_machinery":"The central mechanism is the AKARI-FIS Single Scan Detection List (FISSSDL), a source list built from individual satellite scans without the standard requirement that a source be confirmed in multiple scans separated by months, a requirement that would automatically reject a moving object like Planet Nine. To exploit this list, the authors use the fact that a planet at 300-900 au shifts by 10-25 arcminutes over six months due to parallax, well above the 3.5-arcsecond positional accuracy, while its shift in one hour is only about 0.23 arcseconds. Candidates must therefore be detected at least twice within 24 hours, meaning they are stationary on the short timescale, but must have no detection at the same position six months later, meaning they move on the long timescale. Flux measurements use the most sensitive AKARI band, WIDE-S at 90 microns, with quality cuts FLUX90/FERR90 > 5 and background strength BG90 < 0.2 to reject cirrus contamination, followed by visual inspection of detection probability maps to eliminate cosmic-ray artifacts.","core_discovery":"On its own terms, the paper's central discovery is the identification of two infrared sources, FISSSDL J0250422-150114 and FISSSDL J0301112-164240, in the AKARI Single Scan Detection List that pass every filter designed to select a moving outer Solar System planet. Both were detected twice within 24 hours in the 90-micron WIDE-S band with no counterpart at the same position in scans six months later, which the authors interpret as the signature of a planet's six-month parallax. Their flux densities, measured per scan, lie between 0.5 and 1.6 Jy, matching the black-body emission predicted for the mass, distance, and temperature ranges taken from the Millholland-Laughlin 2017 simulation and Cowan-Holder-Kaib 2016 thermal model. The paper presents these as viable candidates consistent with theoretical predictions, not as a confirmed detection, and recommends follow-up to determine orbits and confirm the Planet Nine hypothesis.","pith_inferences":["The paper's 'no detection after six months' criterion is used as evidence of motion, but the later-epoch detection probability maps at the candidate positions are never calibrated; a stationary 0.5 Jy source might have been missed, so a re-analysis with injected fake sources would test this directly.","The two candidates show flux variations of a factor of about 2.5-2.6 between their two detections within 24 hours; if these variations are real rather than noise, they are unusual for a distant black-body source and could indicate a variable or transient object instead.","The same single-scan selection applied to the full FISSSDL, not just the ML17 region, might reveal additional moving far-infrared sources, including previously unrecognized asteroids or other distant planets.","If either candidate is a real planet, the far-infrared thermal approach would become a viable complement to optical reflected-light searches, since thermal flux falls as distance squared rather than to the fourth power, extending the detectable volume."],"forward_implications":["If either candidate is confirmed by follow-up, it would be the first direct detection of Planet Nine and would support the dynamical explanation for the clustering of extreme Kuiper belt objects.","The search probes distances up to roughly 800 au for a 53 K planet, beyond the reach of previous optical surveys that exclude orbits brighter than V~21.","The two candidates are within reach of a few pointings with the Subaru Hyper Suprime-Cam, which can detect a 26th-magnitude point source, so confirmation or rejection is observationally feasible.","If both candidates are ruled out, the survey's non-detections still place upper limits on the 90-micron flux of Planet Nine in the ML17 sky region, constraining its allowed mass and distance."],"supporting_citations":[{"why":"Supplies the promising sky region (30-50 deg R.A., -20 to 20 deg Dec.) and the 6-12 Earth-mass range used in the candidate selection.","marker":"Millholland and Laughlin 2017"},{"why":"Provides the 28-53 K temperature range for Planet Nine and the parallax and proper motion equations used to derive the expected angular motion.","marker":"Cowan, Holder, and Kaib 2016"},{"why":"Establishes the Planet Nine hypothesis and the initial orbital and mass predictions that motivate the search.","marker":"Batygin and Brown 2016"},{"why":"Provides updated orbital parameters and is used to justify choosing the more distant ML17 parameter space over the largely excluded Brown-Batygin orbits.","marker":"Brown and Batygin 2021"},{"why":"Combined optical surveys that rule out Planet Nine orbits brighter than V~21 (about 500 au), leaving the more distant ML17 region unexplored.","marker":"Brown, Holman, and Batygin 2024"},{"why":"Previous far-infrared search requiring IRAS detections; this work extends to fainter fluxes using only AKARI single scans.","marker":"Sedgwick and Serjeant 2022"},{"why":"Prior IRAS-based search that found one candidate inconsistent with most predictions; provides context for the far-infrared approach.","marker":"Rowan-Robinson 2021"},{"why":"Defines the AKARI FISBSC construction and the 32-arcsecond distinctness criterion used in cross-matching.","marker":"Yamamura et al. 2010"},{"why":"Describes the AKARI Far-Infrared Surveyor and its four filters, establishing the 90-micron WIDE-S band as the data source.","marker":"Kawada et al. 2007"}],"fun_headline_variants":["Infrared sky survey flags two Planet Nine candidates","Two faint infrared blips could be Planet Nine","AKARI far-IR data yield two Planet Nine candidates","Planet Nine? Two candidates from far-infrared search"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper treats a source's absence from scans six months later as evidence that it moved, but it never shows that a stationary source of the same brightness would have been detected in those later scans, so a missed stationary object could masquerade as a moving candidate.","fun_headline_variants_meta":{"raw":{"variants":["Infrared sky survey flags two Planet Nine candidates","Two faint infrared blips could be Planet Nine","AKARI far-IR data yield two Planet Nine candidates","Planet Nine? Two candidates from far-infrared search"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1420,"prompt_tokens":1057,"completion_tokens":363,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":673,"completion_tokens_details":{"reasoning_tokens":302}},"tokens_in":673,"tokens_out":363,"duration_ms":4493,"temperature":1.0,"reasoning_tokens":302,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:37:47.102058+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-analyze the AKARI detection probability maps at the positions of FISSSDL J0250422-150114 and FISSSDL J0301112-164240 in the six-month-separated scans, injecting a fake 0.5 Jy point source to measure the local detection limit; if such a stationary source would be confidently detected, the absence supports motion, and if not, the candidates could be stationary objects.","supporting_citations":[],"review_version":1}